(PDF) Insincere Evidence

1115

VIRGINIA LAW REVIEW

VOLUME 105 OCTOBER 2019 NUMBER 6

ARTICLE

INSINCERE EVIDENCE

Michael D. Gilbert* and Sean P. Sullivan**

Proving a violation of law is costly. Because of the cost, minor

violations of law often go unproven and thus unpunished. To illustrate,

almost everyone drives a little faster than the speed limit, but rarely are

we ticketed for doing so. Failure to enforce the law against minor

infractions is justifiable from a cost-benefit perspective. The cost of

proving a minor violation—for example, that a driver broke the speed

limit by one mile per hour—outstrips the benefit. But systemic non-

enforcement has the downside of underdeterrence. People drive a little

too fast, pollute a little too much, and so on.

This Article explores how insincere rules, meaning rules that misstate

lawmakers’ preferences, might reduce proof costs and improve

enforcement. To demonstrate the argument, suppose lawmakers want

drivers to travel no more than 55 mph. A sincere speed limit of 55 mph

may cause drivers to go 65 mph, while an insincere speed limit of 45

mph may cause drivers to drop down to, say, 60 mph—closer to

lawmakers’ ideal. Insincere rules work by creating insincere evidence.

* Professor of Law, University of Virginia.

** Associate Professor of Law, University of Iowa. For helpful comments, we thank Matt

Adler, Omri Ben-Shahar, Darryl Brown, Tony Casey, Frank Easterbrook, Ofer Eldar, Lee

Fennell, Kim Ferzan, Michael Frakes, Paul Gowder, William Hubbard, Rich Hynes, Ben

Johnson, Richard McAdams, Robert Miller, Greg Mitchell, Mark Osiel, Todd Pettys, J.J.

Prescott, Fred Schauer, Bobbie Spellman, conference participants at the 2018 ComplianceNet

meeting, the 2019 American Law and Economics Association meeting, and the 2019 Political

Economy and Public Law Conference, and finally workshop participants at Duke University,

the University of Chicago, the University of Michigan, and the University of Iowa.

1116 Virginia Law Review [Vol. 105:1115

In the driving example, the reduction in proof costs achieved by an

insincere rule is akin to adding an artificial 10 mph to the reading of

every radar gun.

The logic of insincere evidence is not confined to speed limits. The

conditions necessary for insincerity to work pervade the legal system.

We distinguish insincere evidence from familiar concepts like over-

inclusive rules, prophylactic rules, and proxy crimes. We connect

insincere evidence to burdens of persuasion, showing how it can offset

the effects of higher burdens. Finally, we consider the normative

implications of insincere evidence for trials, truth, and law

enforcement.

I. INTRODUCTION ................................................................................ 1116

II. LAW IN BOOKS, LAW IN ACTION ..................................................... 1122

III. ENFORCEMENT AND INSINCERITY .................................................. 1126

A. The Costs of Proof .................................................................. 1127

B. The Benefits of Insincerity ...................................................... 1136

C. Interpretation and Implications .............................................. 1138

D. Generalizations and Applications .......................................... 1141

E. The Limits of Insincere Evidence ............................................ 1145

F. Summary ................................................................................. 1151

IV. DISTINGUISHING INSINCERE EVIDENCE ......................................... 1152

A. Gilbert on Insincere Rules ...................................................... 1152

B. Conduct and Decision Rules ................................................... 1153

C. Over-Inclusive Rules............................................................... 1155

D. Prophylactic Rules ................................................................. 1156

E. Pathological Laws .................................................................. 1158

F. Proxy Crimes .......................................................................... 1161

V. NORMATIVE IMPLICATIONS: COSTS, TRUTH, AND TRIALS ............... 1162

A. Are There Better Ways to Reduce Costs? ............................... 1163

B. Is This an Affront to Truth? .................................................... 1164

C. Is Cost Reduction Enough? .................................................... 1168

VI. CONCLUSION ................................................................................. 1169

TECHNICAL APPENDIX ......................................................................... 1172

I. INTRODUCTION

Law aims to deter bad behaviors, but the mere enactment of a statute

or regulation is not enough. Laws must be enforced to have effect. Alas,

enforcement is expensive. Sometimes enforcement is so expensive that

2019] Insincere Evidence 1117

society would be better off without it. A modern controversy illustrates.

