For your Final Project, you will have to find four (4) peer reviewed articles that will be supportive of your topic that you introduced in Week 2.
For this paper you must include:
1-2 paragraph summary of each peer review article.
Include the relevancy to your topic;
The meaning or implication of the results of the study that the article is about.
1 paragraph discussing how the author could expand on the results, what the information means in the big picture, what future research should focus on or how future research could move the topic forward.
Discuss your data source as it relates to your topic.
Include how these peer review articles add to your data source and influence the research of your topic.
attached are the 4 articles that need to be completed
Peer Review Article Summary
Due: Midnight Sunday of Unit 4
For your Final Project, you will have to find four (4) peer reviewed articles that will be supportive of your topic that you introduced in Week 2.
For this paper you must include:
· 1-2 paragraph summary of each peer review article.
· Include the relevancy to your topic;
· The meaning or implication of the results of the study that the article is about.
· 1 paragraph discussing how the author could expand on the results, what the information means in the big picture, what future research should focus on or how future research could move the topic forward.
· Discuss your data source as it relates to your topic.
· Include how these peer review articles add to your data source and influence the research of your topic.
Peer Review Article Summary Rubric
(0 – 5 Points)
Proficient to Development Needed
(6 – 8 Points)
(9 – 10 Points)
Summary of peer review articles and data source
Missing summary of peer review articles missing or not appropriately outlined. Data source is also missing or not clearly addressed.
Summary of articles is present along with the data source, but parts are unclear, require further development, or are insufficiently cited.
Summary of articles and the data source is fully described, appropriately applied, and cited.
References missing or not appropriate caliber.
Fewer than 4 peer reviewed references cited.
4 or more peer reviewed references cited.
Clear and professional writing and format
Errors impede professional presentation; guidelines not followed
Few errors that do not impede professional presentation
Writing and format is clear, professional, APA compliant, and error free
Civil Forfeiture, Crime Fighting
and Safeguards for the Innocent
An Analysis of Department of Justice Forfeiture Data
By Jennifer McDonald
In July 2017, the Department of Justice (DOJ) under then-Attorney General Jeff Sessions revived
a controversial federal forfeiture program the previous administration had sharply curtailed:
“adoptive forfeitures” or “adoptions.”1 In the months since, the DOJ has offered an ardent
defense of adoptions and of civil forfeiture in general.2 It claims 1) civil forfeiture
overwhelmingly targets criminals, not innocents, and is thus a valuable crime-fighting tool and
2) the DOJ’s forfeiture program has new “safeguards” in place to ensure adoptions do not cause
innocent people to unjustly lose their properties.3 Based on our best understanding of the DOJ’s
own data—which only partially illuminate the black box that is the federal forfeiture process—
the Institute for Justice (IJ) finds reason to doubt these claims. Specifically, we find that the DOJ
cannot substantiate its claim that civil forfeiture fights crime. We also conclude that the DOJ’s
new safeguards are unlikely to prevent innocent people from losing their property unjustly.
A Brief History of Federal Civil Forfeiture and Its Revitalization
Civil forfeiture is a mechanism that allows law enforcement agencies to seize property on the
mere suspicion that it is connected to a crime and forces property owners to go up against the
government to try to win it back. In contrast to criminal forfeiture, where property owners can
lose their property only after a criminal conviction, civil forfeiture allows law enforcement to
take property from innocent people who have never been formally charged with a crime, let
alone convicted of one. This evasion of the criminal justice system is based on a legal fiction in
which property alleged to be connected to a crime is deemed “guilty” of having somehow
assisted in committing that crime. In criminal forfeiture, the government proceeds against the
person charged with a crime; in civil forfeiture, the government proceeds against the property.4
This feature of American law can be traced to 17th-century English maritime law, which
permitted courts to obtain jurisdiction over property when it was virtually impossible to obtain
1 U.S. Department of Justice Office of Public Affairs. (2017, July 19). Attorney General Sessions issues policy and
guidelines on federal adoptions of assets seized by state or local law enforcement [Press release]. Retrieved from
seized-state; U.S. Department of Justice Office of Public Affairs. (2015, January 16). Attorney General prohibits
federal agency adoptions of assets seized by state and local law enforcement agencies except where needed to
protect public safety [Press release]. Retrieved from https://www.justice.gov/opa/pr/attorney-general-prohibits-
2 Rosenstein, R. J. (2017, November 9). Bernie Madoff and the case for civil asset forfeiture. The Wall Street
Journal. Retrieved from https://www.wsj.com/articles/bernie-madoff-and-the-case-for-civil-asset-forfeiture-
1510237360; U.S. Department of Justice Office of Public Affairs. (2018, February 12). Attorney General Sessions
offers remarks to the National Sheriffs’ Association [Press release]. Retrieved from
3 U.S. Department of Justice Office of Public Affairs, 2017.
4 For a detailed primer on civil forfeiture, please see Carpenter, D. M., Knepper, L., Erickson, A. E., & McDonald, J.
(2015). Policing for profit: The abuse of civil asset forfeiture (2nd ed.). Arlington, VA: Institute for Justice.
Retrieved from http://ij.org/wp-content/uploads/2015/11/policing-for-profit-2nd-edition.pdf
jurisdiction over owners—pirates, for example—who had violated the law.5 For much of the
nation’s history, civil forfeiture was something of a legal backwater. However, its use expanded
greatly during the early 1980s with the War on Drugs. Today, unmoored from the practical
necessities of enforcing maritime law, civil forfeiture is a popular tool of law enforcement.6
This is concerning for at least two reasons. First, because civil forfeiture proceedings are civil
actions against property, not criminal actions against people, owners caught up in them lack
rights afforded the criminally accused. Owners do not have the right to counsel, and the federal
government faces a lower evidentiary threshold to forfeit property than it does to convict a
person of a crime. Even people who had nothing to do with an alleged crime can lose their
property through civil forfeiture unless they can prove their own innocence—flipping the
American legal tradition of innocent until proven guilty on its head. Second, civil forfeiture laws
often give law enforcement agencies a financial stake in forfeitures by awarding them some, if
not all, of the proceeds of successful forfeiture actions. This financial incentive creates a conflict
of interest and encourages the pursuit of property instead of the pursuit of justice.7
Not only do federal law enforcement agencies seize property for forfeiture on their own, but they
also partner with state and local agencies through a program called “equitable sharing.”8 There
are two ways this partnership plays out: joint investigations and adoptions. Joint investigations,
which make up the majority of equitable sharing,9 allow federal agents to partner with state and
local agents to seize and forfeit property under federal law through, for example, a joint task
force. Adoptions occur when state and local agents seize property under state law and then turn it
over to the federal government, which then “adopts” the property and forfeits it under federal
Equitable sharing adds yet another layer to the problem of civil forfeiture because it allows state
and local agencies to circumvent state laws that make it more difficult or less profitable for law
enforcement to forfeit property. For example, the burden of proof required to forfeit property is
higher under New York state law than it is under federal law. And while New York state law
allows seizing agencies to keep a maximum of 60 percent of the proceeds from forfeited
property, the equitable sharing program allows agencies to receive up to 80 percent.11
Unsurprisingly, New York agencies are among the most avid participants in equitable sharing
nationwide.12 North Carolina presents an even more extreme example: The state has completely
5 Boudreaux, D. J., & Pritchard, A. C. (1996). Civil forfeiture and the war on drugs: Lessons from economics and
history. San Diego Law Review, 33, 79–135.
6 Rainbolt, G., & Reif, A. F. (1997). Crime, property, and justice: The ethics of civil forfeiture. Public Affairs
Quarterly, 11(1), 39–55.
7 See Carpenter et al., 2015.
8 While a number of federal agencies engage in forfeiture and equitable sharing, this white paper focuses only on
agencies that participate in the Department of Justice’s Asset Forfeiture Program, which are listed at
9 Between 2000 and 2013, joint investigations accounted for 73 percent of DOJ equitable sharing seizures and 82
percent of the proceeds from DOJ equitable sharing forfeitures. Carpenter et al., 2015, pp. 29, 149.
10 Carpenter et al., 2015.
11 Carpenter et al., 2015, pp. 110, 25.
12 Carpenter et al., 2015, p. 111.
eliminated civil forfeiture and directs all proceeds from criminal forfeiture to public schools.
Likely as a result, state and local law enforcement agencies actively participate in the DOJ’s
equitable sharing program, bringing in more than $11 million annually.13 And scholarly research,
including a 2018 article published in the journal Criminology and Public Policy, confirms this
phenomenon, finding that law enforcement agencies in states with more restrictive and less
financially rewarding civil forfeiture laws receive more equitable sharing dollars.14
Former Attorney General Eric Holder severely curtailed adoptive equitable sharing in January
2015 following public outcry over the practice.15 The DOJ’s data suggest that federal adoptions
virtually ended after Holder instituted this policy change (see Table 1 and Figure 1).
Table 1: Total Value of Property Seized by the DOJ, Joint Investigations vs.
Adoptions, January 17, 2014–January 16, 2016
Percent of Total
Percent of Total
1 Year Pre-Holder
$516,487,639 89% $65,011,586 11%
1 Year Post-Holder
$427,278,555 99.996% $15,628 0.004%
13 Carpenter et al., 2015, pp. 112–13.
14 Holcomb, J. E., Williams, M. R., Hicks, W. D., Kovandzic, T. V., & Meitl M. B. (2018). Civil asset forfeiture
laws and equitable sharing activity by the police. Criminology and Public Policy, 17(1), 101–27; Carpenter et al.,
2015, pp. 25–30.; Holcomb, J. E., Kovandzic, T. V., & Williams, M. R. (2011). Civil asset forfeiture, equitable
sharing, and policing for profit in the United States. Journal of Criminal Justice, 39(3), 273–85.
15 U.S. Department of Justice Office of Public Affairs, 2015; Sallah, M., O’Harrow, R., Jr., Rich, S., Silverman, G.,
Chow, E., & Mellnik, T. (2014, September 6). Stop and seize. The Washington Post. Retrieved from
Figure 1: Total Number of DOJ Adoptions, January 17, 2014–January 16, 2016
In July 2017, the DOJ under then-Attorney General Jeff Sessions lifted the limits on adoptions
with the stated intent of shoring up federal crime-fighting efforts and resources.16 Perhaps in
response to anticipated criticism,17 the DOJ added “safeguards” it said would protect innocent
property owners from unwarranted adoptions. The safeguards include, among others, requiring
1) speedier notice of seizure for property owners, 2) additional documentation of the probable
cause that led to seizures, and 3) additional scrutiny of cash seizures of $10,000 or less.18 IJ’s
analysis of federal forfeiture data casts doubt on the DOJ’s claim that civil forfeiture fights
crime. It also sheds light on the DOJ’s past behavior, giving reason to doubt that new safeguards
will truly protect property owners going forward.
16 U.S. Department of Justice Office of Public Affairs, 2017.
17 Wilson, J. J. (2017, July 19). Civil forfeiture is inherently abusive: Statement responding to DOJ’s announced
policy change increasing use of civil forfeiture [Press release]. Arlington, VA: Institute for Justice. Retrieved from
18 U.S. Department of Justice Criminal Division. (2017). Policy guidance on the Attorney General’s order on federal
adoption and forfeiture of property seized by state and local law enforcement agencies (Policy Directive 17-1).
Retrieved from https://www.justice.gov/criminal-mlars/file/985636/download
To examine the DOJ’s claims that civil forfeiture overwhelmingly targets criminals and that new
“safeguards” will protect innocent owners, IJ used the DOJ’s own data. The DOJ keeps track of
the property its agencies seize and forfeit using a database called the Consolidated Asset
Tracking System (CATS). CATS contains information on the type and value of property seized,
whether property was seized as part of a joint investigation or adoption, what kind of legal
procedures were used to forfeit the property, and much more.19 We initially obtained the CATS
database from the DOJ through a Freedom of Information Act request, but the database is now
publicly accessible on the DOJ’s website.20
For this analysis, we examined CATS data on cash seizures conducted between 1997 and 2015.21
We examined all of the DOJ’s cash seizures—as opposed to just those conducted as part of the
equitable sharing program—because such an analysis provides an indication of the DOJ’s
policies more generally.22 We find that the DOJ cannot substantiate its claim that civil forfeiture
is an essential crime-fighting tool because it does not even track whether its civil forfeiture
efforts are advancing criminal investigations. We also conclude, based on past behavior by the
DOJ and its agencies, that the new safeguards are unlikely to protect innocent people from losing
their property unjustly.
The DOJ does not track, and therefore cannot substantiate, whether its civil forfeiture
efforts are advancing criminal investigations.
The DOJ argues that civil forfeiture is an important tool to fight crime and combat the illegal
drug trade.23 But these claims are difficult, if not impossible, to substantiate because the DOJ
does not track in CATS whether criminal charges or convictions accompany the seizure and
forfeiture of property. Not one of the more than 1,300 variables contained in CATS indicates
19 U.S. Department of Justice Justice Management Division. (2014). Major information systems. Retrieved from
20 U.S. Department of Justice Asset Forfeiture Program. (2017). Freedom of Information Act. Retrieved from
https://www.justice.gov/afp/freedom-information-act. For this analysis, we used the version of the database that was
up to date as of the end of Fiscal Year 2016.
21 This is the longest time period for which we had complete calendar-year data at the time of analysis. We limited
our analysis to cash seizures because we believe they are the most accurately valued assets in CATS. With other
types of property, such as vehicles or electronics, a certain amount of error is likely introduced when the estimated
value is entered into the database. On the other hand, $5,000 in cash has a clear value of $5,000, making for a
22 While the results presented here reflect analyses conducted on all DOJ cash seizures, we ran the same analyses on
the subset of seizures conducted as part of the equitable sharing program. Differences in results were negligible.
23 The DOJ also claims forfeiture allows it to return substantial sums of forfeiture proceeds to crime victims: $4.4
billion since 2000, half of which the DOJ says came from civil forfeiture. Rosenstein, 2017; U.S. Department of
Justice Office of Public Affairs, 2017; C-SPAN. (2017, July 29). Attorney General announcement on civil asset
forfeiture [Video file]. Retrieved from https://www.c-span.org/video/?431550-1/attorney-general-announces-civil-
asset-forfeiture-directives. But $2.2 billion is less than 9 percent of the more than $25 billion the DOJ reportedly
forfeited between fiscal years 2001 and 2016.
whether criminal charges were ever filed in conjunction with a seizure.24 This leaves open the
possibility that innocent people are losing their property unjustly. The DOJ’s own Office of the
Inspector General (OIG) has come to similar conclusions.
Concerned about civil liberty abuses from the DOJ’s forfeiture activity, the OIG conducted its
own investigation into the quality of the DOJ’s oversight of its agencies’ forfeiture activity.
Because the DOJ does not track information about crimes associated with seizures, the OIG was
forced to go directly to the source: It examined Drug Enforcement Administration (DEA) records
for a sample of 100 cash seizures that were executed without a warrant and without the presence
of drugs—indicators that these seizures were likely to violate property owners’ civil liberties. It
found that in more than half of the cases, the DEA could not verify whether warrantless
interdiction seizures advanced a criminal investigation, warning:
When seizure[s] and administrative forfeitures do not ultimately
advance an investigation or prosecution, law enforcement creates
the appearance, and risks the reality, that it is more interested in
seizing and forfeiting cash than advancing an investigation or
The DOJ’s new “safeguards” fail to address the most fundamental problems with federal
civil forfeiture and are unlikely to offer meaningful protection for innocent property
The DOJ touts new “safeguards” within the revived adoptions program it says will protect
innocent owners from losing property unjustly. These safeguards include requiring 1) speedier
notice of seizure for property owners, 2) additional documentation of the probable cause that led
to seizures, and 3) additional scrutiny of cash seizures of $10,000 or less.26 Analysis of past DOJ
forfeiture data suggests that these safeguards will not do enough to protect innocent property
owners because they fail to address three significant problems with federal civil forfeiture:
• Filing a claim for return of property is an uphill battle, increasing the likelihood that
owners will lose their properties without cause.
• Most seizures are adjudicated administratively, not judicially, meaning the government
does not have to prove property’s connection to a crime before a neutral judge.
• The DOJ’s internal processes do not provide a meaningful check on the power of
24 Erickson, A. C., McDonald, J., & Menjou, M. (n.d.). Department of Justice report card. Forfeiture transparency
and accountability: State-by-state and federal report cards. Arlington, VA: Institute for Justice. Retrieved April 24,
2018, from http://ij.org/report/forfeiture-transparency-accountability/?state=DOJ. For a complete list of variables
contained in CATS, see https://www.justice.gov/afp/page/file/441201/download
25 U.S. Department of Justice Office of the Inspector General. (2017). Review of the Department’s oversight of cash
seizure and forfeiture activities. Retrieved from https://oig.justice.gov/reports/2017/e1702.pdf, p. iii.
26 U.S. Department of Justice Criminal Division, 2017.
1. Filing a claim for return of property is an uphill battle, increasing the likelihood that
owners will lose their properties without cause.
The Attorney General’s office boasts that “[a]bout 80% of the time, nobody even tries to claim
the seized assets” and suggests that this is because “[m]ost cases are indisputable.”27 Aside from
the fact that the DOJ cannot substantiate this assertion, this statement ignores the host of
obstacles confronting owners of seized property that deter many from even attempting to fight
the federal government.
When the DOJ seizes or adopts property, it is required to provide the property owner with
written notice of the seizure within 90 days.28 The AG’s office argues that cutting this time in
half will help innocent owners challenge the seizure of their property more promptly,29 but such
a policy change will do nothing to streamline the daunting process required to claim seized
property. Unlike in criminal proceedings, victims of civil forfeiture are not entitled to an
attorney. Claimants must therefore either spend thousands of dollars to retain counsel—often
more than the value of the seized property at stake—or attempt to navigate the byzantine federal
forfeiture process alone (see Figure 2). Faced with these hurdles, many property owners may feel
their only option is to walk away.
27 Rosenstein, 2017.
28 18 U.S.C. § 983(a)(1)(A)(iv).
29 Rosenstein, 2017
Figure 2: Federal Forfeiture Flow Chart30
Even if a property owner decides to take on the federal government and file a claim for return of
property, doing so correctly is difficult, though the DOJ presents it as a fairly straightforward
process. According to the DOJ’s website, once property owners receive notice of the
government’s intent to forfeit their property, all they have to do is mail in a claim to the seizing
agency. Claims must:
• Be in writing.
• Describe the seized property.
• State ownership or other interest in the property.
• Be made subject to penalty of perjury.
• Be made before the deadline indicated on the seizure notice.31
30 Figure reproduced from Heritage Foundation. (2015). Arresting your property: How civil asset forfeiture turns
police into profiteers. Retrieved from http://thf_media.s3.amazonaws.com/2015/pdf/Forfieture-Booklet-FINAL-
Full.pdf, pp. 10–11.
31 U.S. Department of Justice. (n.d.). Claims. Retrieved April 24, 2018, from
But this process is rife with pitfalls, and running into one can cause a claim to be deemed
deficient—and property to be forfeited automatically. Indeed, a large portion of claims are
deemed deficient by the government: one-fifth of all claims made between 1997 and 2015 and
one-third of all claims made for seized cash during that time (see Figure 3).
Figure 3: Portion of Claims Deemed Deficient under Justice Asset Forfeiture
The biggest pitfall is that the seizing agency itself—and not a judge or other neutral arbiter—gets
to determine whether a claim is deficient. And it can deem a claim deficient for a variety of
reasons, as shown in Table 2. By far the most common reason is that a claim was “not executed
and sworn to by the claimant.” Overall, 68 percent of deficient claims for seized cash fell into
this trap. The DEA, which conducts the lion’s share of the seizures recorded in CATS, has
proved to be particularly fastidious about this. For example, it has in the past deemed claims
sworn before a public notary deficient because they do not explicitly say they were sworn
“subject to penalty of perjury,” even though courts have frequently ruled that such language is
not necessary to support a perjury conviction.32
The next largest share of deficient claims—a full 16 percent—were so deemed for “other”
reasons. This finding reveals that the claims process is often a black box.
32 United States v. Jones, No. CR 12-10084-PBS, 2017 U.S. Dist. LEXIS 13147, 2017 WL 421644 (D. Mass. Jan.
Claims for Seized Cash
Table 2: Reasons for Deficient Claims, DOJ Cash Seizures 1997–2015
Percent of Claims
Deemed Deficient Reason for Deficiency
68% Claim was not executed and sworn to by the claimant
5% Claimant neglected to state their interest in such property
4% Claimant did not state intent to contest seizure in District Court
3% Claimant did not state claim of ownership
1% Property was not properly identified by the claimant
1% Claim of indigency not adequately supported
1% In forma pauperis claim insufficient – no affidavit of indigency
0.4% Claim was received without a bond
0.3% Claim submitted was in photocopy and not original
0.1% Bond not payable to U.S. DOJ, returned to claimant
0.1% Insufficient bond received, returned to claimant
2. Most seizures are adjudicated administratively, not judicially, meaning the government
does not have to prove property’s connection to a crime before a neutral judge.
Civil forfeiture cases can be adjudicated in one of two ways: judicially or administratively. With
judicial procedures, the property’s fate is decided, at least in theory, by a judge who serves as a
neutral arbiter weighing the merits of the government’s case against the property owner’s claim.
With administrative procedures, in contrast, law enforcement officials decide whether to forfeit,
allowing them to play the role of judge. Three-quarters of all federal forfeitures are adjudicated
According to the DOJ, filing a claim for the return of seized property automatically triggers
judicial proceedings.34 When property is not claimed, the DOJ generally assumes it is
The DOJ argues that administrative forfeiture is so common because most people are guilty and
know they have no case; therefore, they do not file a claim.36 Yet, as already noted, the DOJ does
33 This rate holds true for all property types, not just cash.
34 U.S. Department of Justice. (n.d.). Civil asset forfeiture backgrounder, p. 2. Copy on file with the Institute for
Justice. The data cast doubt on the DOJ’s claim, however, suggesting that the reality of the federal civil forfeiture
process is less cut and dried: In over 9,000 cases between 1997 and 2015, cash was forfeited administratively even
though someone filed a claim for its return. With some of these forfeitures, the “forfeiture type” switched between
judicial and administrative proceedings at least once before the cash was finally forfeited administratively. Although
we cannot be sure why these changes occur—or indeed why claimed property is ultimately forfeited
administratively, in conflict with the DOJ’s public statements about the process—it might be because claimants
settle with the government or drop out of the onerous process. Either way, the data suggest that even if judicial
proceedings are begun, it is still possible for a case to be decided by DOJ officials rather than a federal judge. Our
analysis uses the most recent forfeiture type assigned to an asset at the time it was forfeited.
35 U.S. Department of Justice, 2017, p. 2.
not track the data necessary to back up this argument, and OIG findings cast doubt on its
veracity. Moreover, as detailed in the previous section, filing a claim can be an uphill battle and
there are many reasons why a person may be unable or unwilling to fight for the return of their
property even if they are innocent of any wrongdoing. It is therefore likely that innocent owners
face administrative forfeiture, with the fate of their property up to the discretion of law
In theory, administrative forfeiture requires law enforcement officials to weigh the evidence
linking a property to a crime. Even in the absence of a claim, officials are supposed to consider
whether property is subject to forfeiture before seizing and forfeiting property administratively.
However, the data give reason to question whether this is truly happening in practice: Between
1997 and 2015, just 2 percent of cash seizures for administrative forfeiture resulted in the cash
(or a portion of it) being returned to its owner. In contrast, fully one-third of cash seizures for
judicial forfeiture resulted in a return of the money. It is possible that innocent owners are more
likely to fight forfeiture, leading to a higher return rate for judicial forfeitures, but it is unlikely,
given the obstacles to filing a claim, that this can fully account for the difference.37 Indeed, that
so many owners with the opportunity and means to fight win some or all of their property back
calls into question whether seizure standards are too lax in the first place. It also suggests that the
presence of a neutral arbiter makes a difference in forfeiture outcomes.
3. The DOJ’s internal processes do not provide a meaningful check on the power of
The DOJ claims that strengthening some of its internal processes for adoptions amounts to new
“safeguards” that will further protect innocent property owners. Specifically, it says that it will
require additional documentation of the probable cause that led to a seizure and additional
scrutiny for lower-value seizures. Using the CATS database and building on past research, our
analysis of how these safeguards have worked in the past gives reason to doubt that they have
provided meaningful protection or that they will do so in the future, even once strengthened.
A. Documentation of probable cause
In order to seize property, law enforcement must have probable cause. This is the same low
evidentiary burden that police must meet before searching a car at a traffic stop without a warrant
and a lower one than that required to forfeit property federally and in nearly all states.38 The DOJ
says that it will now require additional documentation of the probable cause that gave rise to the
seizure of property that is later adopted. While a step in the right direction, ensuring probable
cause existed for a seizure is a basic law enforcement responsibility and no more than what
agencies should already be doing. Additional documentation might weed out the most blatantly
36 Rosenstein, 2017; C-SPAN, 2017.
37 And, as previously mentioned, some properties are forfeited administratively even though a claim is filed. The
data do not allow us to ascertain whether this is because some sort of settlement agreement was reached, because
claimants dropped out of the process for reasons that had nothing to do with their guilt or innocence, or because of
some other reason or process to which the public is not privy.
38 Carpenter, et al., 2015, p. 17
unjust adoptions, but it does not constitute a drastic improvement to a fundamentally unjust
B. Additional scrutiny of lower-value seizures
When the DOJ reinstated adoptions, it also reinstated a rule requiring that lower-value cash
seizures receive additional scrutiny and raised the threshold for this additional scrutiny from
$5,000 to $10,000. Although intended to provide protection for innocent property owners, this
policy is troubling because it implies that the DOJ assumes the more cash a person is carrying,
the more likely it is that the cash is related to criminal activity.
But it is not a crime to carry even large amounts of cash. Indeed, there are any number of reasons
a person might carry a large amount of cash that do not involve criminal activity, such as raising
money for charity,39 buying or selling property,40 working in a primarily cash-based business,41
or avoiding credit card debt.42 Moreover, members of certain populations, including many
disadvantaged groups, are less likely to have a bank account, which makes them more likely to
carry cash.43 The DOJ’s threshold is also arbitrary: It is unclear why someone carrying $10,001
deserves less due process than someone carrying $9,999.
The data also give reason to doubt the logic behind the threshold. Assuming the DOJ really has
given sub-$5,000 seizures additional scrutiny in the past, the data show at most a slight
difference in return rates for cash seizures above and below the threshold (see Table 3). If
higher-value seizures were truly more likely to be indicative of criminal activity—something
which, as discussed above, the DOJ does not even track—we might expect to see a lower return
rate for them than for lower-value seizures. The fact that we do not suggests the threshold is
arbitrary and raising it will have little effect on protecting property owners.
39 Ingraham, C. (2016, April 25). How police took $53,000 from a Christian band, an orphanage and a church. The
Washington Post. Retrieved from https://www.washingtonpost.com/news/wonk/wp/2016/04/25/how-oklahoma-
40 Liberty project intervenes, government drops all charges. (Summer 2017). DKT Liberty Project Newsletter.
Retrieved from http://dktlibertyproject.org/wp-content/uploads/2017/05/DKTLP-Summer-2017.pdf, p. 1; Kirby, B.
(2015, June 22). When cops seize money or property in Alabama, it’s owner’s burden to prove he’s innocent.
AL.com. Retrieved from
41 O’Harrow, R., Jr., Sallah, M., & Rich, S. (2014, September 8). They fought the law. Who won? The Washington
Post. Retrieved from http://www.washingtonpost.com/sf/investigative/2014/09/08/they-fought-the-law-who-
42 The good and bad of carrying cash: Does it really pay? (n.d.). Retrieved April 24, 2018, from
43 Nine million U.S. households did not have a bank account in 2015, with Hispanic, black, working-age disabled,
lower-income, less-educated and younger households more likely to be in this group than the rest of the population.
Federal Deposit Insurance Corporation Division of Depositor and Consumer Protection. (2016). 2015 FDIC national
survey of unbanked and underbanked households: Executive summary. Retrieved from
Table 3: Return Rates Above and Below $5,000, DOJ Cash Seizures, 1997–2015
<$5,000 ≥$5,000 <$5,000 ≥$5,000 Return Rate 44% 34% 2% 1% Where a difference in return rates is observed, however, is between judicial and administrative forfeitures. For example, while seizures processed judicially both above and below the threshold are returned in over one-third of cases, those processed administratively are hardly ever returned. This discrepancy is further evidence that a judge’s involvement in a forfeiture case can mean all the difference to an innocent property owner. Conclusion: The deck is stacked in favor of the government and new “safeguards” are unlikely to protect property owners. The DOJ assumes that the low rate of claims made for return of property is evidence that its civil forfeiture program ensnares primarily the criminals it is intended to target. But IJ’s analysis offers an alternative explanation for this trend: The DOJ and federal civil forfeiture law make it very difficult for innocent owners to get their property back. The civil forfeiture process is designed in such a way that the government has the upper hand at each stage, forcing property owners to navigate a nearly unintelligible legal process, prove their own innocence and do it all without the benefit of a lawyer if they cannot afford to hire one (or if the seized property is worth less than the cost of an attorney). Making matters worse, the vast majority of this activity happens without the oversight of a judge or jury. Modest modifications to internal procedures are unlikely to do much to protect property owners when the system itself and the incentives it provides police and prosecutors remain unchanged. As long as the decision to forfeit property rests with the very individuals who stand to benefit from forfeiting it, administrative “safeguards” will likely offer little to no meaningful protection against an energized federal forfeiture machine. Forfeiture White Paper_final_11-19-2018 - with cover Forfeiture White Paper_final_5-2-2018 - with cover White-Paper-Evergreen Forfeiture White Paper_November 2018 update - no cover Mark Gius Justice Policy Journal, Spring, 2018 The Effects of Civil and Criminal Forfeitures 1 The Effects of Civil and Criminal Forfeitures on Drug-Related Arrests Mark Gius1 Justice Policy Journal Volume 15, Number 1 (Spring, 2018) © Center on Juvenile and Criminal Justice 2018 www.cjcj.org/jpj Abstract The purpose of the present study is to determine if civil and criminal forfeitures have statistically-significant and negative effects on drug-related arrests. The primary focus of this paper will be on the deterrent effects of forfeitures. Using a random effects model and state-level data for the period 2000-2013, it was found that there is a negative relationship between the per capita value of seized assets and the drug-related crime rate. It is important to note, however, that the effect is very minimal; even if the per capita value of seized assets was doubled, the drug- related arrest rate would fall by only 0.05336%. Hence, given the constitutional issues surrounding civil forfeitures and the minimal effects of such forfeitures, it would be in the public interest to amend the Comprehensive Crime Control Act (CCCA) of 1984 so that equitable sharing of forfeiture proceeds among federal, state, and local agencies would no longer be permitted. Amending the CCCA in this manner would remove the incentives that state and local agencies have to engage in seizures and forfeitures. Such a revision of the CCCA would only very minimally affect the drug-related arrest rate but would, at the same time, restore some degree of due process to forfeiture proceedings. 1 Quinnipiac University Corresponding Author: Mark Gius, [email protected] http://www.cjcj.org/jpj mailto:[email protected] 2 The Effects of Civil and Criminal Forfeitures Introduction In 2007, a waitress, her two young sons, and her boyfriend were travelling through Tenaha, Texas when they were stopped by local police. They were told that the reason they were stopped was because they were driving in the passing lane. The police officer searched their car and found a large amount of cash that the couple said they were going to use to buy a car. Nothing illegal was found. Nonetheless, local police detained them and drove them to the local police station. There, the county’s district attorney told them that they could either be charged with felony money laundering, in which case the woman would lose custody of her children, or they could sign over all of their cash to the city of Tenaha and be on their way. They decided to give up their money. Hence, even though they were never convicted or even charged with a crime, the waitress and her boyfriend had to forfeit all of their money that day in Tenaha, Texas. This story originally appeared in the New Yorker magazine (August 12, 2013) and is an excellent example of civil forfeiture. In general, there are two types of forfeitures that may be used by authorities: criminal forfeitures and civil forfeitures. Criminal forfeitures occur when authorities seize the assets of convicted persons. Civil forfeitures occur when law enforcement agencies seize the property of persons allegedly involved in criminal activities; the person involved does not need to have been convicted or even charged with a crime. For criminal forfeitures, the evidentiary standard is “beyond a reasonable doubt.” For civil forfeitures, the evidentiary standard is much lower; in some states, it is only “probable cause.” In addition, once a person’s property has been seized in a civil forfeiture, it is very difficult to recover. In many jurisdictions, the person must prove that they did not engage in the alleged criminal activity. Finally, it is important to note that, in civil forfeitures, the defendant is not the person; the defendant is the property. Hence, this results in some odd sounding case names, such as “State of Texas v. $6,037.” Until the late 20th century, the use of civil forfeitures was very rare and very limited. They were primarily used in situations involving piracy or border customs. In 1970, Congress enacted the Forfeiture Act, which allowed federal authorities to bring civil forfeiture actions against convicted drug dealers. The act was amended in 1984 (the Comprehensive Crime Control Act), and this new law allowed federal authorities to seize any property that was used or was intended to be used in drug trafficking. Although supposedly limited in scope, the Comprehensive Crime Control Act (CCCA) of 1984 has been used to justify the seizure of almost any property, even property not related in any way whatsoever to drug trafficking or any type of drug- related offense. In addition, the CCCA created the Asset Forfeiture Fund which is where the proceeds from forfeitures are deposited so that they may be used at a Mark Gius Justice Policy Journal, Spring, 2018 The Effects of Civil and Criminal Forfeitures 3 later date by law enforcement agencies. Finally, this law also allowed for the equitable sharing of forfeiture proceeds among federal, state, and local agencies. Under this program, federal authorities “adopt” local seizures, and then the proceeds are equitably shared among federal and local agencies. These “adoptions” by federal agencies typically occur in states where laws regarding the disposition of seized assets are somewhat restrictive. In some states, local law enforcement agencies can only keep a small fraction of the seized assets. By having federal authorities “adopt” the seizures, local authorities are able to circumvent these restrictive state laws and keep more of the seized assets for themselves. Therefore, the CCCA created incentives for both federal and state agencies to be much more aggressive in their seizure of assets allegedly involved in criminal activity. These increased incentives created by the CCCA have led local police departments to view the proceeds from forfeitures as relatively steady revenue streams, especially given the volatility of tax revenues available from local governments. Police departments have direct control over the seizure of properties, and given the relatively low evidentiary standards that are required for civil forfeitures, local police departments in many states view forfeiture revenues as an integral part of their budgets. The impact of civil forfeitures on police behavior has been extensively examined (Kelly and Kole, 2016; Holcomb, Kovandzic, and Williams, 2011; Bishopp and Worrall, 2009; Baicker and Jacobson, 2007; Worrall, 2001; Mast, Benson, and Rasmussen, 2000). Most of these studies examined the impact of forfeitures on policing behavior, such as the impact of forfeitures on the rate of drug-related arrests or the impact of forfeiture laws on local government budgetary decisions. In Kelly and Kole (2016), Bishopp and Worrall (2009), Baicker and Jacobson (2007), and Mast, Benson, and Rasmussen (2000), the effects of forfeitures on drug-related arrests were examined. In these studies, it was assumed that the lure of money provided by forfeitures resulted in more aggressive policing, thus increasing the rate of drug-related arrests. Because of lenient civil forfeiture policies, police focused more of their efforts on forfeiture-related arrests (typically drug-related) than on other types of arrests (Kelly and Kole, 2016). If this hypothesis is true, then an increase in forfeitures should result in an increase in drug-related and forfeiture- related arrests. Another possibility, however, is that forfeitures may have a deterrent effect on crime (Kelly and Kole, 2016). An increase in forfeitures should result in a reduction in assets available to drug dealers and other criminals. This reduction in assets 4 The Effects of Civil and Criminal Forfeitures means that criminals would have less money with which to purchase adequate legal defenses and less money to reinvest in their criminal enterprises (Kelly and Kole, 2016). Thus, we may observe a decrease in drug-related arrests whenever there is an increase in forfeiture activity. No prior research has examined the deterrent effects of forfeitures. All pertinent prior studies have tested the hypothesis that an increase in forfeitures results in an increase in drug-related arrests (policing behavior), rather than a decrease in drug arrests (deterrent effect). Hence, the purpose of the present study is to test the validity of the deterrent effect and to determine if forfeitures have statistically-significant and negative effects on drug-related arrests. As noted earlier, there has been no research as of yet that has attempted to determine the statistical significance of the deterrent effect of forfeitures. A brief review of the relevant literature will be presented in the next section. Prior Research As noted previously, several studies have examined the effects of forfeitures on police budgeting decisions (Holcomb, Kovandzic, and Williams, 2011; Worrall, 2001). Given the focus of the present study, these types of studies will not be examined in this literature review. Instead, only those prior studies that examined the effects of forfeitures on drug-related arrests will be reviewed. One of the earliest studies that examined this topic was Mast, Benson, and Rasmussen (2000). This study used data obtained from the National Institute of Justice’s Drug Use Forecasting program. This data set covers only 24 cities in the United States and looks at drug use among arrestees. The data used in this study were for the years 1987 to 1993. The results of this study suggest that law enforcement agencies increase drug arrests when the agencies are allowed to keep the proceeds from forfeitures. Several issues with this study include the use of a very limited data set (only 24 cities) and the use of data that are over 20 years old. In Baicker and Jacobson (2007), the authors examined several issues regarding forfeitures, including the impact of forfeitures on drug-related arrests. This study obtained data on federal seizures at the judicial district level for the period 1990- 1998. Data for California, Florida, Pennsylvania, Arizona, and New York were obtained for various years in the 1990s. Arrest data were obtained from the Uniform Crime Reports. Data are at the county level. Results suggest that when police are allowed to keep more of the seized assets, drug-related arrests increase. Bishopp and Worrall (2009) used data obtained from the Law Enforcement Management and Administrative Statistics (LEMAS) survey for the years 1997 and Mark Gius Justice Policy Journal, Spring, 2018 The Effects of Civil and Criminal Forfeitures 5 2000. It is important to note that the LEMAS surveys only collect data from law enforcement agencies with at least 100 officers, thus omitting many smaller law enforcement agencies. Using an OLS model, results indicate that forfeiture laws have no statistically-significant effects on drug-related arrests. Finally, Kelly and Kole (2016) used data from the LEMAS surveys for the years 2000, 2003, and 2007. Using a fixed effects model, their results suggest that increases in forfeiture activity increase drug-related arrests. However, it is important to note that the effect is very inelastic: the authors found that a 1% increase in forfeitures result in only a 0.019% increase in drug-related arrests. Hence, although forfeitures may influence policing behavior and arrest activity, the effect is minimal. Kelly and Kole (2016) conclude that the introduction of civil forfeiture for drug crimes was not one of the primary factors resulting in an increase in drug-related arrests. The present study differs significantly from this prior research in several ways. First, the primary focus of this paper will be on the deterrent effects of forfeitures and not on the positive effects of forfeitures on drug-related arrests in the 1990s and 2000s. Second, most prior research in this area used data from the LEMAS surveys and used data that were not very recent. In addition, LEMAS data is collected only at the agency level, and LEMAS is conducted only once every 3 or 4 years. The present study will use annual, state-level data for the period 2000-2013, which is much more recent than any other study on the topic of forfeitures and crime rates. Methods As noted previously, the focus of this study will be on the deterrent effects of forfeitures. All prior studies have assumed that the introduction of less restrictive forfeiture laws resulted in an increase in drug-related arrests. The reasoning behind this hypothesis is that lenient asset forfeiture laws create an incentive for police to concentrate more of their efforts on drug-related crimes, primarily because they will be able to keep some or all of the proceeds from the seizures. This increased desire to aggressively pursue individuals possibly involved in drug-related crimes should result in more drug-related arrests. An important extension of this argument, however, is that this increase in drug- related arrests may eventually increase the costs associated with engaging in drug crimes. Thus, would-be criminals may be deterred from participating in drug- related crimes, especially given that not only are police concentrating more of their 6 The Effects of Civil and Criminal Forfeitures efforts on drug crimes but also because the police may seize all of the alleged criminal’s property that is associated with the illicit drug trade (Kelly and Kole, 2016). Thus, would-be criminals may shift their focus to other types of crimes, or they may utilize additional safeguards to lessen the probability that they will be arrested. Hence, an increase in forfeiture activity may result in a reduction in drug- related arrests. In order to test this theory, the following equation will be estimated in the present study: Y = α0 + α1 Assets Seized + α2 Control Variables + α3 State Effects (1) + α4 Year Effects In equation (1), Y denotes drug-related arrests, and Assets Seized is the per capita amount of assets seized by the police in forfeitures. It is expected that the per capita amount of seized assets will be negatively related to drug-related arrests. The reason for this is because criminals would regard the increase in forfeiture activity as an increase in the costs of engaging in criminal activity. Hence, individuals will engage in less drug-related criminal activity, or they will take more safeguards in order to reduce the probability that they will be arrested. A panel data model was used to estimate Equation (1). This model is superior to both cross-sectional and time series models for two reasons. First, panel data models control for potentially important but unobservable state-level and year- specific effects that may be correlated with other determinants. If a panel data model is not used when appropriate, state-level and year-specific effects may be omitted, and omitted variable bias may result. Second, panel data, which combines time-series and cross-sectional data, greatly increases the degrees of freedom; hence, one can examine state-level data even though there may be limited annual data available. There are two ways in which a panel data model may be estimated. If one assumes that parameters estimates are independent of state-level effects, then fixed effects should be used. A fixed effects model is a classical regression model with state and year dummies. If one assumes, however, that parameter estimates vary across states, then a random effects model should be used. A random effects model allows for parameter estimate variation among states by utilizing a generalized regression model where the variance is dependent upon a state-level disturbance term. In order to determine which model is more appropriate for a given model, a Hausman Test is used. For the present study, results of the Hausman Test suggest that random effects would be the more appropriate model. Mark Gius Justice Policy Journal, Spring, 2018 The Effects of Civil and Criminal Forfeitures 7 Hence, equation (1) is estimated using a random effects model that controls for both state-level and year-specific effects. In addition, all observations are weighted using state-level population, standard errors are corrected using a clustering method (clustering is done at the state-level), and a log-linear functional form is used. The reason for weighting all observations by population is to correct for potential heteroscedasticity. Clustering standard errors is necessary in order to account for potentially nonrandom variations within certain groups. Finally, a log-linear function is used because it corrects for nonlinearities in the data. Nonlinearities are sometimes due to a dependent variable with a very large standard deviation. Control variables that are used in the estimation of equation (1) include the following: percentage of state population