Some states want transgender people to use the bathroom that matches

their biological sex.

1

But the cost of enforcing such a law—inspectors in

every stall?

2

—would exceed any benefit (or purported benefit) that the

law would produce.

The bathroom laws are controversial, but laws on other topics like

corruption, pollution, theft, and discrimination are not. In a perfect world,

these laws would fully deter the targeted behavior. Enforcement costs

frustrate that ideal. Can we lower enforcement costs? Scholars and

policymakers have spent decades on this question. They have explored

solutions like ankle bracelets, red light cameras, extreme fines, and other

incentives.

3

This Article analyzes a different method for lowering enforcement

costs, one that has received little attention but that we believe may be

widespread in practice: the use of insincere rules to generate insincere

evidence.

4

To explain, a sincere rule mandates the rule-maker’s preferred

behavior while an insincere rule mandates something else. If the state

wants drivers to stay at or below 55 miles per hour, a sincere rule would

set the speed limit at 55 mph, while an insincere rule might set the speed

limit at, say, 45 mph.

1

See, e.g., Eric Lichtblau & Richard Fausset, U.S. Warns North Carolina that Transgender

Bill Violates Civil Rights Laws, N.Y. Times (May 4, 2016), https://www.nytimes.com/-

2016/05/05/us/north-carolina-transgender-bathroom-bill.html [https://perma.cc/6TCK-8N3-

H] (describing the law in North Carolina).

2

Samantha Michaels, We Asked Cops How They Plan to Enforce North Carolina’s Bath-

room Law, Mother Jones (Apr. 7, 2016), https://www.motherjones.com/politics/2016/04/-

north-carolina-lgbt-bathrooms-hb2-enforcement/ [https://perma.cc/3JPM-9BR4] (describing

the challenges of enforcement and quoting a police spokesperson: “We’re not checking birth

certificates. We just don’t have the police power to be able to do that in bathrooms.”).

3

See generally Gary S. Becker, Crime and Punishment: An Economic Approach, 76 J. Pol.

Econ. 169, 174 (1968) (“It would be cheaper to achieve any given level of activity . . . the

more highly developed the state of the arts, as determined by technologies like fingerprinting,

wire-tapping, computer control, and lie-detecting.”); id. at 190–99 (providing a germinal

analysis of fines); James Byrne & Gary Marx, Technological Innovations in Crime Prevention

and Policing: A Review of the Research on Implementation and Impact, 20 Cahiers

Politiestudies 17 (2011) (reviewing technological innovations in law enforcement); Bruno S.

Frey, Punishment – and Beyond, 5 Contemp. Econ. 90 (2011) (discussing the limits of

punishment as a tool for crime prevention and considering alternatives).

4

One of us introduced insincere rules in a separate paper. See Michael D. Gilbert, Insincere

Rules, 101 Va. L. Rev. 2185 (2015). The current Article explores a different mechanism

through which insincere rules can improve compliance. We distinguish the two projects in

Section IV.A, infra.

1118 Virginia Law Review [Vol. 105:1115

Insincere rules lower enforcement costs by making violations of law

easier to prove. Enforcing laws requires many steps. At a minimum, the

state must (1) observe a bad act, (2) identify and apprehend the

perpetrator, (3) prove that the perpetrator committed the bad act, and (4)

enjoin or punish the perpetrator. Many papers focus on the first, second,

and fourth steps in this chain.

5

We focus on the third step. Proving the

commission of a wrong is not easy. The state must gather evidence,

prepare witnesses and exhibits, try the case, and ultimately convince a

judge or jury of the wrong beyond the appropriate burden of persuasion,

to say nothing of appeals and retrials. Every step of this process involves

costs. Insincere rules make the process easier—make enforcement

cheaper—by turning hard cases into easy cases.

6

Below we develop the

argument with rigor. Here we illustrate with an example.