that is African-American, real per capita income, percentage of population that has a bachelor’s degree, unemployment rate, dummy variables denoting region of country (Northeast, South, and West), percentage of population ages 18-24, population density, per capita alcohol consumption, per capita prison population, percentage of population living in large cities, and police department employees per capita. All of these variables were used in prior studies that examined the determinants of criminal activity (Kelly and Kole, 2016; Bishopp and Worrall, 2009; Mast, Benson, and Rasmussen, 2000). Data and Results All data used in this study are at the state-level and are for the period 2000-2013. This is one of the largest and most recent data sets ever used to test the impact of forfeitures on drug-related arrests. Although prior studies in this area examined local police agencies or counties, the state is the more appropriate geographic level of analysis. The primary reason for this is because laws governing forfeitures are state laws. Thus, the rules by which police agencies can seize properties are the same for all agencies in a given state. Most prior research on this topic used LEMAS survey data, which is at the law enforcement agency level and which exclude police agencies with fewer than 100 officers. Hence, LEMAS data exclude a substantial number of police agencies, including the police department of Tenaha, Texas, which was discussed in the introduction of this paper. Using state-level data should result in a more inclusive and exhaustive data set. One issue with forfeiture data, however, is that few states report the value of seized assets. Some states report only a few years of data, while other states do not report any data at all. Unfortunately, there are also issues with federal forfeiture 8 The Effects of Civil and Criminal Forfeitures data. Civil forfeiture and criminal forfeiture data are combined in the federal surveys. A civil forfeiture, which is the primary focus of this paper, is an action brought against property. The property is the defendant, and no criminal charges against the owner of the property are necessary in order for a civil forfeiture to proceed. A criminal forfeiture is an action brought as part of the criminal prosecution of a defendant. It is an action against a person (the defendant in a criminal proceeding) and requires that the government indict the property used or derived from the crime. It would be ideal if the value of assets seized in a civil forfeiture were separate from the value of assets seized in a criminal forfeiture, but, unfortunately, the federal forfeiture data combine the values of assets seized for both types of forfeitures. Hence, it is not possible to examine only the value of assets seized in a civil forfeiture. In addition, there are two sources of data for seized assets: the U.S. Department of Justice (DOJ) and the U.S. Department of the Treasury (DOT). Unfortunately, Department of Justice data is calendar year while Department of Treasury data is fiscal year; hence, they cannot be aggregated. Given the lack of reliable data on civil forfeiture seizures, only the individual state’s share of DOJ seizures will be used in the present study. It is important to note that this data include only the state’s share of equitable sharing of forfeiture proceeds. Although the data used in this study exclude state and local agency forfeiture proceeds that were not obtained through the DOJ equitable sharing program as well as state-level equitable sharing proceeds obtained through the DOT, the data used in the present study are the best indicator of forfeiture activity in a given state. Data on state-level equitable sharing forfeiture proceeds were obtained from Carpenter, Knepper, Erickson, and McDonald (2015). This report presents data for those states that report state forfeiture activity and for DOJ and DOT equitable sharing forfeitures. The federal data lumps together both civil and criminal forfeitures. As noted previously, for purposes of this study, only the state’s share of DOJ equitable sharing proceeds will be used as a proxy for the overall level of forfeiture activity in a given state. State-level drug-related arrest data were obtained from various issues of Crime in the United States. All other data are at the state level and were obtained from various reports published by the U.S. Census Bureau. Dollar-denominated values were deflated using the Consumer Price Index-Urban, base year 1982-1984. Descriptive statistics are presented on Table 1, and bivariate correlations are presented on Tables 2-5. Regarding some of more interesting statistics on asset forfeitures and crime rates, the average per capita value of seized assets, in real dollars, was $0.58, and Mark Gius Justice Policy Journal, Spring, 2018 The Effects of Civil and Criminal Forfeitures 9 the average rate of drug-related arrests (per 100,000 persons) was 410. The per capita value of seized assets ranged from a low of $0 (Idaho in 2004) to a high of $10.39 (West Virginia in 2007). The West Virginia data appears to be an outlier as the next highest per capita value of seized assets was $1.61 (Florida in 2001). The vast majority of per capita values of seized assets were less than $1 per person. Table 1. Descriptive Statistics Variable Mean Standard Deviation Real per capita value of seized assets $0.58 $3.61 Percentage African-American 0.102 0.0945 Real per capita income $17,472 $2,694 Percentage with bachelor’s degree 0.267 0.049 Unemployment rate 0.059 0.021 Percentage ages 18-24 0.099 0.0093 Population density 192 257 Per capita alcohol consumption 2.36 0.48 Per capita prison population 431 177 Percentage in large cities 0.139 0.155 Police employees per capita 308 113 Rate of drug-related arrests 410 245 N = 690 Table 2: Bivariate Correlations for Seized Assets, Percentage of African- Americans, and Per Capita Income Real per capita value of seized assets Percentage African- American Real per capita income Real per capita value of seized assets 1 0.0618 -0.0233 Percentage African-American 0.0618 1 -0.0816 Real per capita income -0.0233 -0.0816 1 Percentage with bachelor’s degree 0.0034 -0.150 0.753 10 The Effects of Civil and Criminal Forfeitures Unemployment rate 0.03282 0.198 -0.0029 Percentage ages 18-24 -0.0218 -0.0494 -0.251 Population density -0.0146 0.193 0.601 Per capita alcohol consumption 0.01809 -0.221 0.324 Per capita prison population 0.00901 0.510 -0.163 Percentage in large cities -0.02837 0.0113 0.0468 Police employees per capita 0.00754 0.211 0.181 Rate of drug-related arrests 0.02931 0.1611 0.134 Regarding correlations between the variables, the strongest correlations are between percentage African-American and per capita prison population (0.51), real per capita income and percentage with bachelor’s degree (0.753), real per capita income and population density (0.601), per capita prison population and police employees per capita (0.49241), and per capita prison population and rate of drug- related arrests (0.41791). These correlations suggest that there is substantial correlation between measures of economic success and between measures of criminal activity in a given state. Interestingly, there is weak correlation between the value of seized assets and drug-related arrests. Hence, it appears that civil and criminal forfeitures may have minimal impact on drug-related crime. However, a correlation analysis may be inadequate in capturing the relationship between these two variables, especially in the context of panel data and because of the significant variability in seized asset data. Table 3: Bivariate Correlations for Percentage with Bachelor’s Degree, the Unemployment Rate, and Percentage Ages 18-24 Percentage with bachelor’s degree Unemployment rate Percentage ages 18-24 Percentage with bachelor’s degree 1 -0.00815 -0.08412 Unemployment rate -0.00815 1 -0.11997 Percentage ages 18-24 -0.08412 -0.11997 1 Population density 0.48734 0.10623 -0.2356 Per capita alcohol consumption 0.24921 -0.03833 -0.1662 Per capita prison population -0.32289 0.12033 -0.00605 Percentage in large cities 0.00626 0.2045 -0.11333 Mark Gius Justice Policy Journal, Spring, 2018 The Effects of Civil and Criminal Forfeitures 11 Police employees per capita 0.03393 -0.02764 -0.04592 Rate of drug-related arrests 0.00968 0.0401 -0.15158 Table 4: Bivariate Correlations for Population Density, Per Capita Alcohol Consumption, and Per Capita Prison Population Population density Per capita alcohol consumption Per capita prison population Population density 1 0.00511 -0.10085 Per capita alcohol consumption 0.00511 1 0.01449 Per capita prison population -0.10085 0.01449 1 Percentage in large cities -0.10027 -0.01916 0.17288 Police employees per capita 0.13948 0.06884 0.49241 Rate of drug-related arrests 0.07248 0.04657 0.41791 As noted previously, in order to determine if civil and criminal forfeitures deterred drug-related crimes, a random effects model that controls for both state- level and year-specific random effects was used. All observations were weighted using state-level population, standard errors were corrected using a clustering method (clustering is done at the state-level), and a log-linear functional form was used. Results are presented on Table 6. These results suggest that there is a negative relationship between the per capita value of seized assets and the drug- related crime rate. These results provide evidence of the deterrent effect of forfeitures. The greater the value of seized assets, the lower is the drug-related crime rate. It is important to note, however, that the effect is very minimal. According to these results, for every $1 increase in the value of per capita seized assets, the drug-related arrest rate falls by 0.092%. Given that the average value of per capita seized assets for all states and for the entire period examined is $0.58, the effect of forfeitures on drug arrests is negative, but very minimal. 12 The Effects of Civil and Criminal Forfeitures Table 5: Bivariate Correlations for Percentage in Large Cities, Police Per Capita, and Drug-Related Arrest Rate Percentage in large cities Police employees per capita Rate of drug- related arrests Percentage in large cities 1 0.13134 0.19475 Police employees per capita 0.13134 1 0.67416 Rate of drug-related arrests 0.19475 0.67416 1 Regarding the control variables, states with greater percentages of college- educated individuals, higher per capita incomes, higher unemployment rates, more 18-24 year olds, greater per capita alcohol consumption, and larger per capita prison populations have higher rates of drug-related arrests. These results are consistent with the results of prior studies in this area (Kelly and Kole, 2016; Mast, Benson, and Rasmussen, 2000). In terms of magnitude, the variables that have the greatest effects on drug-related arrests are the percentage of the population ages 18-24 and the unemployment rate. Hence, states with more unemployed young adults have higher rates of drug-related arrests. Table 6: Random Effects Regression Results Variable Coefficient Test Statistic Real per capita value of seized assets -0.0092 -2.81*** Percentage African-American -0.12 -0.17 Real per capita income 0.00011 7.97*** Percentage with bachelor’s degree 2.602 3.56*** Unemployment rate 5.21 4.80*** Percentage ages 18-24 10.17 6.16*** Population density -0.0002 -0.83 Per capita alcohol consumption 0.44 5.78*** (Continued) Variable Coefficient Test Statistic Per capita prison population 0.00097 3.82*** Percentage in large cities 1.237 5.12*** Police employees per capita 0.00032 1.46 Region of residence – northeast -0.068 -0.40 Mark Gius Justice Policy Journal, Spring, 2018 The Effects of Civil and Criminal Forfeitures 13 Region of residence – south 0.547 3.71*** Region of residence - west 0.0746 0.57 Note: p-value<=1% ***; 1%
presented in parentheses. Panel A presents the unweighted results, although we will discuss the
impact of weighting by population later (Panel B). Each column of the table reports a different
level of crime data aggregation (i.e., total, violent, non-violent, and various sub-categories of
crimes). Following the intuition of the Becker (1968) model, we expect crime to fall as the
returns to such activity are reduced with the increased risk of forfeiture.
In Column (1) we report the impact of forfeiture on total reported crime. In states where
federal forfeiture was more generous than existing state law, we find that crime fell by an
average of 2,691 crimes. Evaluated at the mean number of crimes in the sample, increasing
police agencies’ ability to retain 80-90 percent of forfeited assets led to a 17.3 percent drop in
crime. Not surprisingly, the overall decrease in reported crime is associated with changes in non-
violent or property crime. In Columns (2) and (3), we report the estimates for violent and non-
violent crime separately. This finding is consistent with the notion that criminals are more likely
to respond to financial incentives when the crime is not an act of passion. In Columns (4) and (5)
we separately report the estimates for burglary and larceny which comprise over 90 percent of all
non-violent crimes (FBI, 2010). Within non-violent crime we estimate that about 58 percent of
the decline in total crime reported is driven by decreases in larceny. About 38 percent of the
reduction in non-violent crime is due to a decrease in burglary.
In Table 1, Panel B we report the estimates of forfeiture on the offenses reported, but
weight each observation by population. We do this to highlight the difference between the
impacts of forfeiture for the average citizen versus the average county. When we weight by
population, our estimates generally increase in absolute magnitude, however the relative
magnitude is slightly smaller (10.1 percent decrease in crime). The difference in magnitudes
suggests that the impact of expanded forfeiture powers are heterogeneous across counties (Solon,
Haider, and Wooldridge, 2015). We explore this and other sources of heterogeneity in Section 5
Next, we consider police departments’ responses to the financial incentives of expanded
forfeiture opportunities. We expect that a profit-maximizing police chief would reallocate effort,
as measured by arrests, toward more profitable crimes. In Table 2 we report the reduced form
change in the number of total arrests, violent and non-violent crime arrests, as well as the change
in arrests for specific Index I crimes. Overall, we find that the total number of arrests in the
treatment states increased after 1984, however, the estimates are imprecise. If police respond to
financial incentives, we would expect the number of drug arrests to increase. While we estimate
a positive coefficient, it is imprecise over the entire post-1984 period. If police departments were
slow to respond to the financial incentives, then it is possible that the delay could push the entire
post-1984 estimate towards zero. In the next section we estimate the year-by-year effects of the
change in policy.
Finally, we consider whether local governments respond to increased forfeiture activity
by cutting budgets and by investing in additional police strength. Given the nature of Census
reporting, we are only able to measure total police expenditures, regardless of the funding
source. Thus, as long as local governments do not completely offset police budgets by the value
of seizures, we should estimate a non-decreasing relationship between expenditures and
forfeiture. In Table 3, Columns (1) and (3), we report the estimated relationship between
forfeiture and police expenditures. We estimate a positive relationship, however the magnitude is
generally small and not statistically significant at traditional levels. Broadly, our estimates
correspond to Baicker and Jacobsen (2007) who find that municipalities only partially reduce
appropriations in response to abnormally large seizures. We also report the estimated
relationship between the forfeiture and the size of the police force, measured by the number or
sworn officers. Should the police departments experience a budgetary bonanza, we might expect
that they respond by hiring additional officers. Our estimates, reported in Table 3, Columns (2)
and (4), do not support this notion, consistent with stable budgets. We do not find any evidence
of local law enforcement expanding faster in states more heavily influenced by the CCCA.
However, it is possible that our estimated decline in manpower could be the result of officers
diverted to new drug task forces, which were reported as separate agencies once they were
created. Unfortunately, data on the task forces’ manpower is ill-reported, so it is challenging to
empirically examine this conjecture. Alternatively, the reduction in officer numbers could be
driven by demand, as crime fell in the treated states relative to the controls.
4.2 Dynamic Effects
To this point we have limited our discussion of the results to the average impact in the
post-1984 period. However, our baseline specification may mask interesting dynamics. To better
understand the impacts of forfeiture over time, we augment our baseline specification by
estimating an event study model as follows:
, , = 0 + ∑ (1 − ,1984) × 1( == ) ,
+ ,1970 + , + + + , . (2)
This specification allows equitable sharing to have a different impact in each year of the sample
(relative to 1983). The event study specification also serves as a check on the validity of our
empirical strategy, as the CCCA should have no differential impact on outcomes in the years
prior to its enactment in 1984.
In Table 4 we report our year-by-year estimates for offenses reported (total crime, violent
crime, non-violent crime, burglary, and larceny). For brevity, we focus on total crime (Column
1) and non-violent crime (Column 3). First, note that the coefficients prior to the enactment of
the CCCA between 1977 and 1982. In each year we generally cannot reject that the coefficient is
zero.9 This result provides additional evidence that states with differing state forfeiture statutes
before 1984 were on similar trends just prior to the CCCA. After the change in forfeiture law, we
estimate a decrease in total and non-violent crimes through 1992, although the magnitude
declines beginning in the 1990s. The magnitudes are largest between 1986 and 1990. Why might
the reported crime rates initially dip and then return to the steady state? As noted earlier,
dynamic behavioral models of crime suggest that criminals update their information with
experience. If criminals believe that police will aggressively pursue forfeiture, then we should
estimate a large initial drop, especially after police departments have sufficient time to reallocate
their effort. If criminals learn over time that their priors overweight the probability of forfeiture,
then future decisions to commit a crime should account for these updated beliefs, perhaps
resulting in more crime later.
9 In 1977 there appear to be meaningful differences for all classes of crime except violent crime.
However, in the years immediately before the enactment of equitable sharing, 1980-1982, there are no
statistically significant differences for any crime outcome.
In Table 5 we report our time-varying estimates for the arrest outcomes. As before, we
report the estimates for total arrests, violent, non-violent, and drug crimes. Prior to the enactment
of the CCCA in 1984 we cannot reject that the estimate is zero. The lack of a statistically
significant difference in arrest outcomes in the pre-treatment period is reassuring and reinforces
the baseline empirical design that assumes states with different forfeiture policy prior to the
CCCA were on the same pre-trends. In the post-1984 period, we find little evidence that either
the total number or arrests, or arrests for violent and non-violent crimes changed. However, we
find that there is a large and statistically significant increase in the number of arrests for drug
crimes beginning in 1988. Between 1988 and 1992, the average increase in the number of drug
arrests is 732.9 per jurisdiction-year. Given that the average county had 1,951 drug arrests
annually, our estimates suggest that drug arrests increased by approximately 37.5 percent after
1988. The sizable increases in drug arrests suggest that police departments were responding to
the financial incentives that equitable sharing introduced. One open question is why the effect
took three to four years to appear. We believe that institutional frictions can help to explain the
delayed response. It took time for police departments to establish special drug task forces to take
advantage of the new federal law. Evidence of this policy lag can be seen in police surveys
themselves. The Law Enforcement Management and Administrative Statistics (LEMAS) survey
administered by the Bureau of Justice Statistics did not initially ask local law enforcement
agencies about the existence of special drug task forces. However, beginning in the 1990 version
of the survey, the agency began inquiring about such task forces and asset forfeitures. The
updated survey design provides anecdotal evidence that the widespread creation of such
specialized units was not immediate.
In Table 6 we report the year-by-year estimates using police expenditures and the number
of sworn officers as the outcomes of interest. Between 1978 and 1980, there is a difference in
expenditures, with less generous forfeiture states (i.e., the treatment states) providing lower rates
of funding. Yet in the years just prior to the enactment of the CCCA, spending across
jurisdictions in the different types of states was similar. For the post-1984 period, we find that
police expenditures are generally increasing over time, however, the estimates are imprecise. In
1984 and 1985, the coefficient is positive and statistically significant, with a magnitude
suggesting that police expenditures increased by 4.8 to 8.4 percent. While imprecise, this
evidence suggests that local governments were, at the very least, not taking away resources from
police faster than local law enforcement could seize revenue. Quickly, though, it seems that local
governments better understood the new source of revenue generated from seizures and budgets
In Table 6 Column (2) we report the year-by-year effect on sworn officer strength. Prior
to 1984, there were no significant differences between jurisdictions in different types of
forfeiture states. After the rollout of the CCCA, however, we estimate that police officer strength
declined by 32 to 97 officers, or between 4.5 and 13.9 percent relative to the mean, in states that
saw the largest incentive shock from the federal law. What might explain this relative decline in
officer strength? We partially attribute the decline to the formation of new gang and drug related
task forces. Overall, officer strength increased over time, as highlighted in Figure 3, however, the
formation of new drug task forces in existing jurisdictions could have siphoned officers from
existing units. As evidence of the siphoning effect, consider data from the 2007 Census of Law
Enforcement Gang Units. Of the 537 agencies in the Census, 89 had a gang unit as of 1992.10 Of
those 89, 73 were created between 1985 and 1992. The expansion of special units coincides with
the enactment of the CCCA and provides anecdotal evidence that our findings may be partially
explained by the changes in the composition of police units.
To recap, we find evidence that following the enactment of the federal equitable sharing
provision, reported crime, driven by non-violent property crimes, declined by approximately 17
percent, while drug arrests increased up to 37 percent, albeit after an initial delay following
enactment of the CCCA. Combined, these estimates suggest that civil asset forfeiture can be an
effective tool against certain types of crime, but may also incentivize police departments to
reallocate effort toward revenue-generating activities that may lead to its own unintended
consequences. We take up this idea in Section 5.6 below.
5. Robustness and Validity
To better understand our findings we conduct several robustness and validity exercises. We
begin by providing additional evidence that localities in each type of forfeiture state followed
similar trends prior to the enactment of the CCCA. There is also a potential concern that the time
period that we study is unique. The 1980s encompassed the crack epidemic in the United States,
so one might question whether the time fixed effects fully account for the intensity of the crack
boom across time and space? To address this concern, we estimate a set of placebo regressions
using crime data from Canada, which never adopted forfeiture statutes during our sample period,
but experienced similar trends in drug use.11
10 The 2007 Census of Law Enforcement Gang Units sampled police agencies with at least 100 sworn
officers and at least 1 full time officer allocated to gangs.
11 The first civil asset forfeiture statute adopted in Canada was enacted in Ontario in 2001.
As noted above, the CCCA was a broad piece of crime-fighting legislation. Even though we
compare changes in outcomes between locations based on pre-existing differences in state
forfeiture law, these pre-existing laws might be correlated with other factors related to attitudes
toward crime or historic episodes, which were altered by the CCCA. Thus, we may not be
estimating the impact of forfeiture, but a composite effect of many provisions contained within
the CCCA. To address these concerns we consider specifications where we control for additional
variables related to a state’s position on crime prior to the federal law and historic measures of
racial tension. Next, in our presentation of our baseline results we highlighted the potential for
heterogeneity. Thus, we explore how our results differ according to county characteristics, such
as population and racial composition. Additionally, the quality of reported crime data is a
concern. To test whether crime or reported crime is changing in response to the CCCA, we
collect various measures of property value and test whether they respond to the enactment of
equitable sharing. Finally, we provide corroborating evidence that police may reallocate effort
toward more profitable drug enforcement by examining changes in traffic fatalities.
5.1 Parallel Trends
Our baseline strategy implicitly assumes that states with both generous and stingy pre-
existing forfeiture statutes followed similar time paths prior to the enactment of the CCCA.
Previously, we noted that the year-by-year coefficients were generally zero in pre-treatment
years. To further test the validity of the parallel trends assumption, we restrict the sample to the
period prior to the enactment of the CCCA and specify the following regression:
, , = 0 + 1 , + 2 + 3 , × + ,1970 + , + + , . (3)
In this regression we define , as any county in a state that permitted local law
enforcement to retain less than 80 percent of the proceeds from a forfeiture and is a linear time
trend. The coefficient 3 tests whether jurisdictions that eventually benefitted from the
enactment of forfeiture are on a different time trend than the control counties.
Our results from this exercise are reported in Table 7. In Panel A we report the estimates
of 3 for the outcomes related to reported offenses. In Panel B we report the outcomes related to
arrests. Across all outcomes we cannot reject that both types of counties/jurisdictions followed
the same time trends prior to CCCA enactment, lending additional support to our empirical
5.2 Canadian Placebo
Our primary dataset encompasses a unique period in the United States. In terms of
macroeconomic conditions, the country exited a recession in the early 1980s, experienced
growth, and then entered another mild recession at the end of the Bush Administration. The
nation also suffered from a severe drug epidemic related to crack cocaine. Given these changes
one might question whether year fixed effects fully absorb these changes in macroeconomic
conditions. To mitigate concerns that our baseline estimates are the result of a spurious
correlation, we specify a set of placebo regressions using crime data from Canada. Canada is
uniquely suited to test the robustness of our main findings. Macro conditions throughout the
1980s and the experience with the drug epidemic were similar in Canada, yet Canada nor its
provinces had instituted a law enabling forfeiture during our sample period. Canada also collects
Uniform Crime Reports for a variety of criminal offenses from its police jurisdictions. If our
main findings are the result of the enactment of the CCCA and not the result of a spurious
correlation, we should not find any effect of the equitable sharing rule in the Canadian sample.
To operationalize this falsification experiment we randomly assign a placebo province-level
retention rate for seized property that matches the observed distribution in the United States. This
random assignment ensures that at least one province always begins with a retention rate of 100
percent, while the majority of provinces have rates set below 80 percent, which mimics the U.S.
case. We then substitute these rates into Equation (1) and estimate the equation. In total, we
specify 15 permutations of these placebo retention rates at the province level.
We present our 15 sets of estimates for total crime, violent crime, and non-violent crime
in Table 8. Each coefficient in the table is the estimate from a separate regression. Each row
refers to a specific permutation of the placebo retention rates, while each column refers to a
separate outcome using that permutation. As expected, almost all of our estimates are statistically
insignificant, and are distributed about zero. This finding lends credence that our baseline
specification is not the result of a spurious correlation, but instead captures the impact of the U.S.
5.3 Changes in Attitude Toward Crime
The CCCA was a broad piece of legislation that expanded police powers, altered rules for
establishing bail, facilitated federal sentencing reform, and possibly changed local attitudes
toward crime. If the federal government, through the enactment of the CCCA, pressured states to
become tougher on crime, then our estimates may reflect the bundle of changes associated with
the CCCA that extend beyond forfeiture. To address this concern we add additional covariates to
our main regression to capture pre-existing differences across states in their enforcement and
attitudes toward crime. First, we include a measure of the per capita incarceration rate at the state
level in 1981, drawn from the DOJ’s Sourcebook of Criminal Justice Statistics, and interact it
with year fixed effects. By controlling for the pre-existing state-level incarceration rates, we
allow states that were initially tough or soft on crime to evolve differently over time. Second, as
the War on Drugs expanded, it may have altered the distribution of cases and effort across
federal and state/local agencies across the country. To address this concern, we measure the
share of federal defendants who originated from each state prior to the enactment of the CCCA
(again, drawing data from the DOJ’s Sourcebook of Criminal Justice Statistics). Our intent is to
proxy for the pre-existing intensity of federal prosecution in a given location, as locations with
initially high levels of federal prosecution may evolve differently than those where the federal
government was initially less active. Finally, during the 1960s and early 1970s, many
communities experienced race riots. The experiences in these communities during the 1960s
could have persistent effects with regards to the interaction with police forces and future policies
that altered police power. To control for this possible race riot channel, we add a control for the
number of race riots in each county during the 1960s, drawing on data from Carter (1986) and
Collins and Margo (2007).
In Table 9, we report estimates of Equation (1) using reported crime as the outcome of
interest. In column (1) we display our baseline estimates (replicated from Table 1), while
columns (2) – (4) report the additional estimates. When we include the initial incarceration rate
per capita, the share of federal cases originating in each state, or the intensity of race riots during
the 1960s, our estimates remain qualitatively similar. Indeed, each point estimate with the
additional controls lies within the confidence interval of the baseline estimates. Given the
similarity of the estimates after the inclusion of the additional controls, it is more likely that the
estimates are the result of equitable sharing and not changes in attitudes towards crime that may
have accompanied the CCCA.
To further investigate the impact of county population and the size of minority populations,
we augment Equation (1) by interacting either population or the percent black with the main
treatment variables. In Table 10 we report our estimates for the total level of reported crime,
violent crime, non-violent crime, burglary, and larceny. Our estimates suggest the presence of
heterogeneity based on population. For burglary and larceny, which account for approximately
90 percent of all crimes, we estimate that increases in population result in fewer crimes. Based
on our estimates an increase in population from the sample mean to a county population of
500,000 would result in an additional decrease of 813 burglaries (21 percent relative to the mean)
and 308 larcenies (3.6 percent relative to the mean).
Given the potential for police departments to discriminate against minorities, we might
expect larger declines in crime in places where minority groups are concentrated, especially if
minorities view the penalties as more likely or more severe (Mustard, 2001; Donohue and Levitt,
2001b; Antonovics and Knight, 2009; Price and Wolfers, 2010; Abrams et al., 2012; Rehavi and
Starr, 2014; Horrace and Rohlin, 2016). We generally estimate negative coefficients on the
interaction term between race and treatment, suggesting larger declines in burglaries in locations
with relatively larger minority populations. These estimates might be perceived as evidence of
potential discrimination in the use of forfeiture, yet in most instances the estimates lack statistical
precision, making such strong conclusions difficult to support empirically.
5.5 Property Values
Using reported crime as a proxy for the true crime conditions in a location could raise
concern. Changes in reporting may itself be a function of expanded forfeiture activity.
Individuals who were victimized may perceive that their property is at risk if they report a crime,
or the policy may generally generate ill will towards the police, especially in light of the
potential for the abuse of civil liberties. Therefore, it is not clear that our estimates reflect
changes in reporting or real changes in crime. To address this we replace the outcome in
Equation (1) with either the decennial Census’s median home value for the decades 1970-1990,
or use the state-level quarterly price index reported by the Federal Housing Financing Agency
(FHFA) for the period 1977-1992.12 This strategy relies on the existing literature that has
documented the relationship between property values and crime (Lynch and Rasmussen, 2004;
Collins and Margo, 2007; Tracey and Rockoff, 2008; Pope and Pope, 2012, Ajzenman, Galiani,
Seira, 2015; Cunningham, 2016). If reporting responds to forfeiture, while the actual level of
crime remains fixed, property values should not respond to the enactment of the law. Conversely,
if crime (and not just reporting) declines in response increased forfeiture activity, we should
estimate a positive relationship between forfeiture and property values.
Indeed, in Table 11 we report that both the median home values measured by the Census and
the FHFA price index both increase in places more strongly treated after the enactment of the
CCCA. In Column (2) we find that home values increase by 1.67 percent for a 10 percent
increase in assets retained by local law enforcement. Using the FHFA measure we estimate
slightly larger magnitudes. Combined, these different measures of property values suggest that
the actual level of crime was decreasing, rather than just changes in reporting behavior.
5.6 Reallocation of Police Effort
12 The FHFA also constructs a price index at the MSA level, however, only 39 MSAs began reporting on
or before 1977 that fall within our 38 state sample. The results using the quarterly MSA data are
qualitatively similar to our estimates at the state level.
Police departments have relatively few mechanisms to generate revenue. One key source of
revenue over time has been traffic citations. Indeed, traffic ticket revenues often fund portions of
police salaries and citations are sensitive to local fiscal conditions (Markowsky and Stratmann,
2009; Garrett and Wagner, 2009). If forfeiture becomes a new tool to generate revenue, with a
potentially higher marginal revenue than routine traffic stops, profit maximization suggests that
police departments would reallocate effort from traffic enforcement to policing drug crime. The
economics literature has found that drivers are responsive to changes in enforcement on roads
and highways. In Oregon, for example, following a severe budget cut in 2003, state highway
patrols were cut by 35 percent, resulting in a 10-20 percent increase in auto-related injuries and
fatalities (DeAngelo and Hansen, 2014). Similarly, there is evidence that increased traffic
monitoring during campaigns such as Click-it-or-Ticket also reduces motor vehicle fatalities
(Luca, 2015). Traffic citations also respond to budget shocks, which leads to increased road
safety (Makowsky and Stratmann, 2011).
To indirectly test whether police reallocated effort away from traffic enforcement to other
revenue-generating activities, we follow DeAngelo and Hansen (2014) and estimate a Poisson
regression as follows:
( ℎ ) = exp[ 0 + 1(1 − ,1984) × 1( ≥ 1984) + + + ] (4).
The dependent variable is either the number of people in fatal accidents or the number of
vehicular fatalities in a given county-year. The fatality and accident data are compiled from the
microdata reported by the National Highway Traffic Safety Administration’s (NHTSA) Fatality
Analysis Reporting System (FARS) Accident series for the period 1977-1992. Unlike DeAngelo
and Hansen (2014), who used state-level data, we do not explicitly scale for vehicle miles
traveled because they are not reported at the county level. Instead, we control for the annual
state-level VMT that we collected from yearly volumes of the Federal Highway Administration’s
“Highway Statistics” report. Controlling for VMT, rather than scaling by VMT does not force
the coefficient to be equal to 1. We also control for county fixed effects and year fixed effects.
In Table 12 we report the estimated deadly accident and fatality elasticities associated
with the enactment of the CCCA for two samples. The first sample consists of all counties in the
38 states for which we have coded forfeiture statutes. Our second sub-sample is constructed to
match the counties in our reported crime sample. Because we impose a population cutoff of
50,000 to appear in the crime samples, the accident and fatality sub-sample consists of counties
that are more populous and urbanized relative to the full sample of counties. In Column (1), we
report the elasticity for the number of people involved in a fatal crash, while in Column (2) we
report the elasticity with respect to fatalities using the full sample. In Columns (3) and (4) we
report the analogous estimates using the restricted (more urban) sample. Focusing on the
elasticity with respect to fatalities (Column (2)), we estimate that for a 10 percent increase in
forfeited assets retained, traffic fatalities increase by 2.8 percent. In the restricted sample, the
estimate is similar, a 10 percent increase in forfeited assets retained leads to a 2.3 percent
increase in fatalities. Provided that many states in our sample experienced an 80 percent increase
in local proceeds from forfeiture, we estimate that fatalities in the treated states increased by
18.4-22.7 percent relative to untreated states. These estimates are similar in magnitude to the
estimates reported in DeAngelo and Hansen (2014).
Given the mean number of fatalities in each sample, the estimates imply that treated
counties experience anywhere from 3.2-7.75 additional traffic fatalities per year. As of 2016, the
U.S. Department of Transportation (USDOT) assesses a value of a statistical life at $9.6 million.
Applying the USDOT’s valuation to our estimates suggests that there is a hidden cost of $30.7-
$74.4 million associated with the enforcement of the CCCA.13 This finding suggests that there
are potentially large externality costs associated with the implementation of civil asset forfeiture
that have not previously been documented in the literature. While this finding is not direct
evidence that police reduced their traffic patrols in response to the enactment of equitable
sharing, it is suggestive that such a reallocation occurred and has led to a significant unintended
consequence of the law.
Civil asset forfeiture was initially proposed as a means to “cut the head off the snake” and to
limit the gains that criminal organizations obtained through their illicit activities. Lawmakers
hoped that by reducing the financial returns from crime, crime would be reduced. The financial
incentives created within the forfeiture environment, however, have the potential to alter police
incentives. Indeed, these incentives are strong enough to have raised significant civil liberties
concerns. Recently, over concerns of abuse and the potential violation of habeas corpus, the
Department of Justice suspended federal civil asset forfeiture in the absence of a criminal charge
or warrant. Several states have also begun to reform their approach to forfeiture, either by
increasing the standard of proof or requiring that a forfeiture be accompanied by a criminal
conviction. Understanding the impacts of civil asset forfeiture is especially important as
legislators debate its reform. There is currently an open question regarding the distributional
impacts of forfeiture and whether specific groups have been targeted by police. Recent forfeiture
level data from Chicago indicate that forfeitures are overwhelmingly carried out in low income
and minority communities; and while forfeiture is intended to target kingpins, the median seizure
has totaled just over $1,000 (often in cash), which has led some to challenge police practices
with regard to forfeiture.14 The recent spotlight that has been shining on forfeiture as a crime-
fighting tool makes it increasingly important to understand how the system affects the behavior
of participants in the criminal justice system and generates intended and unintended
We estimate that between 1984 and 1992 changes in forfeiture opportunities contained in
the CCCA led to a 17 percent reduction in non-violent property crimes in places where police
could suddenly keep a larger share of their seizures. The bulk of this reduction is driven by
changes in burglary and larceny. Thus, reducing the financial incentives of crime through
equitable sharing seems to have achieved one of the primary objectives of forfeiture – that is, to
reduce crime. The altered financial incentives that police departments faced as a result of the
CCCA led them to reallocate effort toward fighting drug crimes, where the opportunity to seize
cash was greatest. We estimate that between 1989 and 1992, drug arrests increased by
approximately 37 percent in the treated states. We also find suggestive evidence that police were
reallocating effort away from traffic enforcement and toward drug crime. We estimate that
traffic fatalities increased by about 22 percent after the enactment of equitable sharing. While our
analysis reveals that civil asset forfeiture may help to achieve some of the policy goals that
proponents tout, it also raises new concerns about the unintended negative consequences of the
law, beyond the civil liberties concerns that opponents emphasize.
Abrams, D.., Bertrand, M. and Mullainathan, S. “Do Judges Vary in Their Treatment of
Race?.” The Journal of Legal Studies 41, no. 2 (2012): 347-383.
Aisch, Gregor and Alicia Parplpiano. “What do you Think is the Most Important Problem Facing
this Country Today?.” New York Times. Feb. 27, 2017. Accessed August 17, 2017.
Ajzenman, Nicolas, Sebastian Galiani, and Enrique Seira. “On the distributive costs of drug-
related homicides.” The Journal of Law and Economics 58, no. 4 (2015): 779-803.
Antonovics, K., and Knight, B. “A new look at racial profiling: Evidence from the Boston Police
Department.” The Review of Economics and Statistics 91, no. 1 (2009): 163-177.
Baicker, K., and Jacobson, M.. “Finders keepers: Forfeiture laws, policing incentives, and local
budgets.” Journal of Public Economics 91.11 (2007): 2113-2136.
Becker, G. “Crime and Punishment: An Economic Approach.” Journal of Political Economy 76,
no. 2 (1968): 169-217.
Benson, B., Rasmussen, D. and Sollars, D. “Police bureaucracies, their incentives, and the war
on drugs.” Public Choice 83, no. 1 (1995): 21-45.
Boudreaux, D., and Pritchard, A. “Civil forfeiture and the war on drugs: Lessons from economics
and history.” San Diego L. Rev. 33 (1996): 79.
Carpenter, D. Knepper, L., Erickson, A., and McDonald, J. “Policing for profit: the abuse of civil
asset forfeiture.” Institute for Justice 2 (2015).
Carter, Gregg Lee. “The 1960s black riots revisited: city level explanations of their
severity.” Sociological Inquiry 56, no. 2 (1986): 210-228.
Chalfin, Aaron, and Justin McCrary. “Criminal deterrence: A review of the literature.” Journal of
Economic Literature 55, no. 1 (2017): 5-48.
Collins, William J., and Robert A. Margo. “The economic aftermath of the 1960s riots in
American cities: Evidence from property values.” The Journal of Economic History 67, no. 4
Cunningham, Jamein P. “An evaluation of the Federal Legal Services Program: Evidence from
crime rates and property values.” Journal of Urban Economics 92 (2016): 76-90.
DeAngelo, Gregory, and Benjamin Hansen. “Life and death in the fast lane: Police enforcement
and traffic fatalities.” American Economic Journal: Economic Policy 6, no. 2 (2014): 231-257.
Donohue III, John J., and Steven D. Levitt. “The impact of legalized abortion on crime.” The
Quarterly Journal of Economics 116, no. 2 (2001): 379-420.
Donohue, John J., and Steven D. Levitt. “The impact of race on policing and arrests.” The
Journal of Law and Economics 44, no. 2 (2001): 367-394.
Garrett, Thomas A., and Gary A. Wagner. “Red ink in the rearview mirror: Local fiscal
conditions and the issuance of traffic tickets.” The Journal of Law and Economics 52, no. 1
Handley, Joel, and Freddy Martinez. “Inside the Chicago Police Department’s secret budget.”
Chicago Reader. September 29, 2016. Accessed August 14, 2017.
Haines, Michael R. “Historical, Demographic, Economic, and Social Data: The United States,
1790-2002 [Computer file]. ICPSR02896-v3. Inter-university Consortium for Political and
Social Research.” Ann Arbor, MI (2010).
Holcomb, Jefferson E., Tomislav V. Kovandzic, and Marian R. Williams. “Civil asset forfeiture,
equitable sharing, and policing for profit in the United States.” Journal of Criminal Justice 39,
no. 3 (2011): 273-285.
Horrace, William C., and Shawn M. Rohlin. “How dark is dark? Bright lights, big city, racial
profiling.” Review of Economics and Statistics 98, no. 2 (2016): 226-232.
Kelly, Brian D., and Maureen Kole. “The effects of asset forfeiture on policing: a panel
approach.” Economic Inquiry 54, no. 1 (2016): 558-575.
Levitt, Steven D. “The effect of prison population size on crime rates: Evidence from prison
overcrowding litigation.” The quarterly journal of economics 111, no. 2 (1996): 319-351.
Levitt, Steven D. “The relationship between crime reporting and police: Implications for the use
of Uniform Crime Reports.” Journal of Quantitative Criminology 14, no. 1 (1998): 61-81.
Levitt, Steven D. “The limited role of changing age structure in explaining aggregate crime
rates.” Criminology 37, no. 3 (1999): 581-598.
Levitt, Steven D. “Understanding why crime fell in the 1990s: Four factors that explain the
decline and six that do not.” The Journal of Economic Perspectives 18, no. 1 (2004): 163-190.
Luca, Dara Lee. “Do traffic tickets reduce motor vehicle accidents? Evidence from a natural
experiment.” Journal of Policy Analysis and Management 34, no. 1 (2015): 85-106.
Lynch, Allen K., and David W. Rasmussen. “Proximity, neighborhood and the efficacy of
exclusion.” Urban Studies 41, no. 2 (2004): 285-298.
Makowsky, Michael D., and Thomas Stratmann. “Political economy at any speed: what
determines traffic citations?.” The American Economic Review 99, no. 1 (2009): 509-527.
Makowsky, Michael D., and Thomas Stratmann. “More tickets, fewer accidents: How cash-
strapped towns make for safer roads.” The Journal of Law and Economics 54, no. 4 (2011): 863-
Mas, Alexandre. “Pay, reference points, and police performance.” The Quarterly Journal of
Economics 121, no. 3 (2006): 783-821.
Mast, Brent D., Bruce L. Benson, and David W. Rasmussen. “Entrepreneurial police and drug
enforcement policy.” Public Choice 104, no. 3-4 (2000): 285-308.
Mumola, Christopher J., and Jennifer C. Karberg. Drug use and dependence, state and federal
prisoners, 2004. Washington, DC: US Department of Justice, Office of Justice Programs, Bureau
of Justice Statistics, 2006.
O’Harrow, Robert Jr., Sari Horwitz, and Steven Rich. “Attorney General Holder limits civil
seizure process that split billions of dollars with local and state police.” The Washington Post.
January 16, 2015. Accessed August 14, 2017.
Pope, Devin G., and Jaren C. Pope. “Crime and property values: Evidence from the 1990s crime
drop.” Regional Science and Urban Economics 42, no. 1 (2012): 177-188.
Price, Joseph, and Justin Wolfers. “Racial Discrimination Among NBA.” The Quarterly Journal
of Economics (2010): 1859-1887.
Rehavi, M. Marit, and Sonja B. Starr. “Racial disparity in federal criminal sentences.” Journal of
Political Economy 122, no. 6 (2014): 1320-1354.
Schwarcz, Steven L., and Alan E. Rothman. “Civil Foreiture: A Higher Form of Commerical
Law.” Fordham L. Rev. 62 (1993): 287.
Shi, Lan. “The limit of oversight in policing: Evidence from the 2001 Cincinnati riot.” Journal of
Public Economics 93, no. 1 (2009): 99-113.
Snead, Jason. “An Overview of Recent State-Level Forfeiture Reforms.” The Heritage
Foundation. August 23, 2016. Accessed August 14, 2017. http://www.heritage.org/crime-and-
Solon, Gary, Steven J. Haider, and Jeffrey M. Wooldridge. “What are we weighting
for?.” Journal of Human resources 50, no. 2 (2015): 301-316.
Stellwagen, Lindsey D., and Kimberly A. Wylie. Strategies for supplementing the police budget.
US Department of Justice, National Institute of Justice, Office of Development, Testing and
Tracey, D. and Rockoff, J.E., 2008. Estimates of the impact of crime risk on property values
from Megan’s laws. The American Economic Review, 98(3), pp.1103-1127.
Trott, Stephen S. “Implementing criminal justice reform.” Public Administration Review (1985):
United States Comptroller General, and United States of America. “Asset Forfeiture: A Seldom
Used Tool in Combatting Drug Trafficking.” (1981).
United States. Congress. House. Committee on the Judiciary. Civil Asset Forfeiture Reform Act:
report together with dissenting views (to accompany H.R. 1658) (including cost estimate of the
Congressional Budget Office). United States Government Printing Office. Washington, D.C.
United States. Congress. Senate. Committee on the Judiciary. Subcommittee on Criminal
Justice. Forfeiture of narcotics proceeds: hearings before the Subcommittee on Criminal Justice
of the Committee on the Judiciary, United States Senate, Ninety-sixth Congress, second session,
on forfeiture of profits of narcotics traffickers, July 23 and 24, 1980. United States Government
Printing Office. Washington, D.C. 1981.