Suppose the state wants drivers to go 55 mph. The state sets the speed

limit at 55 mph, meaning the rule is sincere. If a driver gets caught going

85, the case is easy. Yes, the state burns resources monitoring the road,

stopping the driver, and administering a fine.

7

But on the question of guilt

the case is open-and-shut: he sped. Now what if a driver gets caught going

56 mph? This case is hard. The driver might fight the ticket. He might

argue that the officer made a mistake by failing to hold the radar still. He

might argue that the radar measures with error on sharp curves like the

one he traversed. Even if the driver is lying, proving it takes work. The

officer would likely need to testify, and the state may have to present

evidence on the trustworthiness of the radar. The state might lose despite

this effort. Deterring someone from going 56 mph is not worth the trouble.

By the same logic, deterring someone from going up to 65 mph may not

5

See, e.g., A. Mitchell Polinsky & Steven Shavell, The Theory of Public Enforcement of

Law, in 1 Handbook of Law & Economics 403, 405 (A. Mitchell Polinsky & Steven Shavell

eds., 2007) (“The theoretical core of our analysis addresses the following basic questions:

Should the form of the sanction imposed on a liable party be a fine, an imprisonment term, or

a combination of the two? Should the rule of liability be strict or fault-based? If violators are

caught only with a probability, how should the level of the sanction be adjusted? How much

of society’s resources should be devoted to apprehending violators?”).

6

Cf. Charles L. Barzun, Rules of Weight, 83 Notre Dame L. Rev. 1957 (2008) (proposing

a distinct but conceptually related mechanism in which legal rules might be used to impose

specific fact-finding weights on various items of evidence).

7

In the economics literature on enforcement, fines are usually treated as costless transfers

of wealth from perpetrators to the state. In reality, fines must have some social cost, however

small, because they must be administered. See Polinsky & Shavell, supra note 5, at 430–31

(analyzing enforcement when fines are socially costly); see also Richard A. Posner, An

Economic Theory of the Criminal Law, 85 Colum. L. Rev. 1193, 1207–08 (1985) (noting

other frictions that complicate the calibration and administration of penal fines).

2019] Insincere Evidence 1119

be worth the trouble. Savvy drivers realize this and see that they can drive

as fast as 65 mph with impunity.

Now rerun the example with one change: an insincere rule. Instead of

setting the speed limit at 55 mph, suppose that the state makes it 45 mph.

The case of the driver going 85 mph remains easy. But the case of the

driver going 65 mph is now easy as well. Even after accounting for

potential errors by the officer and radar, judges and jurors probably would

conclude that a driver clocked at 65 mph in a 45-mph zone had broken

the law. Foreseeing that proof will be easy, the officer will ticket the

driver going 65 mph. And foreseeing the ticket, the driver will slow down.

He may slow to 60 mph—closer to the speed that lawmakers wanted in

the first place. By making hard cases easy, the insincere rule improves

behavior.

Our title captures the mechanism by which insincerity lowers the cost

of proof. An insincere rule can be viewed as giving the state a block of

“free” or “artificial” evidence to help prove a violation of law. Consider

the driving example. From a proof-cost perspective, adopting an

insincerely low speed limit is like handing the officer a rigged radar gun.

It is easy to see that a radar gun which secretly adds an extra 10 mph onto

every measurement would make it easier to prove a speeding violation.

Every arguable 5-mph infraction would suddenly appear to be an

unambiguous 15-mph infraction. Insincere rules have a similar effect.

Reducing the speed limit from the sincere 55 mph level to the insincere

45 mph level does not change the speed that a 60-mph driver was actually

going. It simply turns what had been a 5-mph infraction into what is now

a 15-mph infraction.

Insincere rules offer a mechanism by which lawmakers—legislators,

agencies, judges—can lower enforcement costs and induce better

behavior in regulated parties. The mechanism is very general.

8

Insincere

rules do not require deception; even if everyone knows a rule is insincere,

it can still lower proof costs. Insincere rules do not require irrationality

among regulated parties; they work even when everyone is strongly

rational. Insincere rules do not require objective evidentiary uncertainty;

they can improve behavior in the face of perceived uncertainty as well.