United States Department of Justice, Federal Bureau of Investigation. Uniform Crime Reporting
Program Data [United States]: County-Level Detailed Arrest and Offense Data, 1977-1993. Ann
Arbor, MI: Inter-university Consortium for Political and Social Research [producer and
distributor], 1977-1992. c
United States Department of Justice. Federal Bureau of Investigation. Uniform Crime Reporting
Program Data [United States]: Arrests by Age, Sex, and Race, 1977-1992. ICPSR23344-v1. Ann
Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 2009-03-
United States Department of Commerce, Bureau of the Census. ANNUAL SURVEY OF
GOVERNMENTS: FINANCE STATISTICS. Washington, DC: U.S. Dept. of Commerce,
Bureau of the Census [producer], 1977-1992. Ann Arbor, MI: Inter-university Consortium for
Political and Social Research [distributor], 1984. http://doi.org/10.3886/ICPSR08146.v1
Worrall, John L. “Addicted to the drug war: The role of civil asset forfeiture as a budgetary
necessity in contemporary law enforcement.” Journal of Criminal Justice 29, no. 3 (2001): 171-
Worrall, John L., and Tomislav V. Kovandzic. “Is policing for profit? Answers from asset
forfeiture.” Criminology & Public Policy 7, no. 2 (2008): 219-244.
Table 1: Offenses Reported – Index I Crimes
Panel A: Offenses Reported – Unweighted
(1) (2) (3) (4) (5)
VARIABLES Total Violent
Violent Burglary Larceny
(1-pre84keep)*(t≥1984) -2,691*** 87.06 -2,779*** -1,053*** -1,631***
(757.4) (188.2) (727.9) (315.8) (411.5)
Observations 9,456 9,456 9,456 9,456 9,456
R-squared 0.983 0.951 0.984 0.970 0.982
Mean 15493 1710 13783 3804 8429
Clusters 591 591 591 591 591
Standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1 Panel B: Offenses Reported - Population Weighted (1) (2) (3) (4) (5) VARIABLES Total Violent Non- Violent Burglary Larceny (1-pre84keep)*(t≥1984) -8,002 2,239 -10,241** -3,809** -5,750*** (5,891) (2,531) (4,297) (1,717) (2,210) Observations 9,456 9,456 9,456 9,456 9,456 R-squared 0.989 0.965 0.990 0.984 0.988 Mean 79221 11658 67563 18812 37721 Clusters 591 591 591 591 591 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Notes and Sources: The data for this table are constructed from the FBI UCR Offenses Reported Files, 1977-1992. To appear in the sample a county must appear in the UCR for 15/16 sample years, have a population over 50,000, and be located in one of the 38 states for which we have unambiguous pre-existing state-level forfeiture statutes. Each coefficient is from a different regression and reports the relationship corresponding to 1in equation (1). Standard errors are clustered at the county level and are reported in parenthesis. Each regression includes county fixed effects, year fixed effects, and controls for the county-level 1970 level of education, unemployment rate, percent male, percent black, percent urban, and the proportion of the population aged 18-25, all interacted with year fixed effects. We also control for the annual population. In Panel A, all estimates are unweighted, while in Panel B the estimates are weighted by population. -42- Table 2: Arrests – Index I Crimes Panel A: Arrests- Unweighted Estimates (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) VARIABLES Total Violent Non- Violent Drug Robbery Agg Assault Oth Assault Burglary Larceny Vehicular (1-pre84keep)*( t≥1984) 598.2 -86.38 564.6 120.0 82.27 140.2** -255.4** 241.1 213.3 110.2 (705.8) (188.6) (655.7) (296.5) (121.5) (70.33) (112.2) (230.8) (376.3) (77.04) Observations 2,431 2,431 2,431 2,431 2,431 2,431 2,371 2,431 2,431 2,431 R-squared 0.956 0.949 0.936 0.928 0.964 0.948 0.920 0.913 0.917 0.969 Mean 8170 2418 3801 1951 488.7 686.2 1275 865.5 2561 375.4 Clusters 187 187 187 187 187 187 187 187 187 187 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Panel B: Population Weighted Estimates (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) VARIABLES Total Violent Non- Violent Drug Robbery Agg Assault Oth Assault Burglary Larceny Vehicular (1-pre84keep)*( t≥1984) 2,446 482.1 235.4 1,728 -292.2 1,035** 261.4 132.8 -545.9 648.6** (2,677) (422.8) (2,282) (1,377) (498.2) (453.3) (450.3) (690.7) (1,562) (316.8) Observations 2,431 2,431 2,431 2,431 2,431 2,431 2,371 2,431 2,431 2,431 R-squared 0.987 0.984 0.978 0.983 0.994 0.986 0.948 0.978 0.966 0.991 Mean 66824 21329 24938 20557 6765 6620 8106 5858 15171 3909 Clusters 187 187 187 187 187 187 187 187 187 187 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 -43- Notes and Sources: The data for this table are constructed from the FBI UCR Arrests Files, 1980-1992. To appear in the sample a police jurisdiction must appear in the UCR for 12/13 sample years, have a population over 50,000, and be located in one of the 38 states for which we have unambiguous pre-existing state-level forfeiture statutes. Each coefficient is from a different regression and reports the relationship corresponding to 1in equation (1). Standard errors are clustered at the police jurisdiction level and are reported in parenthesis. Each regression includes police jurisdiction fixed effects, year fixed effects, and controls for the county-level 1970 level of education, unemployment rate, percent male, percent black, percent urban, and the proportion of the population aged 18- 25, all interacted with year fixed effects. We also control for the annual population. In Panel A, all estimates are unweighted, while in Panel B the estimates are weighted by population. -44- Table 3: Police Expenditures and Sworn Officers Police Expenditures and Sworn Officers Unweighted Weighted (1) (2) (3) (4) VARIABLES Police Expenditures ($100,000) Sworn Officers Police Expenditures ($100,000) Sworn Officers (1-pre84keep)*( t ≥1984) 43.28 -82.50* 225.6 -366.2** (29.43) (47.26) (189.6) (162.0) Observations 9,040 9,376 9,040 9,376 R-squared 0.936 0.982 0.974 0.994 Mean 196.2 698.5 1238 3491 Clusters 565 586 565 586 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Notes and Sources: Sworn Officer data is drawn from the FBI UCR LEOKA 1977-1992, while Police Expenditure data is from the Census of Government Finances, 1977-1992. To appear in the sample a county must appear in the UCR for 15/16 sample years, have a population over 50,000, and be located in one of the 38 states for which we have unambiguous pre-existing state- level forfeiture statutes. Each coefficient is from a different regression and reports the relationship corresponding to 1in equation (1). Standard errors are clustered at the county level and are reported in parenthesis. Each regression includes county fixed effects, year fixed effects, and controls for the county-level 1970 level of education, unemployment rate, percent male, percent black, percent urban, and the proportion of the population aged 18-25, all interacted with year fixed effects. We also control for the annual population. Columns (1) and (2) report the unweighted estimates, while Columns (3) and (4) report population weighted estimates. -45- Table 4: Offenses Reported: Dynamic Effect (Unweighted) Offenses Reported- Dynamic Effects (1) (2) (3) (4) (5) VARIABLES Total Violent Non-Violent Burglary Larceny (1-pre84keep)*(t=1977) 1,190** -113.6 1,303*** 708.8*** 512.6** (502.9) (172.7) (432.3) (218.2) (254.0) (1-pre84keep)*(t=1978) 893.7* -78.13 971.9** 675.0*** 210.0 (499.1) (149.1) (418.2) (202.5) (230.7) (1-pre84keep)*(t=1979) 727.6 -46.21 773.8** 436.8** 261.4 (446.9) (118.1) (380.9) (169.6) (207.1) (1-pre84keep)*(t=1980) 235.2 -169.4 404.6 304.6 159.0 (502.4) (143.3) (414.9) (218.0) (199.6) (1-pre84keep)*(t=1981) -4.328 -28.26 23.93 -27.37 46.35 (425.6) (85.59) (377.4) (194.1) (253.2) (1-pre84keep)*(t=1982) -17.78 -34.32 16.55 -75.74 91.12 (372.1) (67.28) (328.4) (153.6) (178.6) (1-pre84keep)*(t=1983) 0 0 0 0 0 (1-pre84keep)*(t=1984) -1,500** -110.1 -1,390** -549.1** -633.7* (759.2) (120.3) (647.2) (227.5) (353.1) (1-pre84keep)*(t=1985) -1,558*** -9.316 -1,549*** -546.1*** -879.1*** (432.6) (74.61) (395.6) (157.9) (234.3) (1-pre84keep)*(t=1986) -2,424*** -14.50 -2,409*** -942.2*** -1,343*** (673.7) (118.8) (622.9) (241.4) (330.5) (1-pre84keep)*(t=1987) -3,056*** 8.720 -3,065*** -1,137*** -1,809*** (886.5) (119.1) (846.0) (312.1) (460.4) (1-pre84keep)*(t=1988) -2,893*** 3.679 -2,897*** -1,078*** -1,760*** (1,001) (144.8) (931.1) (342.3) (500.4) (1-pre84keep)*(t=1989) -2,971*** 47.16 -3,018*** -1,058*** -1,858*** (1,075) (192.1) (983.7) (351.8) (522.0) (1-pre84keep)*(t=1990) -2,690*** -22.77 -2,667*** -836.8** -1,761*** (970.1) (243.5) (871.4) (336.2) (483.6) (1-pre84keep)*(t=1991) -1,930** 83.73 -2,014** -517.6* -1,542*** (907.5) (281.4) (791.7) (312.9) (467.1) (1-pre84keep)*(t=1992) -1,355* 194.1 -1,549** -234.3 -1,471*** (743.9) (251.5) (708.4) (339.2) (413.7) Observations 9,456 9,456 9,456 9,456 9,456 R-squared 0.983 0.951 0.984 0.970 0.982 Mean 15493 1710 13783 3804 8429 Clusters 591 591 591 591 591 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 -46- Notes and Sources: The data for this table are constructed from the FBI UCR Offenses Reported Files, 1977-1992. To appear in the sample a county must appear in the UCR for 15/16 sample years, have a population over 50,000, and be located in one of the 38 states for which we have unambiguous pre-existing state-level forfeiture statutes. Each coefficient corresponds to in equation (2) for the given outcome. Standard errors are clustered at the county level and are reported in parenthesis. Each regression includes county fixed effects, year fixed effects, and controls for the county-level 1970 level of education, unemployment rate, percent male, percent black, percent urban, and the proportion of the population aged 18-25, all interacted with year fixed effects. We also control for the annual population. -47- Table 5: Arrests: Dynamic Effects Arrests - Dynamic Effects (1) (2) (3) (4) VARIABLES Total Violent Non- Violent Drug (1-pre84keep)*(t=1980) 1,174 43.91 433.5 696.7 (1,733) (635.4) (749.1) (607.8) (1-pre84keep)*(t=1981) 578.6 188.3 62.60 327.8 (744.8) (399.4) (248.3) (318.3) (1-pre84keep)*(t=1982) -375.5 -254.9 -310.3 189.7 (559.7) (302.0) (246.7) (190.6) (1-pre84keep)*(t=1983) 0 0 0 0 (1-pre84keep)*(t=1984) 133.6 -30.42 249.3 -85.26 (415.1) (127.6) (197.8) (254.2) (1-pre84keep)*(t=1985) -16.53 -108.9 48.63 43.77 (662.1) (223.3) (373.9) (425.7) (1-pre84keep)*(t=1986) -671.0 -231.9 -241.6 -197.5 (959.3) (331.0) (405.3) (548.6) (1-pre84keep)*(t=1987) 933.2 -38.85 547.4 424.7 (1,707) (533.7) (986.0) (266.3) (1-pre84keep)*(t=1988) 1,159 64.76 712.9 381.4** (1,496) (493.6) (952.4) (192.9) (1-pre84keep)*(t=1989) 1,669 105.0 944.2 620.2*** (1,580) (506.9) (1,008) (230.8) (1-pre84keep)*(t=1990) 1,507 -359.2 1,070 795.9*** (2,087) (853.3) (1,150) (290.6) (1-pre84keep)*(t=1991) 1,992 17.78 1,065 909.8** (1,885) (511.9) (1,156) (361.1) (1-pre84keep)*(t=1992) 1,875 -238.5 1,156 957.2** (1,934) (529.6) (1,185) (372.3) Observations 2,431 2,431 2,431 2,431 R-squared 0.956 0.949 0.936 0.928 Mean 8170 2418 3801 1951 Clusters 187 187 187 187 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Notes and Sources: The data for this table are constructed from the FBI UCR Arrests Files, 1980-1992. To appear in the sample a police jurisdiction must appear in the UCR for 12/13 sample years, have a population over 50,000, and be located in one of the 38 states for which we have unambiguous pre-existing state-level forfeiture statutes. Each coefficient corresponds to in equation (2) for the given outcome. Standard errors are clustered at the police jurisdiction level and are reported in parenthesis. Each regression includes police jurisdiction fixed effects, -48- year fixed effects, and controls for the county-level 1970 level of education, unemployment rate, percent male, percent black, percent urban, and the proportion of the population aged 18-25, all interacted with year fixed effects. We also control for the annual population. -49- Table 6: Police Spending and Sworn Officers Police Expenditures and Sworn Officers – Dynamic Effects (1) (2) VARIABLES Police Expenditures ($100,000) Sworn Officers (1-pre84keep)*(t=1977) -39.07 49.06 (25.60) (48.13) (1-pre84keep)*(t=1978) -38.05* 26.04 (20.16) (47.69) (1-pre84keep)*(t=1979) -31.62* 42.76 (16.64) (44.37) (1-pre84keep)*(t=1980) -25.06* 12.81 (13.26) (52.51) (1-pre84keep)*(t=1981) -11.42 20.05 (8.074) (51.49) (1-pre84keep)*(t=1982) -2.150 1.556 (4.065) (51.81) (1-pre84keep)*(t=1983) 0 0 (1-pre84keep)*(t=1984) 9.530** 3.793 (4.346) (15.47) (1-pre84keep)*(t=1985) 16.63* -32.75* (9.324) (18.94) (1-pre84keep)*(t=1986) 20.09 -48.16** (12.30) (22.90) (1-pre84keep)*(t=1987) 21.90 -57.94** (17.98) (26.46) (1-pre84keep)*(t=1988) 13.01 -79.21*** (18.16) (23.69) (1-pre84keep)*(t=1989) 20.78 -74.16** (21.22) (29.44) (1-pre84keep)*(t=1990) 22.43 -75.31** (26.71) (29.82) (1-pre84keep)*(t=1991) 29.25 -87.23*** (32.63) (30.14) (1-pre84keep)*(t=1992) 46.53 -97.12*** (38.18) (33.17) Observations 9,040 9,376 R-squared 0.936 0.982 Mean 196.2 698.5 Clusters 565 586 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 -50- Notes and Sources: Sworn Officer data is drawn from the FBI UCR LEOKA 1977-1992, while Police Expenditure data is from the Census of Government Finances, 1977-1992. To appear in the sample a county must appear in the UCR for 15/16 sample years, have a population over 50,000, and be located in one of the 38 states for which we have unambiguous pre-existing state- level forfeiture statutes. Each coefficient corresponds to in equation (2) for the given outcome. Standard errors are clustered at the county level and are reported in parenthesis. Each regression includes county fixed effects, year fixed effects, and controls for the county-level 1970 level of education, unemployment rate, percent male, percent black, percent urban, and the proportion of the population aged 18-25, all interacted with year fixed effects. We also control for the annual population -51- Table 7: Pre-Trends Panel A: Offenses Reported - Pre-Trends (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) VARIABLES Total Violent Non- Violent Burglary Larceny Vehicular Murder Rape Robbery Aslt (Treat*TimeTrend) 93.25 14.01 79.24 6.176 78.37 -5.305 -0.348 -0.113 17.77 -3.646 (194.8) (36.08) (160.5) (48.85) (85.62) (41.16) (0.605) (1.524) (22.77) (13.47) Observations 4,137 4,137 4,137 4,137 4,137 4,137 4,137 4,137 4,137 4,137 R-squared 0.536 0.389 0.555 0.518 0.598 0.393 0.385 0.441 0.350 0.403 Mean 14607 1430 13177 4000 7891 1285 23.82 92.05 615 723.3 Clusters 591 591 591 591 591 591 591 591 591 591 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Panel B: Arrests- Pre-Trends (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) VARIABLES Total Violent Non- Violent Drug Robbery Agg Assault Oth Assault Burglary Larceny Vehicular (Treat*TimeTrend) -124.4 -167.0 -104.3 147.0 -21.42 51.25 -192.9 -28.39 -59.29 -16.65 (649.3) (240.8) (272.0) (239.7) (24.68) (39.30) (211.5) (66.85) (224.5) (20.32) Observations 748 748 748 748 748 748 722 748 748 748 R-squared 0.708 0.691 0.694 0.703 0.726 0.708 0.412 0.696 0.641 0.667 Mean 8170 2418 3801 1951 488.7 686.2 1275 865.5 2561 375.4 Clusters 187 187 187 187 187 187 187 187 187 187 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Notes and Sources: The data for Panel A are constructed from the FBI UCR Offenses Reported Files, 1977-1983. The data for Panel B are constructed from the FBI UCR Arrests Files, 1980-1983. To appear in the sample a police jurisdiction must appear in the UCR -52- for 12/13 sample years, have a population over 50,000, and be located in one of the 38 states for which we have unambiguous pre- existing state-level forfeiture statutes. Each coefficient is from a different regression and reports the relationship corresponding to 3in equation (3). Standard errors are clustered at the police jurisdiction level and are reported in parenthesis. Each regression includes either county of police jurisdiction fixed effects, year fixed effects, and controls for the county-level 1970 level of education, unemployment rate, percent male, percent black, percent urban, and the proportion of the population aged 18-25, all interacted with year fixed effects. We also control for the annual population. -53- Table 8: Canadian Placebo Regressions Placebo Regressions - Canada (1) (2) (3) VARIABLES Total Violent Non-Violent (1-pre84keep)*( t ≥1984) -82.72 -28.82 17.45 (182.1) (29.16) (64.54) (1-pre84keep)*( t ≥1984) -227.9 -7.806 -128.9* (187.4) (22.21) (74.22) (1-pre84keep)*( t ≥1984) -338.1 -51.91** -154.9* (208.8) (21.53) (93.52) (1-pre84keep)*( t ≥1984) -69.34 -45.00 41.64 (204.2) (34.66) (69.17) (1-pre84keep)*( t ≥1984) -37.08 21.24 -24.76 (215.6) (20.83) (92.53) (1-pre84keep)*( t ≥1984) -34.94 9.464 -5.222 (228.4) (28.31) (81.22) (1-pre84keep)*( t ≥1984) 30.14 14.17 -55.99 (170.0) (18.81) (74.98) (1-pre84keep)*( t ≥1984) 134.4 20.02 38.66 (193.5) (27.82) (78.64) (1-pre84keep)*( t ≥1984) 128.1 23.38 1.216 (147.6) (20.73) (59.50) (1-pre84keep)*( t ≥1984) 70.32 -7.661 65.81 (172.9) (24.51) (67.28) (1-pre84keep)*( t ≥1984) -154.5 -56.31 10.54 (229.2) (39.49) (76.72) (1-pre84keep)*( t ≥1984) 384.2*** 40.95*** 159.2*** (90.24) (12.82) (34.15) (1-pre84keep)*( t ≥1984) 53.08 22.64 8.960 (193.2) (20.59) (81.39) (1-pre84keep)*( t ≥1984) 51.25 37.37* 4.143 (163.3) (20.63) (66.94) (1-pre84keep)*( t ≥1984) -300.2 -58.02** -102.4 (188.6) (28.19) (71.40) N 22,967 22,967 22,967 Clusters 1351 1351 1351 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Notes and Sources: Data are drawn from the Canadian UCR 1.0, All Police Services 1977-2003. The sample is constructed by first restricting the data to the period 1977-1993. We then require that each location report non-zero crime for all 17 years in the period. Each coefficient is from a separate regression and reports the relationship corresponding to 1in equation (1), following the procedure outlined in Section 5.2. Standard errors are clustered at the location level and are -54- reported in parenthesis. Specifically, we randomly assign pre-1984 forfeiture retention rates to Canadian provinces based on the pre-1984 distribution in the United States. We report the coefficients from 15 different permutations of this exercise. -55- Table 9: Offenses Reported – Additional Controls Offenses Reported - Additional Controls OUTCOME (1) (2) (3) (4) Total Crime -2,691*** -2,820*** -2,386*** -2,760*** (757.4) (783.7) (726.9) (740.1) Violent 87.06 49.36 -252.2* 50.94 (188.2) (176.0) (140.4) (175.2) Non-Violent -2,779*** -2,869*** -2,134*** -2,811*** (727.9) (735.4) (676.2) (722.3) Burglary -1,053*** -1,121*** -469.8* -1,058*** (315.8) (320.9) (260.1) (317.1) Larceny -1,631*** -1,648*** -1,324*** -1,646*** (411.5) (407.0) (396.6) (409.5) Federal Defendant Control N Y N N State Incarceration Controls N N Y N 1960s Race Riot Controls N N N Y Observations 9,449 9,449 9,449 9,449 Clusters 591 591 591 591 Standard errors in parenthesis *** p<0.01, ** p<0.05, * p<0.1 Notes and Sources: See notes to Table 1. Column (1) reports the main results from Table 1, Panel A. Column (2) controls for the number of inmates at the state level, measured in 1981. Column (3) controls for the share of federal defendants originating in the state. Both the inmate and federal defendant data are drawn from the DOJ’s Sourcebook of Criminal Justice Statistics (1981). Column (4) controls for the number of race riots that took place in the county during the 1960’s drawing on data reported in Carter (1986) and Collins and Margo (2007). The three sets of additional control variables are interacted with the full set of year fixed effects. -56- Table 10: Offenses Reported – Heterogeneous Effects Heterogeneous Effects by Population and Percent Black Population Percent Black (1) (2) (3) (4) (5) (1) (2) (3) (4) (5) VARIABLES Total Violent Non- Violent Burglary Larceny Total Violent Non- Violent Burglary Larceny (1-pre84keep)*(year>1984) -2,288* -948.8** -1,339 250.5 -1,103** -2,297*** -92.56 -2,204*** -596.5** -1,444***
(1,366) (391.4) (994.9) (474.6) (457.8) (690.1) (142.3) (660.7) (263.0) (412.3)
(1-pre84keep)* Post84* Population -0.00101 0.00342** -0.00443 -0.00419** -0.00158
(0.00510) (0.00138) (0.00373) (0.00176) (0.00163)
(1-pre84keep)* Post84*Percent Black -2,738 1,745 -4,483 -3,958* -1,281
(5,288) (1,774) (4,733) (2,303) (2,866)
Observations 9,449 9,449 9,449 9,449 9,449 9,449 9,449 9,449 9,449 9,449
R-squared 0.984 0.960 0.985 0.977 0.983 0.984 0.951 0.984 0.970 0.983
Mean 15493 1710 13783 3804 8429 15493 1710 13783 3804 8429
Clusters 591 591 591 591 591 591 591 591 591 591
Robust standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1 Notes and Sources: See notes to Table 1. -57- Table 11: Housing Values Census County Level Data FHFA State Level (1) (2) (3) (4) VARIABLES Median House Value Log(Median House Value) HPI Log(HPI) (1-Pre84Keep) *(year>1984) 9,289*** 0.0167** 13.72 0.0899*
(817.9) (0.00794) (9.110) (0.0521)
Observations 7,995 7,993 2,583 2,583
R-squared 0.877 0.981 0.902 0.930
Mean 33547 10.15 128.6 4.807
Clusters 2666 2665 38 38
Standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1 Notes and Sources: Each coefficient corresponds to 1 in equation (1) for the given housing value outcome. The results reported in Columns (1) and (2) use county-level data drawn from the Census of Population in 1970, 1980, and 1990. Standard errors are reported in parenthesis and are clustered at the county level. Columns (3) and (4) utilize quarterly, state-level price indices, from 1977 to 1993 from the Federal Housing Financial Agency (https://www.fhfa.gov/DataTools/Downloads/Pages/House-Price-Index-Datasets.aspx#mpo). Standard errors are reported in parenthesis and are clustered at the state level. As in other specifications, we restrict the sample to the 38 states that report unambiguous forfeiture statutes. Each regression includes location fixed effects and year fixed effects. https://www.fhfa.gov/DataTools/Downloads/Pages/House-Price-Index-Datasets.aspx#mpo -58- Table 12: Fatal Crashes (1) (2) (3) (4) VARIABLES Persons in Fatal Crashes Fatalities in Crashes Persons in Fatal Crashes Fatalities in Crashes All Counties in 38 State Sample Counties in Offenses Reported Sample (1-Pre84Keep) * Post84 0.0125*** 0.0284*** 0.00842* 0.0229*** (0.00381) (0.00583) (0.00469) (0.00726) Observations 45,254 45,254 10,030 10,030 Mean 33.93 14.46 100.9 42.17 Clusters 2,662 2,662 590 590 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Notes and Sources: Each coefficient corresponds to 1 in the Poisson specification, as outlined in equation (4). The county-level count data used in this sample are constructed by aggregating the National Highway Traffic Safety Administration’s (NHTSA) Fatal Accident Reporting System (FARS) (ftp://ftp.nhtsa.dot.gov/fars/) microdata to the county-year over the period 1977 to 1993. Columns (1) and (3) report the estimates using the number of persons involved in fatal accidents as the outcome of interest, while Columns (2) and (4) use the actual number of fatalities as the outcome. In columns (1) and (2), the sample consists of all counties in the 38 states that report unambiguous forfeiture statutes. In columns (3) and (4) we restrict the sample to match the counties in the FBI UCR Offenses Reported sample (see Table 1 notes for details). The regression controls for county fixed effects, year fixed effects, and state vehicle miles travelled (VMT). State-level highway VMT data are collected from the NHTSA’s annual volume, Highway Statistics, Table VM-2, 1977-1993. Standard errors are reported in parenthesis and are clustered at the county level. ftp://ftp.nhtsa.dot.gov/fars/ -59- Appendix Tables Table A1: State Level Civil Forfeiture Sources Pre-CCCA State Civil Forfeiture Statutes Sources State Link Alabama http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssal0075&size=2&collection=ssl&id=205 Alaska http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssak0051&size=2&collection=ssl&id=433 Arizona http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssaz0050&size=2&collection=ssl&id=444 Arkansas http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssar0089&size=2&collection=ssl&id=956 California http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssca0108&size=2&collection=ssl&id=461 Colorado http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssco0098&size=2&collection=ssl&id=534 Connecticut http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssct0123&size=2&collection=ssl&id=332 Delaware http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssde0025&size=2&collection=ssl&id=621 Florida http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssfl0117&size=2&collection=ssl&id=35 Georgia http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssga0090&size=2&collection=ssl&id=502 Hawaii http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/sshi0027&size=2&collection=ssl&id=377 Idaho http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssid0052&size=2&collection=ssl&id=693 Illinois http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssil0107&size=2&collection=ssl&id=1306 Indiana http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssin0119&size=2&collection=ssl&id=631 Iowa http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssia0029&size=2&collection=ssl&id=273 Kansas http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssks0042&size=2&collection=ssl&id=1189 Kentucky http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssky0115&size=2&collection=ssl&id=160 Louisiana http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssla0103&size=2&collection=ssl&id=104 Maine http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssme0068&size=2&collection=ssl&id=946 Maryland http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssmd0197&size=2&collection=ssl&id=2460 Massachusetts http://heinonline.org/HOL/Page?handle=hein.ssl/ssma0060&start_page=1&collection=ssl&id=705 Michigan http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssmi0039&size=2&collection=ssl&id=778 Minnesota http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssmn0072&size=2&collection=ssl&id=359 Mississippi http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssms0108&size=2&collection=ssl&id=920 -60- Montana http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssmt0046&size=2&collection=ssl&id=425 Nebraska http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssne0048&size=2&collection=ssl&id=1160 Nevada http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssnv0044&size=2&collection=ssl&id=491 New Hampshire http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssnh0026&size=2&collection=ssl&id=161 New Jersey http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssnj0044&size=2&collection=ssl&id=47 New Mexico http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssnm0062&size=2&collection=ssl&id=616 New York http://heinonline.org/HOL/Page?handle=hein.ssl/ssny0092&start_page=2743&collection=ssl&id=360 North Carolina http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssnc0039&size=2&collection=ssl&id=493 North Dakota http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssnd0034&size=2&collection=ssl&id=672 Ohio http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssoh0100&size=2&collection=ssl&id=1639 Oklahoma http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssok0048&size=2&collection=ssl&id=411 Oregon http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssor0038&size=2&collection=ssl&id=390 Pennsylvania http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/sspa0057&size=2&collection=ssl&id=137 Rhode Island http://heinonline.org/HOL/Page?handle=hein.ssl/ssri0151&size=2&collection=ssl&id=795 South Carolina http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/sssc0053&size=2&collection=ssl&id=585 South Dakota http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/sssd0029&size=2&collection=ssl&id=458 Tennessee http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/sstn0056&size=2&collection=ssl&id=961 Texas http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/sstx0076&size=2&collection=ssl&id=2064 Utah http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssut0035&size=2&collection=ssl&id=113 Vermont http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/ssvt0054&size=2&collection=ssl&id=360 Virginia http://heinonline.org/HOL/Page?handle=hein.ssl/ssva0092&start_page=1&collection=ssl&id=325 Washington http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/sswa0059&size=2&collection=ssl&id=1388 West Virginia http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/sswv0068&size=2&collection=ssl&id=332 Wisconsin http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/sswi0029&size=2&collection=ssl&id=102 Wyoming http://heinonline.org/HOL/Page?men_tab=srchresults&handle=hein.ssl/sswy0032&size=2&collection=ssl&id=505 -61- Table A2: Summary Statistics of , Characteristics Summary Statistics of 1970 Census Characteristics (1) (2) (3) (3)-(2) Variable Whole Sample Control County Treated County Difference Population Share Aged 18-25 47.50 47.49 47.52 -0.0308 Education - Less than High School 23.19 22.94 24.45 -1.5054 Education - High School Graduate 15.68 15.99 14.09 1.9065 Education - Some College 5.10 5.03 5.41 -0.3728 Education - College Graduate 5.09 5.10 5.09 0.0053 Unemployment Rate 1.58 1.62 1.42 0.1996 Poverty Rate 0.68 0.63 0.94 -0.3105 Percent Male 49.05 49.04 49.11 -0.0694 Percent Black 8.69 7.78 13.32 -5.5350 Percent Urban 36.96 36.56 38.99 -2.4271 Bold Indicates Statistically Significant Difference at the 5 percent level. Notes and Sources: This table reports the mean for each of the 1970 Census of Population variables that are used as covariates in Equation (1). Column (1) reports the mean for all 591 counties in the sample, while Columns (2) and (3) report the means in Control and Treatment counties respectively. Column (4) reports the difference in means between the treatment and control counties, differences shown in bold typeface indicate a statistically significant difference at the 5 percent level. -62- Table A3: Additional Specifications – Offenses Reported Offenses Reported - Unweighted Outcome (1) (2) (3) (4) Total Crime -767.7 717.5*** -4,566*** -2,691*** (1,603) (268.9) (1,057) (757.4) Violent Crime 342.7 514.9*** -247.2 87.06 (296.1) (137.7) (211.9) (188.2) Non-Violent Crime -1,110 202.6 -4,319*** -2,779*** (1,337) (183.8) (903.9) (727.9) Burglary -942.7*** -661.8*** -1,360*** -1,053*** (344.9) (142.7) (229.2) (315.8) Larceny -444.7 466.1*** -2,494*** -1,631*** (756.5) (111.8) (542.8) (411.5) Observations 9,456 9,456 9,456 9,456 County FE N Y Y Y Year FE N N Y Y X Controls N N N Y Clusters 591 591 591 591 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Notes and Sources: For data, see notes to Table 1. Each coefficient is from a separate regression, reporting the estimate of 1 from Equation (1). Standard errors are reported in parenthesis and are clustered at the county. Column (1) reports the estimates excluding all fixed effects and controls. Column (2) adds county fixed effects to the specification. Column (3) adds year fixed effects, while the estimates in Column (4) replicate the results presented in Table 1, Panel A. -63- Table A4: Additional Specifications – Arrests Arrests - Unweighted OUTCOME (1) (2) (3) (4) Total Arrests 811.7 -1,346*** -426.8 598.2 (1,238) (244.0) (395.5) (705.8) Violent Arrests 550.2 -227.9*** -416.9*** -86.38 (407.4) (86.86) (136.0) (188.6) Non-Violent Arrests -636.6* -1,496*** -520.1 564.6 (378.3) (351.6) (467.8) (655.7) Drug 898.0* 378.2* 510.2** 120.0 (542.2) (219.4) (243.0) (296.5) Robbery 32.96 -220.9*** -161.6* 82.27 (97.67) (82.49) (91.72) (121.5) Agg. Assault 208.2 26.64 78.22 140.2** (141.3) (42.08) (50.59) (70.33) Other Assault 302.2 3.375 -224.0** -255.4** (226.9) (60.38) (94.71) (112.2) Burglary -259.1*** -586.7*** -225.1 241.1 (78.42) (157.2) (188.6) (230.8) Larceny -464.8* -880.1*** -294.3 213.3 (262.2) (180.4) (260.3) (376.3) Vehicular 87.29 -29.56 -0.717 110.2 (69.13) (35.44) (51.36) (77.04) Observations 2,431 2,431 2,431 2,431 County FE N Y Y Y Year FE N N Y Y X Controls N N N Y Clusters 187 187 187 187 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Notes and Sources: For data, see notes to Table 2. Each coefficient is from a separate regression, reporting the estimate of 1 from Equation (1). Standard errors are reported in parenthesis and are clustered at the county. Column (1) reports the estimates excluding all fixed effects and controls. Column (2) adds county fixed effects to the specification. Column (3) adds year fixed effects, while the estimates in Column (4) replicate the results presented in Table 2, Panel A. -64- Table A5: Offenses Reported-Assign Ambiguous States Offenses Reported - Assign Ambiguous States Outcome (1) (2) (3) Total Crime -2,691*** -2,709*** -1,225*** (757.4) (715.3) (444.3) Violent Crime 87.06 1.369 246.8 (188.2) (166.6) (153.5) Non-Violent Crime -2,779*** -2,710*** -1,472*** (727.9) (691.2) (421.4) Burglary -1,053*** -1,039*** -655.2*** (315.8) (290.8) (211.9) Larceny -1,631*** -1,544*** -834.8*** (411.5) (397.2) (246.2) Observations 9,456 11,664 11,664 Ambig. State Rate N/A 0 100 Clusters 591 729 729 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Notes and Sources: For data, see notes to Table 1. Each coefficient is from a separate regression, reporting the estimate of 1 from Equation (1). Standard errors are reported in parenthesis and are clustered at the county. Column (1) replicates the main results from Table 1, Panel A. Column (2) assumes that the retention rate for seized assets is 0 percent in the states with missing forfeiture data. Column (3) assumes that the retention rate for seized assets is 100 percent in the states with missing forfeiture data. -65- Table A6: Arrests – Assign Ambiguous States Arrests - Assign Ambiguous States Outcome (1) (2) (3) Total 598.2 223.3 -753.5** (705.8) (601.0) (346.0) Violent -86.38 -38.04 -298.9*** (188.6) (117.6) (106.4) Non-Violent 564.6 181.9 -599.9* (655.7) (519.3) (324.0) Drug 120.0 79.51 145.2 (296.5) (112.3) (172.0) Robbery 82.27 -13.09 -142.1** (121.5) (82.72) (65.48) Agg. Assault 140.2** 43.51 -16.72 (70.33) (40.68) (36.33) Assault -255.4** -76.78 -133.2 (112.2) (51.18) (85.05) Burglary 241.1 53.99 -237.2** (230.8) (169.2) (119.1) Larceny 213.3 47.71 -353.3** (376.3) (295.5) (178.0) Vehicular 110.2 80.17 -9.317 (77.04) (65.25) (39.62) Observations 2,431 2,990 2,990 Ambig. State Rate N/A 0 100 Clusters 187 230 230 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Notes and Sources: For data, see notes to Table 2. Each coefficient is from a separate regression, reporting the estimate of 1 from Equation (1). Standard errors are reported in parenthesis and are clustered at the county. Column (1) replicates the main results from Table 2, Panel A. Column (2) assumes that the retention rate for seized assets is 0 percent in the states with missing forfeiture data. Column (3) assumes that the retention rate for seized assets is 100 percent in the states with missing forfeiture data. Texas A&M Journal of Property Texas A&M Journal of Property Law Law Volume 5 Number 3 Student Articles Article 10 6-1-2019 Guilty Until Proven Innocent: Rethinking Civil Asset Forfeiture and Guilty Until Proven Innocent: Rethinking Civil Asset Forfeiture and the Innocent Owner Defense the Innocent Owner Defense Luis Suarez Texas A&M University School of Law (Student), [email protected] Follow this and additional works at: https://scholarship.law.tamu.edu/journal-of-property-law Part of the Criminal Law Commons Recommended Citation Recommended Citation Luis Suarez, Guilty Until Proven Innocent: Rethinking Civil Asset Forfeiture and the Innocent Owner Defense, 5 Tex. A&M J. Prop. L. 1001 (2018). Available at: https://doi.org/10.37419/JPL.V5.I3.10 This Student Article is brought to you for free and open access by Texas A&M Law Scholarship. It has been accepted for inclusion in Texas A&M Journal of Property Law by an authorized editor of Texas A&M Law Scholarship. For more information, please contact [email protected] https://law.tamu.edu/ https://law.tamu.edu/ https://scholarship.law.tamu.edu/journal-of-property-law https://scholarship.law.tamu.edu/journal-of-property-law https://scholarship.law.tamu.edu/journal-of-property-law/vol5 https://scholarship.law.tamu.edu/journal-of-property-law/vol5/iss3 https://scholarship.law.tamu.edu/journal-of-property-law/vol5/iss3/10 https://scholarship.law.tamu.edu/journal-of-property-law?utm_source=scholarship.law.tamu.edu%2Fjournal-of-property-law%2Fvol5%2Fiss3%2F10&utm_medium=PDF&utm_campaign=PDFCoverPages http://network.bepress.com/hgg/discipline/912?utm_source=scholarship.law.tamu.edu%2Fjournal-of-property-law%2Fvol5%2Fiss3%2F10&utm_medium=PDF&utm_campaign=PDFCoverPages https://doi.org/10.37419/JPL.V5.I3.10 mailto:[email protected] \jciprod01productnTTWR5-3TWR310.txt unknown Seq: 1 15-APR-19 14:42 GUILTY UNTIL PROVEN INNOCENT: RETHINKING CIVIL ASSET FORFEITURE AND THE INNOCENT OWNER DEFENSE By: Luis Suarez † ABSTRACT Law enforcement departments across the country use civil asset forfeiture as a method to fund the work of law enforcement departments under the guise of combatting the “War on Drugs.” Attorney General Jeff Sessions made in- creasing civil asset forfeiture a DOJ priority. If civil asset forfeiture continues to rise to the level that Attorney General Sessions would like to see it, then we will soon find ourselves fighting to keep what is rightfully ours. This Comment will argue that the government should be required to prove that the owner of forfeited property had actual knowledge that the property was connected to an underlying crime. Dick M. Carpenter, Director for the Institute for Justice, believes that civil asset forfeiture is a thing of the past that today’s legal system should eschew. Civil asset forfeiture plays a relevant role in contemporary law enforcement, but additional safeguards should be enacted to ensure that civil asset forfeiture is not used at the expense of citizens’ property rights. Uniform reporting reform regarding forfeiture should occur amongst the states, and the government must prove that the innocent owner is not innocent. This is not to say that the government is required to succeed in a criminal prosecution before property can be forfeited, but this Comment argues that the government must prove from the onset that any owner of the property is not innocent and detached from the crime. Property owners should never be forced to prove their innocence without the constitutional protections guaran- teed in criminal courts. I. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1001 R II. HISTORY OF CIVIL ASSET FORFEITURE . . . . . . . . . . . . . . . . . 1004 R A. Modern Day Forfeiture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1004 R III. BURDEN OF PROOF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1006 R IV. PROCEDURAL ISSUES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1007 R V. FINANCIAL INCENTIVES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1008 R A. Inadequate Forfeiture Reporting . . . . . . . . . . . . . . . . . . . . . 1009 R B. Equitable Sharing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1010 R VI. DIFFICULTY OF PROVING INNOCENCE . . . . . . . . . . . . . . . . . . . 1012 R VII. COMPARISON AMONG STATE LAWS . . . . . . . . . . . . . . . . . . . . . 1016 R VIII. CONCLUSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1018 R I. INTRODUCTION On July 19, 2017, Attorney General Sessions signed an order that strengthened civil asset forfeiture programs by allowing the Depart- ment of Justice and other federal agencies to forfeit assets seized by state or local law enforcement, a process known as “federal adop- † Luis A. Suarez is a J.D. Candidate at Texas A&M University School of Law, class of 2019. 1001 DOI: https://doi.org/10.37419/JPL.V5.I3.10 \jciprod01productnTTWR5-3TWR310.txt unknown Seq: 2 15-APR-19 14:42 1002 TEX. A&M J. PROP. L. [Vol. 5 tions.”1This new order creates a stronger federal forfeiture program, which is troubling news for citizens like Charles Clarke. In 2014, the Cincinnati law enforcement officials seized $11,000 in cash from Charles Clarke.2 Clarke lost his entire life savings because the officials claimed that his bag smelled of marijuana.3 Officials did not find any drugs, illegal items, or evidence connected to a criminal activity on Clarke’s body or in his carry-on.4 Regardless, the law en- forcement officials seized all of the cash found in Clarke’s carry-on before proving that the cash had a probable relation to a crime.5 Current civil asset forfeiture law allows law enforcement to seize property they suspect is connected to criminal activity.6 Police may seize cash, cars, jewelry, and homes—even without a warrant.7 So long as law enforcement officials have probable cause to believe that the property being seized is subject to forfeiture and the seizure is made pursuant to a lawful arrest or search, then law enforcement officials can effectively seize a person’s property without any judicial over- sight.8 Officials can bring actions seeking forfeiture as in rem proceed- ings, which allow the government to circumvent bringing an action against the property owner by filing an action against the property itself.9 When beginning civil forfeiture proceedings, the prosecution must send a written notice to interested parties within sixty days of seizing the property.10 An interested party must file the claim to the property no later than thirty days after the date of final publication of the no- tice of seizure.11 The government designates an interested party if no person claims an interest in the property.12 After a claim has been made, the government has ninety days to file a complaint for forfei- ture against the property.13 The government has an advantage over 1. Att’y Gen. Order No. 3946-2017 (July 19, 2017); see also U.S. Dep’t of Just., Crim. Div., Policy Directive 17-1 on Attorney General’s Order on Federal Adoption and Forfeiture of Property Seized by State and Local Law Enforcement Agencies (2017). 2. DICK M. CARPENTER II ET AL., INST. FOR JUST., POLICING FOR PROFIT: THE ABUSE OF CIVIL ASSET FORFEITURE 8 (2d ed. 2015), https://ij.org/wp-content/ uploads/2015/11/policing-for-profit-2nd-edition.pdf [https://perma.cc/G6KF-VFGF]. 3. Id. 4. Id. 5. Id. 6. Id. 7. 18 U.S.C. § 981(b)(2) (2012). 8. Id. 9. Rachel L. Stuteville, Reverse Robin Hood: The Tale of How Texas Law En- forcement Has Used Civil Asset Forfeiture to Take From Property Owners and Pad The Pockets Of Local Government-the Righteous Hunt for Reform Is On, 46 TEX. TECH L. REV. 1169, 1176 (2014). 10. § 983(a)(1)(A)(i). 11. § 983(a)(2)(B). 12. § 983(a)(1)(A)(v). 13. § 983(a)(3)(A). \jciprod01productnTTWR5-3TWR310.txt unknown Seq: 3 15-APR-19 14:42 2019] CIVIL ASSET FORFEITURE 1003 interested parties at the time the complaint is filed because 18 U.S.C. § 983 does not allow the dismissal of a complaint for failure to estab- lish that the property was forfeitable on the ground that the govern- ment possessed inadequate evidence at the time the complaint was filed.14 The burden of proof during forfeiture proceedings is on the govern- ment to establish by a preponderance of the evidence that the prop- erty is subject to forfeiture.15 Furthermore, if the government relies on the theory that the property was used to facilitate or commit a crime, then the government must also establish that a substantial connection existed between the property and the offense.16 Moreover, the pro- ceeding is the first time that the owner of the property is offered the opportunity to regain ownership of his or her property. The owner may raise an “innocent owner” defense, which would stop the govern- ment from forfeiting the property interest under any civil forfeiture statute.17 Innocent owners must prove by a preponderance of the evidence that they “did not know of the conduct giving rise to forfeiture; or upon learning of the conduct giving rise to forfeiture, did all that rea- sonably could be expected under the circumstances to terminate such use of the property.”18 Therefore, because the forfeiture is character- ized as a civil matter, property owners are not afforded the constitu- tional procedural protections guaranteed in criminal proceedings, such as a right to a jury trial and requiring the government to prove its case beyond a reasonable doubt.19 Historically, courts permitted civil asset forfeiture actions but acknowledged that the actions were crimi- nal in nature.20 This Comment proposes a change of civil asset forfeiture laws at both the federal and state levels due to the unjust difficulties that property owners face throughout the country. The Comment begins by discussing the roots of civil asset forfeiture and how the criminal nature of forfeiture should impact the standard of proof and the pro- cedural posture of forfeiture proceedings. The Comment’s latter half discusses the potential for abuse that civil asset forfeiture creates when considering law enforcement agencies’ financial incentive to pursue forfeiture and the difficulty property owners face when chal- lenging a forfeiture. Finally, the Comment concludes with a compari- son of forfeiture laws amongst the states, and a proposal of uniformity of those laws to better protect citizens’ property rights. 14. § 983(a)(3)(D). 15. § 983(c)(1). 16. § 983(c)(3). 17. § 983(d)(1). 18. § 983(d)(2)(A). 19. Leonard v. Texas, 137 S. Ct. 847, 849 (2017) (mem.). 20. Boyd v. United States, 116 U.S. 616, 633-34 (1886). \jciprod01productnTTWR5-3TWR310.txt unknown Seq: 4 15-APR-19 14:42 1004 TEX. A&M J. PROP. L. [Vol. 5 II. HISTORY OF CIVIL ASSET FORFEITURE Civil asset forfeiture traces back to biblical times when it was com- mon practice to relinquish anything connected to one’s wrongdoing over to God.21 Many believed that an object or piece of property could be involved in a wrongdoing and the object or property itself should be held responsible for its wrongdoing.22 Early English com- mon law further developed this line of thought with the Deodand pro- cedure that allowed the King to forfeit any property that caused the deaths of English citizens.23 Originally, the property seized under the Deodand procedure was used for religious purposes. The procedure evolved, however, into a source of revenue for the King.24 Early United States forfeiture law expanded this concept in the realm of admiralty law.25 The United States invoked forfeiture pro- ceedings against ships that violated the law of the high seas for crimes such as slave trafficking.26 In effect, the United States would bring an action against a ship for a crime such as slave trafficking and then seize the ship through forfeiture proceedings. The Supreme Court de- termined that a violation of the law of nations necessitated seizure of ships because seizure was the only adequate means of suppressing the offense or wrong.27 Additionally, another earlier Court held that a forfeiture proceeding can be against the ship rather than the ship’s owner, thereby disregarding the will of the owner and the innocent owner defense.28 The logic used by these earlier Court decisions in admiralty law created the framework for future civil asset forfeiture proceedings. A. Modern Day Forfeiture Modern day civil asset forfeiture did not become widely used until the “war on drugs” began during the Nixon and Reagan administra- tions.29 In the 1980s, Congress strengthened civil asset forfeiture by amending the Comprehensive Drug Abuse and Prevention Act of 1970.30 The Act’s original goals were to prevent drug abuse, to pro- vide more effective means of law enforcement, and to provide a bal- 21. Mike Fishburn, Gored by the Ox: A Discussion of the Federal and Texas Laws that Empower Civil-Asset Forfeiture, 26 RUTGERS L. REC. 4 (2002). 22. Id. 23. Id. 24. Id. 25. Stuteville, supra note 10, at 1178. 26. Id. 27. Harmony v. United States, 43 U.S. 210, 233 (1844). 28. United States v. The Little Charles, 26 F. Cas. 979, 982 (1818). 29. Stuteville, supra note 10, at 1179. 30. Id. \jciprod01productnTTWR5-3TWR310.txt unknown Seq: 5 15-APR-19 14:42 2019] CIVIL ASSET FORFEITURE 1005 anced scheme of criminal penalties for drug offenses.31 The amendments that followed authorized forfeiture of proceeds from drug related offenses, as well as forfeiture from any property that fa- cilitated drug offenses.32 In 1984, Congress passed another amend- ment to the Act that created the Department of Justice Assets Forfeiture Fund.33 The fund authorized the Attorney General to reim- burse any federal, state, or local agency for any costs necessary for forfeiture or costs that were incident to forfeiture.34 In 2000, Congress passed the Civil Asset Forfeiture Reform Act (“CAFRA”) to “provide a more just and uniform procedure for Fed- eral civil forfeitures.”35 During the June 24, 1999, floor debate, Con- gresswoman Deborah Pryce stated that the “current civil asset forfeiture laws, at their core, deny basic due process, and the Ameri- can people have reason to be both offended and concerned by the abuse of individual rights. . .”36 CAFRA became law at a time when people began to realize that civil asset forfeiture was tilted too far in the government’s favor. Even Attorney General Sessions, then a United States Senator, re- alized that civil asset forfeiture laws needed reform. During the Sen- ate Proceedings and Debates on October 6, 1999, Senator Sessions introduced the Sessions–Schumer Civil Asset Forfeiture Reform Act of 1999.37 The Sessions–Schumer bill, which would later be incorpo- rated in part with CAFRA, raised the burden of proof on the govern- ment in forfeiture proceedings from probable cause to preponderance of the evidence.38 CAFRA also created the “innocent owner defense” for owners, which stated that “an innocent owner’s interest in prop- erty shall not be forfeited under any civil forfeiture statute.”39 When asserting the innocent owner defense, however, the burden is on the claimant to prove by a preponderance of the evidence that he or she is actually innocent.40 CAFRA provided greater protection to individuals by allowing them to prove their innocence in forfeiture proceedings by a prepon- 31. David L. Luck, Guns, Drugs, and . . . Federalism? —Gonzales v. Raich Enfee- bles the Rehnquist Court’s Lopez-Morrison Framework, 61 U. MIAMI L. REV. 237, 246 (2006). 32. Stuteville, supra note 10, at 1179; See 21 U.S.C.S. § 881(a)(7) (2012). 33. Comprehensive Crime Control Act of 1984, 28 U.S.C. § 524 (2012). 34. Id. 35. Civil Asset Forfeiture Reform Act of 2000, Pub. L. No. 106-185, 114 Stat. 202 (2000). 36. 145 CONG. REC. H4,852 (daily ed. June 24, 1999) (statement of Rep. Pryce), 145 Cong Rec H 4851-01, at *H4,852 (Westlaw). 37. 145 CONG. REC. S12,108 (daily ed. Oct. 6, 1999) (statement of Sen. Sessions), 145 Cong Rec S 12101-03, at *S12,108 (Westlaw). 38. Id. 39. Civil Asset Forfeiture Reform Act of 2000, Pub. L. No. 106-185, 114 Stat. 202 (2000). 40. Id. \jciprod01productnTTWR5-3TWR310.txt unknown Seq: 6 15-APR-19 14:42 1006 TEX. A&M J. PROP. L. [Vol. 5 derance of the evidence. Nevertheless, CAFRA still provided law en- forcement the ability to benefit from the proceeds of civil forfeiture.41 Since CAFRA was adopted, the reach of civil forfeiture proceedings has expanded at the federal and state levels.42 As a result, forfeiture has expanded long past its biblical roots. Despite its protections, United States citizens are still unjustly deprived of their property every year. III. BURDEN OF PROOF Most state forfeiture laws currently require law enforcement offi- cials to show probable cause that the property is connected to a crime.43 Because the proceeding is civil, property owners are not enti- tled to the right to counsel. If property owners want to regain their property, then the property owners must find legal counsel to re- present them or represent themselves pro se.44 By making the pro- ceedings civil, the government may circumvent the procedural safeguards that are offered in criminal cases such as the right to coun- sel. The lack of procedural safeguards makes forfeiture susceptible to abuse by law enforcement. Further, law enforcement departments are motivated to seize properties and assets because they profit from its forfeiture.45 Al- lowing law enforcement agencies to benefit from forfeiture promotes abuse by law enforcement, especially when there are few procedural safeguards prior to the civil proceedings.46 Citizens would be more protected if the government was required to prove that the property owner had actual knowledge of the underlying crime. Law enforce- ment would then less likely seize property simply for monetary gain. Raising the burden of proof on the government at the time of the seizure would minimize the frequent application of forfeiture pro- ceedings. A study conducted by the Inspector General found that sev- enty-nine out of 100 sample Drug Enforcement Administration (“DEA”) seizures occurred absent any preexisting intelligence of a specific drug crime.47 This could explain the reason why 87% of for- feitures at the federal level are pursued through civil rather than crim- inal proceedings.48 Furthermore, most forfeiture proceedings are never challenged because of the associated legal difficulties.49 For ex- 41. CARPENTER II ET AL., supra note 3, at 10. 42. Id. 43. Id. at 11. 44. Id. at 12. 45. Id. at 11-12. 46. Id. at 11. 47. OFFICE OF INSPECTOR GEN., U.S. DEP’T OF JUSTICE, REVIEW OF THE DEPART- MENT’S OVERSIGHT OF CASH SEIZURE AND FORFEITURE ACTIVITIES 22 (2017), https:/ /oig.justice.gov/reports/2017/e1702.pdf [https://perma.cc/3LTQ-3WRF]. 48. CARPENTER II ET AL., supra note 3, at 12. 49. Id. \jciprod01productnTTWR5-3TWR310.txt unknown Seq: 7 15-APR-19 14:42 2019] CIVIL ASSET FORFEITURE 1007 ample, defendants in Illinois are required to pay the greater of $100 or 10% of the property’s value before challenging the seizure proceeding to prove the individual’s innocence.50 The difficulty and costs associ- ated with trying to retrieve the property partially is the reason why 90% of judicial cash forfeitures are uncontested.51 IV. PROCEDURAL ISSUES Forfeiture proceedings may be brought in rem, which allows the government to bring an action against the property rather than the individual.52 In rem proceedings are beneficial to the government for a few reasons. First, in rem proceedings have different evidentiary rules and procedures than criminal proceedings. Also, the culpability of the party does not need to be proven for in rem proceedings.53 This means that the government does not need to accompany the in rem proceed- ing with a criminal proceeding at all.54 Further, the standard of proof is lower for in rem proceedings as opposed to criminal proceedings. The standard used in most states is “preponderance of the evidence,” which requires that the government prove the property is more likely connected to a crime than not.55 Civil asset forfeiture standards are considered constitutional because the proceeding is in rem and there- fore is not offending any individuals’ due process rights.56 The low standard of proof makes civil forfeiture cases easy for the government but difficult for the claimants to win.57 The federal government can also avoid judicial proceedings through a process known as administrative forfeiture.58 Administrative forfei- ture can be used for personal property, including cash.59 If no party files a claim within the twenty-day deadline prescribed by statute, then an administrative official can declare the property forfeited.60 After the government seizes property, it must provide written notice to in- terested parties within sixty days of the seizure.61 Once the notice is publicized or sent to the interested parties, they have thirty days to file 50. Id. 51. Eric Blumenson & Eva Nilsen, Policing for Profit: The Drug War’s Hidden Economic Agenda, 65 U. CHI. L. REV. 35, 50 (1998). 52. See 18 U.S.C. § 985(c)(3) (2012). 53. William Carpenter, Reforming the Civil Drug Forfeiture Statutes: Analysis and Recommendations, 67 TEMP. L. REV. 1087, 1096 (1994). 54. The Palmyra, 25 U.S. (12 Wheat.) 1, 15 (1827). 55. CARPENTER II ET AL., supra note 3, at 16. 56. David Ross, Civil Forfeiture: A Fiction That Offends Due Process, 13 REGENT U.L. REV. 259, 263 (2001). 57. CARPENTER II ET AL., supra note 3, at 18. 58. Caleb Nelson, The Constitutionality of Civil Forfeiture, 125 YALE L.J. 2446, 2449 (2016). 59. 19 U.S.C. § 1607(a) (2012). 60. Id. § 1609(a). 61. 18 U.S.C. § 983(a)(1) (2012). \jciprod01productnTTWR5-3TWR310.txt unknown Seq: 8 15-APR-19 14:42 1008 TEX. A&M J. PROP. L. [Vol. 5 a claim.62 The interested party must then file an answer to the govern- ment’s complaint within twenty days.63 Once the twenty days have passed, a customs officer declares the property forfeited and sells the property at a public auction.64 The interested party can petition the seizing agency after the administrative forfeiture; however, the seizing agency evaluates the petitioner’s interest rather than the court.65 Administrative forfeiture can occur on any property valued under $500,000.66 Statistics suggest that approximately 81% of all DEA cash seizures have been administratively forfeited at a value of over $3.2 billion.67 Uncontested forfeitures have two likely causes: (1) the for- feiture is valid, and therefore, no one is willing to contest it; or (2) the process to challenge the forfeiture is unduly burdensome, and there- fore, claimants are hesitant to challenge the forfeited property.68 Therefore, it is not difficult to understand why 80% of forfeitures go uncontested in light of the government’s initial benefit from the pro- ceedings and the difficulties the proceedings impose on property owners. For Attorney General Sessions, civil asset forfeitures and the ac- companying proceedings benefit the public at large because “it helps return property to the victims of crime.”69 However, a study of 100 DEA cases, which involved warrantless searches and seizures that turned up no illicit narcotics, showed that over half of the seizures had “no discernable connection between the seizure and the advancement of law enforcement efforts.”70 The administrative forfeiture gives the government a lower standard of proof to meet, allowing them to sub- sequently forfeit the cash when it is not found to further law enforce- ment efforts. V. FINANCIAL INCENTIVES At the federal level, the Department of Justice (“DOJ”) created the Asset Forfeiture Program (“AFP”) in 1984 to support the use of asset forfeiture.71 The AFP is comprised of agencies that deposit any assets 62. Id. § 983(a)(4)(A). 63. Id. § 983(a)(4)(B). 64. 19 U.S.C. § 1609(a) (2012). 65. OFFICE OF INSPECTOR GEN., supra note 51, 13-14. 66. Id. at 7. 67. Id. at 13. 68. See CARPENTER II ET AL., supra note 3, at 18. 69. Attorney General Sessions Issues Policy and Guidelines on Federal Adoptions of Assets Seized by State or Local Law Enforcement, U.S. DEP’T JUST. (July 19, 2017), https://www.justice.gov/opa/pr/attorney-general-sessions-issues-policy-and-guidelines- federal-adoptions-assets-seized-state [https://perma.cc/F3HZ-XKPE]. 70. OFFICE OF INSPECTOR GEN., U.S. DEP’T OF JUSTICE, REVIEW OF THE DEPART- MENT’S OVERSIGHT OF CASH SEIZURE AND FORFEITURE ACTIVITIES 21 (2017). 71. OFFICE OF INSPECTOR GEN., U.S. DEP’T OF JUSTICE, AUDIT OF THE ASSETS FORFEITURE FUND AND SEIZED ASSET DEPOSIT FUND ANNUAL FINANCIAL STATE- MENTS FISCAL YEAR 2016 4 (2016). \jciprod01productnTTWR5-3TWR310.txt unknown Seq: 9 15-APR-19 14:42 2019] CIVIL ASSET FORFEITURE 1009 into the Assets Forfeiture Fund (“AFF”).72 In return, those agencies are eligible to receive an annual allocation of resources from the fund.73 The AFF and the Seized Asset Deposit Fund (“SADF”) com- bine to form a financial reporting entity of the DOJ, which includes cash and property seized by forfeiture.74 Assets are held in the SADF until the end of a successful forfeiture action, at which time they are transferred to the AFF, and then the AFF expends the assets to the agencies that are in the AFP.75 The AFF’s funds cover the operating costs of the AFP such as payments to innocent third-party claimants; federal and state task force expenses; and forfeiture training.76 The AFF receives revenue from the forfeited cash, other assets, and the sale of forfeited property.77 In 2017, the total amount of resources in the AFF was $1,455,113.78 In 2017, an independent audit on the AFF and SADF noted defi- ciencies in both funds’ internal controls over financial reporting.79 The audit also noted that the management of neither fund had controls in place to ensure that: (1) revenue was recognized in the appropriate accounting period; (2) journal entries properly represented the ac- counting events; and (3) budgetary information in the financial state- ments was properly reported and presented.80 Further, the audit noted that the lack of management of the AFF and SADF caused “incom- plete and inaccurate information in the Consolidated Assets Tracking System impacting revenue cut-off and regulation” as well as “insuffi- cient review of manual journal entries.”81 Issues with reporting and record keeping are even worse at the state level. Twenty-six states do not require any public reporting system at all.82 Fifteen states practice an accounting system similar to the federal government’s system, whereby an agency compiles an aggregate report on forfeiture.83 The detail and level of the report, however, varies among the states.84 A. Inadequate Forfeiture Reporting Most state reports fail to provide adequate information needed for proper assessment of law enforcement’s forfeiture use. Some states require that law enforcement agencies report their forfeiture use for 72. Id. at 3. 73. Id. at 5. 74. Id. at 4. 75. Id. 76. Id. 77. Id. at 5. 78. Id. at 4. 79. Id. at 21. 80. Id. at 23. 81. Id. at 18-19. 82. CARPENTER II ET AL., supra note 3, at 33-35. 83. Id. 84. Id. \jciprod01productnTTWR5-3TWR310.txt unknown Seq: 10 15-APR-19 14:42 1010 TEX. A&M J. PROP. L. [Vol. 5 purposes of record keeping or public accounting.85 Even though some states and the federal government require that data be made public, another issue arises when the information is either inadequate or in- complete.86 For example, most state reports do not distinguish be- tween criminal or civil forfeitures or whether the person from which the property was seized was ever convicted.87 Further, most state re- ports lack essential information such as the type of property seized or forfeited; the size or amount of seized property; the average size of seized property; and the value in property retained by the local law enforcement.88 An additional issue regarding state reports is the complete omission of data, as was the case for at least nine states.89 In 2013, Minnesota’s Office of the State Auditor reported that fifty-three law enforcement agencies throughout the state failed to file a criminal forfeiture re- port.90 Minnesota is not alone in this regard. Michigan was missing data from fifty-six agencies; Kentucky was missing information from 178 agencies; and Washington was missing data from forty-three agen- cies.91 Several other law enforcement agencies across the country failed to report forfeiture data that was required by state law.92 Few states report data, and of those few states, most of them either: (1) fail to report useful data, (2) fail to report data from multiple law enforce- ment agencies within the state, (3) or fail to report both. State and federal forfeiture reporting is meager at best. Aside from the federal government, no jurisdiction requires reports regarding the expenditure of the funds acquired by forfeiture.93 The federal report- ing requirements, as discussed earlier, do not require the audit to go into specific detail about how the funds are spent. The DOJ’s reports omit information about how individual agencies spend their money from the AFF.94 The expenditure reports do cover a few general top- ics, including equitable sharing payments to states.95 B. Equitable Sharing Equitable sharing promotes cooperation between the state govern- ments, state law enforcement agencies, and the federal government in 85. Id. at 36. 86. Id. 87. Id. 88. Id. 89. Id. 90. See REBECCA OTTO, OFFICE OF STATE AUDITOR, CRIMINAL FORFEITURES IN MINNESOTA (2013). 91. CARPENTER II ET AL., supra note 3, at 37. 92. Id. at 36. 93. Id. at 39. 94. See generally OFFICE OF INSPECTOR GEN., supra note 76. 95. CARPENTER II ET AL., supra note 3, at 39. \jciprod01productnTTWR5-3TWR310.txt unknown Seq: 11 15-APR-19 14:42 2019] CIVIL ASSET FORFEITURE 1011 order to forfeit property under federal laws.96 Equitable sharing also allows states to process forfeiture under other forms besides state law.97 Essentially, any law enforcement agency that forfeits any prop- erty under the presumption that the property owner violated a federal law may request a share of the forfeiture’s net proceeds through equi- table sharing.98 A federal authority then approves or denies the state’s request for equitable sharing.99 In practice, the state or local law enforcement agency may seize the property and then requests a federal agency to adopt the seizure and proceed with the forfeiture.100 The federal agency is authorized to adopt the seizure if it evidences a violation of federal law.101 Equitable sharing may arise in joint investigations when state and local law en- forcement agencies work with federal agencies to enforce federal criminal laws.102 Federal agencies have minimum threshold require- ments to consider before they adopt a forfeiture, such as a minimum of a $2,000 value on forfeitures of currency, bank accounts, monetary instruments, and jewelry.103 These minimum requirements, however, may be waived if the “forfeiture will serve a compelling law enforce- ment interest.”104 The United States Attorney General has the discretion to share fed- erally forfeited property with participating state and local law enforce- ment agencies.105 The amount that state law enforcement agencies receive through equitable sharing is calculated from the forfeiture’s net proceeds, but the amount is ultimately decided by an equitable sharing deciding authority.106 In determining the amount to be shared, the deciding authority takes the gross amount from the forfeited prop- erty less a number of factors such as qualified third-party interests, money paid to victims, and any award paid to a federal informant.107 After the factors are subtracted from the gross amount of the forfeited property, the deciding authority may distribute funds to any local or state law enforcement agencies or may deposit funds into AFF.108 96. OFFICE OF INSPECTOR GEN., supra note 76, at 6. 97. Simon James, Note, Civil Asset Forfeiture in Virginia: An Imperfect System, 74 WASH. & LEE L. REV. 1295, 1312-13 (2017). 98. ASSET FORFEITURE & MONEY LAUNDERING SECTION, U.S. DEP’T OF JUSTICE, GUIDE TO EQUITABLE SHARING FOR STATE AND LOCAL LAW ENFORCEMENT AGEN- CIES 3 (2009). 99. Id. 100. Id. at 6. 101. Id. 102. Id. 103. Id. at 7. 104. Id. 105. 21 U.S.C. § 881(e)(1)(A) (2012). 106. Id. at 12. 107. Id. at 15. 108. Id. \jciprod01productnTTWR5-3TWR310.txt unknown Seq: 12 15-APR-19 14:42 1012 TEX. A&M J. PROP. L. [Vol. 5 Equitable sharing in practice gives state and local law enforcement agencies a wider branch of law in which to forfeit property because it allows them to also forfeit under federal forfeiture laws while taking advantage of lower standards of proof in states where the burden of proof is higher than the federal standard.109 Further, states utilize eq- uitable sharing for instances when the action is legal under state law but violates federal law, such as marijuana possession in states that have legalized it.110 There is a strong incentive for state and local law enforcement agen- cies to participate in equitable sharing because they can receive up to 80% of the proceeds from a federal-based forfeiture.111 Equitable sharing gives states the opportunity to circumvent state law by using the federal law and still receive a lump sum, 80%, of the profit from the forfeiture.112 From 2000 to 2013, equitable sharing payments from the DOJ to state and local law enforcement agencies tripled, rising from $198 million to $643 million.113 Further, state and local law en- forcement agencies rely heavily on the funding from forfeiture and equitable sharing. Equitable sharing is referred by some as “a virtual cash cow.”114 A 2014 investigation showed that 298 departments and 210 task forces across the nation used equitable sharing to make up approximately 20% of their annual budgets.115 The need for greater protection of individuals’ property rights is ev- ident when considering the lack of reporting mixed with the ability of departments to fund their budgets through forfeiture and other ave- nues such as equitable sharing. Even in states with reporting require- ments, the reports often do not have the necessary information to hold the state and local law enforcement agencies accountable for their forfeiture practices. The difficulty for property owners to prove their innocence, coupled with the powerful financial incentive that forfeiture gives to law enforcement officials, creates a system ripe for abuse. VI. DIFFICULTY OF PROVING INNOCENCE In the United States, the “innocent until proven guilty” standard should apply to civil forfeiture proceedings. For example, an analysis using the innocent until proven guilty standard could have been life changing for Rochelle Bing. In Philadelphia, Bing purchased a home 109. James, supra note 103, at 1313. 110. Id. 111. CARPENTER II ET AL., supra note 3, at 6. 112. Id. 113. Id. 114. Id. at 28. 115. Michael Sallah, Stop and Seize, WASH. POST (Sept. 6, 2014), http://www.wash ingtonpost.com/sf/investigative/2014/09/06/stop-and-seize/?utm_term=.3d27c64cedc6 [https://perma.cc/MA6N-3UFX]. \jciprod01productnTTWR5-3TWR310.txt unknown Seq: 13 15-APR-19 14:42 2019] CIVIL ASSET FORFEITURE 1013 to house eighteen of her grandchildren while their parents were at work, and she was forced to prove her innocence.116 Unbeknownst to Bing, one of her grandchildren was accused of selling crack from the house.117 Bing’s case lasted for two years and she had to appear at court with her attorney twenty-three times.118 The case was eventually settled and Bing was able to retain ownership under certain condi- tions.119 An instance like this illustrates the importance of applying the innocent until proven guilty standard to forfeiture proceedings. In American criminal and civil legal practice, the presumption of innocence acts as a barrier against punishment before conviction.120 The maxim “innocent until proven guilty” applies only to criminal de- fendants. However, throughout history courts around the world have applied the maxim in civil cases as well.121 For example, in France the presumption of innocence is an individual’s personal right even in civil instances.122 The maxim’s importance in criminal trials is great; how- ever, the maxim is equally necessary when an individual is losing his or her property. The creation of the innocent owner defense in CAFRA was a big step forward to give property owners protection from law enforce- ment. Nevertheless, fighting the government over property may be- come more difficult than living life without it.123 Individuals asserting the innocent owner defense face a significant disadvantage due to the costs of litigation, length of time, and the re- sources required to sustain litigation.124 In Arizona, Rhonda Cox en- ded her pursuit to win back the title to her truck after discovering the obstacles of challenging a civil forfeiture.125 Initiating the proceeding can cost up to $304 in filing fees.126 The filing fees alone could be enough to dissuade some families from pursuing their own property. Like Cox, the individual may be told that on top of the filing fees, he or she must pay for the county’s attorney fees and investigations costs, which could potentially exceed the property’s value.127 The struggles 116. Isaiah Thompson, How Civil Forfeiture Laws Are Costing Innocent People Their Homes, PAC. STANDARD (Aug. 22, 2013), https://psmag.com/news/law-to-clean- up-nuisances-costs-innocent-people-their-homes-64809 [https://perma.cc/J2X7-REK A]. 117. Id. 118. Id. 119. Id. 120. Francois Quintard-Morenas, The Presumption of Innocence in the French and Anglo-American Legal Traditions, 58 AM. J. COMP. L. 107 (2010). 121. See generally id. 122. Id at 149. 123. CARPENTER II ET AL., supra note 3, at 18. 124. Id. at 18. 125. Id. 126. Id. 127. Cox v. Voyles, Et. Al., ACLU (Aug. 21, 2017) https://www.aclu.org/cases/cox-v- voyles-et-al [https://perma.cc/3LF7-TFQV]. \jciprod01productnTTWR5-3TWR310.txt unknown Seq: 14 15-APR-19 14:42 1014 TEX. A&M J. PROP. L. [Vol. 5 Cox encountered are commonplace for individuals attempting to win back their property. Additionally, Philadelphia requires individuals to show up in Court- room 478, which was run by district attorneys, with no presiding judge or jury.128 As a result, 83% of the proceedings were decided in the first appearance in Courtroom 478, and of those proceedings, 96% were decided in the government’s favor.129 If a case did not end in the initial proceeding, the district attorneys could “relist” the case, requir- ing the individuals to return to the courtroom at a later date.130 If an individual misses just one relisted court date, the government could then immediately forfeit the property.131 The schemes like those in Arizona and Philadelphia, where state actors effectively persuaded in- dividuals to not pursue the innocent owner defense, is the first burden individuals must pass before retrieving their own property. Federal law does allow for individuals asserting the defense to de- mand the release of their property while the proceedings are con- ducted.132 In order to force the government to release the seized property, the claimant must have a possessory interest in the property and sufficient ties to the community to assure that the property will be available at trial.133 Furthermore, the claimant must prove that contin- ued possession by the government will cause the claimant substantial hardship and that the hardship outweighs the risk that the property will be destroyed if returned to the claimant.134 However, the individ- ual who lost their property cannot demand that it be released during the proceeding if the property is contraband, currency, electronic funds, or other monetary instruments.135 Also, the individual cannot demand release of any property that is (1) used as evidence of a viola- tion of law; (2) is particularly suited for use in illegal activities; or (3) is likely used to commit additional criminal acts if returned to the claimant.136 The ability to demand release of the property while pro- ceedings are in place seems beneficial to the claimants; however, prov- ing that continued possession by the government will result in substantial hardship is difficult in practice.137 128. Nick Sibilla, Philadelphia Earns Millions By Seizing Cash and Homes From People Never Charged With a Crime, FORBES (Aug. 26, 2014, 09:17 AM), https://www .forbes.com/sites/instituteforjustice/2014/08/26/philadelphia-civil-forfeiture-class-ac tion-lawsuit/#601575345d9e [https://perma.cc/WJQ4-EDNJ]. 129. Id. 130. Id. 131. Id. 132. 18 U.S.C. § 983(f)(1) (2012). 133. Id. § 983(f)(1)(A)–(B). 134. Id. § 983(f)(1)(C)–(D). 135. Id. § 983(f)(8)(A). 136. Id. § 983(f)(8)(B)–(D). 137. See Generally United States v. $9,451.00 United States Currency, No. 14-CV- 275-LM, 2015 WL 965819 (D.N.H. Mar. 3, 2015). \jciprod01productnTTWR5-3TWR310.txt unknown Seq: 15 15-APR-19 14:42 2019] CIVIL ASSET FORFEITURE 1015 Once the individual commits significant time and money, the indi- vidual’s struggle to prove his or her innocence begins. Federal law and thirty-five states place the burden of proof on the owners to prove that they had no knowledge that the property was being used for ille- gal purposes, nor did they consent to use of the property for illegal purposes.138 Proving innocence is difficult because the individual must prove a negative—that they did not know about criminal activity re- lated to their property.139 In $18,800 in United States Currency v. State, the claimant failed to affirmatively assert that she had no knowledge of the criminal activity.140 The evidence in the record showed that the claimant had recently kicked out her boyfriend who was accused of the crime, was not present at the arrest, and had not been charged with any connection to the offense.141 The Court still held against the owner, reasoning that insufficient evidence existed to establish that the claimant did not know of the criminal activity.142 Further, innocent owners may have no redress when the property is jointly held. In Laase v. 2007 Chevrolet Tahoe, a married couple in Minnesota were joint owners in a Chevrolet Tahoe that was seized after the police arrested the wife for a driving while intoxicated and charged for a second-degree criminal test refusal.143 The husband had no knowledge that his wife was drinking that evening nor that his wife was driving while intoxicated.144 However, the Supreme Court of Min- nesota held in favor of the government, determining that all owners must be innocent for the innocent owner defense to apply.145 The innocent owner defense creates the illusion that individuals are afforded proper redress to retain their property when it is taken by the government. Initiating the procedure can be costly and time con- suming. When and if the individual actually gets to litigation, the diffi- culty in proving one’s innocence is a task that even our nation’s founders believed would never have to be surmounted. As James Madison once said, “[t]he personal right to acquire property, which is a natural right, gives to property, when acquired, a right to protection, as a social right.”146 Proving one’s innocence to retain one’s right to property is a burden that is extremely difficult to prove, and individu- als learn of the uphill struggle to retain their property rights when their property is seized by law enforcement. Therefore, it is not sur- 138. CARPENTER II ET AL., supra note 3, at 20. 139. Id. 140. 961 S.W.2d 257, 262 (Tex. App.—Houston [1st Dist.] 1997, no writ). 141. Id. 142. Id. 143. 776 N.W.2d 431, 433 (Minn. 2009). 144. Id at 432-33. 145. Id at 439. 146. Steve Straub, James Madison, Speech on the Virginia Constitutional Conven- tion, FEDERALIST PAPERS (Sept. 3, 2013, 09:03 AM), https://thefederalistpapers.org/ founders/madison/james-madison-speech-on-the-virginia-constitutional-convention [https://perma.cc/3QYX-298Q]. \jciprod01productnTTWR5-3TWR310.txt unknown Seq: 16 15-APR-19 14:42 1016 TEX. A&M J. PROP. L. [Vol. 5 prising that 90% of forfeitures are uncontested while considering that only 8% of cash seizures made by the DEA between 2007 and 2016 were eventually returned to their owners.147 Individuals know that the likelihood of regaining possession of property is slim to none, and if they are able to retrieve their property, it may be cost-prohibitive. VII. COMPARISON AMONG STATE LAWS Being modeled after federal law, most states’ civil asset forfeiture laws are disadvantageous towards their residents. However, jurisdic- tions with looser forfeiture laws make forfeiting actions easier for law enforcement, which increase the forfeiture victimization rate. For example, Massachusetts arguably has the loosest civil forfeiture laws in the country.148 Similar to federal law, Massachusetts law also places the burden on innocent owners to prove that they are inno- cent.149 Civil forfeiture laws in Massachusetts are possibly the most unjust in the country because of factors such as the standard of proof that is required of state law enforcement officials to forfeit property. As in most states, preponderance of the evidence is a relatively easy burden of proof. However, Massachusetts law only requires the state to prove probable cause to bring a civil forfeiture cause of action.150 Probable cause might be considered the easiest burden to prove. Fur- ther, the financial incentive from the forfeited funds could be another factor that persuades Massachusetts law enforcement agencies to pur- sue civil forfeiture more aggressively. All monies that are collected from forfeiture proceedings are placed in special trust funds for the local or state police and the district attor- ney.151 Probable cause as the standard of retaining forfeited property, combined with the possibility of retaining 100% of funds, creates an almost insurmountable incentive for law enforcement officials to ag- gressively pursue civil forfeiture in Massachusetts. Unlike Massachusetts, many states do not make civil forfeiture an easy project, and a few states are giving their residents greater consti- tutional protections against it. In ten states and the District of Colum- bia, the government bears the burden of proving that the owners committed a criminal, civil, or municipal violation before forfeiting their property.152 Out of the ten states, New Mexico provides a model for other states. 147. Blumenson, supra note 55; Office of Inspector Gen., Review of the Depart- ment’s Oversight of Cash Seizure and Forfeiture Activities, U.S. DEP’T OF JUST. 14 (Mar. 2017) https://oig.justice.gov/reports/2017/e1702.pdf [https://perma.cc/6UTL- RQ6T]. 148. CARPENTER II ET AL., supra note 3, at 88. 149. MASS. GEN. LAWS ch. 94C, § 47(d) (West 2006). 150. Id. 151. Id. 152. CARPENTER II ET AL., supra note 3, at 20. \jciprod01productnTTWR5-3TWR310.txt unknown Seq: 17 15-APR-19 14:42 2019] CIVIL ASSET FORFEITURE 1017 In 2015, New Mexico’s legislature passed a civil asset forfeiture re- form bill to ensure that only criminal forfeiture is allowed in the state.153 The legislature passed the bill hoping to protect their citizens’ property rights.154 Under the bill, a person’s property is only subject to forfeiture if the person was arrested for an offense to which forfei- ture applies, the person was convicted of the offense, and the state establishes by clear and convincing evidence that the property was subject to forfeiture.155 The requirement of conviction highlights a big win for the residents of New Mexico, which is the elimination of civil asset forfeiture. The New Mexico legislature worried, as did most pro- ponents of civil asset forfeiture, that passing the bill would severely defund local law enforcement and increase drug crimes.156 Nonethe- less, the New Mexico Governor signed the bill in April of 2015.157 New Mexico’s civil asset reform bill advances and ensures citizens’ property rights by redefining the innocent owner defense. The reform bill puts the burden on the innocent owner to prove that he or she holds the legal right in the property and that he or she had an owner- ship interest in the property at the time of seizure.158 However, the owner’s burden is lifted after proving a legitimate interest in the prop- erty. After proving these elements, the bill requires that the govern- ment immediately return the property to the innocent owner.159 The burden then shifts to the state, who then must prove by clear and con- vincing evidence that the innocent owner had actual knowledge of the underlying crime that gave rise to the forfeiture.160 The reform bill in practical terms is far from resolving the issue. New Mexico residents believed that civil asset forfeiture would cease to exist by July 1, 2015, the forfeiture reform law’s effective date.161 Ashley Martinez, a resident of New Mexico, learned the hard way that law enforcement agencies were simply ignoring the forfeiture reform bill.162 Martinez recounts the story of Albuquerque law en- forcement forfeiting her parents’ 2006 Pontiac stating,163 “[w]hy should they take something that we worked hard for . . . that I had nothing to do with and because they don’t want to go by the laws that 153. H.R. 560, 52nd Leg., 1st Sess. (N.M. 2015). 154. N.M. STAT. ANN. § 31-27-2(A)(2) (West Supp. 2016). 155. Id. § 31-27-4(A). 156. 2015 Legis. Bill Hist. NM H.B. 560. 157. H.R. 560, 52nd Leg., 1st Sess. (N.M. 2015). 158. N.M. STAT. ANN. § 31-27-7.1(B) (West Supp. 2016). 159. Id. 160. Id. § 31-27-7.1(F). 161. New Mexico Civil Forfeiture, INST. FOR JUST. (Nov. 18, 2015), http://ij.org/case/ new-mexico-forfeiture/ [https://perma.cc/JQ3P-AJRR]. 162. Renee Montagne, New Mexico Ended Civil Asset Forfeiture. Why Then Is It Still Happening?, NPR (June 7, 2016, 05:02 AM), https://www.npr.org/2016/06/07/ 481058641/new-mexico-ended-civil-asset-forfeiture-why-then-is-it-still-happening. 163. Id. \jciprod01productnTTWR5-3TWR310.txt unknown Seq: 18 15-APR-19 14:42 1018 TEX. A&M J. PROP. L. [Vol. 5 they’re supposed to be following right now?”164 Like other victims of civil forfeiture, following July 1, 2015, Martinez, was shocked when law enforcement officials proceeded to forfeit property even though doing so had become illegal. Martinez retained legal counsel to re- claim her stolen property, but to no avail, as her family’s case was dismissed for procedural issues.165 Law enforcement continues engag- ing in civil asset forfeiture, even when state laws prohibit civil asset forfeiture. One possible reason for law enforcement’s flagrant disre- gard for the law is likely that the vast amount of funding law enforce- ment agencies around the country stand to lose if civil asset forfeiture becomes stricter or outright banned. VIII. CONCLUSION In order to avoid problems of inequitable or unlawful forfeitures, like the cases of Charles Clarke in Cincinnati or Rochelle Bing in Phil- adelphia, our system of civil asset forfeiture at both the state and fed- eral level must be reformed. The current scope of civil asset forfeiture has expanded beyond its historical scope. The “war on drugs” cam- paign tolerates the abuse of civil asset forfeiture and has directly af- fected citizens’ property rights. Attorney General Sessions defends civil asset forfeiture as helping the victims of crime, but when the DEA can seize property without a warrant, without any accompany- ing illicit narcotics, and possibly without a burden of proof to meet, it seems as if civil asset forfeiture can create more victims then it helps.166 Reform should begin at the state level by having a uniform forfei- ture reporting system. A uniform system will provide better access to necessary information so that the state and local law enforcement agencies can be held accountable. The public would have more faith and trust in their local law enforcement if they could see, through re- formed reporting standards, how civil asset forfeiture is utilized. Im- proving faith and trust in local law enforcement can in turn raise morale and in the future, could lead to greater cooperation with law enforcement, which would aid in the fight against criminal activity. Second, the seizing agency should be required to prove by clear and convincing evidence that the innocent owner had actual knowledge of the underlying crime that gave rise to the forfeiture. Considering the historical root of asset forfeiture and the criminal nature of forfeiture proceedings, the burden of proof should be imposed on the govern- 164. Id. 165. Id. 166. See Attorney General Sessions Issues Policy and Guidelines on Federal Adop- tions of Assets Seized by State or Local Law Enforcement, U.S. DEP’T JUST. (July 19, 2017), https://www.justice.gov/opa/pr/attorney-general-sessions-issues-policy-and- guidelines-federal-adoptions-assets-seized-state [https://perma.cc/KWK4-VTGV]. \jciprod01productnTTWR5-3TWR310.txt unknown Seq: 19 15-APR-19 14:42 2019] CIVIL ASSET FORFEITURE 1019 ment and applied uniformly amongst all the states.167 Furthermore, the criminal nature of the proceeding should raise the general pre- sumption of innocent until proven guilty. However, forfeiture laws have found a way to circumvent that presumption by hiding behind the guise of a civil proceeding. Hiding behind the guise of civil pro- ceedings remains problematic because “even people who had nothing to do with an alleged crime can lose their property through civil forfei- ture unless they can prove their innocence – flipping the American legal tradition of innocent until proven guilty on its head.”168 Requiring the state to prove that the innocent owner had actual knowledge of the underlying crime can become burdensome. How- ever, that burden of knowledge can prevent law enforcement officials from using forfeiture as a means to an end. More importantly, it safe- guards society’s civil liberties. This heightened standard could help en- sure that law enforcement officials are seizing property that they truly believe is connected to criminal activity. This reform will not eliminate civil asset forfeiture because the seiz- ing agency can still forfeit the property without achieving criminal charges or convictions. However, this will prevent costly litigation and hardship for individuals who were unaware of their property’s crimi- nal use by others. Placing this burden on the seizing agency will ensure that law enforcement agencies are conducting their forfeiture pro- ceedings in both ethical and legal ways. This minor burden shift to the government greatly enhances citizens’ property rights. Thus, the slightest state and federal reforms would offer citizens the protection they deserve at a cost that the government can bear. 167. See United States v. The Brig Burdett, 34 U.S. 682 (1835); Leonard v. Texas, 137 S. Ct. 847, 849 (2017). 168. CARPENTER II ET AL., supra note 3, at 8. \jciprod01productnTTWR5-3TWR310.txt unknown Seq: 20 15-APR-19 14:42 Guilty Until Proven Innocent: Rethinking Civil Asset Forfeiture and the Innocent Owner Defense Recommended Citation untitled