Insincere rules can benefit the state whether its objective is deterrence,

corrective justice, or any other end connected to the behavior of regulated

parties.

8

See infra Section III.D.

1120 Virginia Law Review [Vol. 105:1115

Here is a sampling of the kinds of laws to which our analysis applies:

blood alcohol content while operating a vehicle; emissions of carbon,

lead, or other pollutants; the length, weight, and number of fish caught in

protected waters; noise limits at airports and nightclubs; the amounts (e.g.,

milligrams) of particular ingredients allowed in each dose of medicine;

distances between guns and schools, protesters and abortion clinics,

smokers and public buildings. These examples canvass precise laws: for

example, laws prohibiting the possession of more than “[o]ne ounce of

useable marijuana.”

9

But insincere rules can improve enforcement of

vague laws too. For example, laws requiring pharmaceutical companies

to “[b]e maintained in a clean and orderly condition,”

10

or prohibiting

mergers “the effect of [which] may be to substantially lessen

competition . . . .”

11

Given the benefits and general applicability of insincere rules, we

suspect that this mechanism, while new to scholars, is not new to

lawmakers. Scholars have long observed gaps between the law in books

and the law in action.

12

Laws mandate one thing while regulated parties

do another. Scholars bemoan these gaps,

13

but perhaps their anguish is

misplaced. Perhaps some of these gaps reflect strategic and beneficial

uses of insincere rules, with lawmakers intentionally getting the law in

books “wrong” in order to get the law in action “right.”

Beyond identifying the potential benefits of insincere evidence, our

analysis uncovers an overlooked connection between the substance of law

and its processes. Speeding tickets are typically civil offenses. Thus, the

state’s burden of persuasion would usually be preponderance of the

evidence. What would happen if the state’s burden were to prove guilt

9

Wash. Rev. Code § 69.50.360(3)(a) (2015).

10

21 C.F.R. § 205.50(a)(4) (2019).

11

Clayton Antitrust Act, ch. 323, § 7, 38 Stat. 730, 731–32 (1914) (current version at 15

U.S.C. § 18 (2012)).

12

Roscoe Pound, Law in Books and Law in Action, 44 Am. L. Rev. 12, 34 (1910). For

perspective on the large literature on “gap studies,” see Jon B. Gould & Scott Barclay, Mind

the Gap: The Place of Gap Studies in Sociolegal Scholarship, 8 Ann. Rev. L. & Soc. Sci. 323

(2012).

13

For example, Max Rheinstein has called pervasive differences between law and practice

in family law “inan[e].” Max Rheinstein, Marriage Stability, Divorce, and the Law 351–53

(1972). Richard Abel states that early gap studies “were frequently directed by the belief that

the gap, once revealed, could and should be eliminated.” Richard L. Abel, Law Books and

Books About Law, 26 Stan. L. Rev. 175, 188 (1973) (reviewing Max Rheinstein, Marriage

Stability, Divorce, and the Law (1972)). Roscoe Pound blamed gaps on “our machinery of

justice” that is “too slow, too cumbersome and too expensive” and argued that lawyers must

“make the law in action conform to the law in the books.” Pound, supra note 12, at 35–36.

2019] Insincere Evidence 1121

beyond a reasonable doubt? Given a speed limit of 55 mph, the driver

clocked at 85 mph would probably still present an easy case and the driver

clocked at 56 mph would still present a hard case. But the higher burden

could be an obstacle to the state securing a verdict for many speeds in

between. Lowering the speed limit—adopting an insincere rule—would

make some of these cases easier. Under the preponderance of the evidence

standard, a speed limit of 50 mph might be enough to make a driver

clocked at 60 mph an easy case. For proof beyond a reasonable doubt, it

might take more insincerity—a speed limit of 45 mph—to make the driver

clocked at 60 mph an easy case.

To generalize, insincere evidence can offset the cost of meeting higher

burdens of persuasion. The more stringent the burden, the more insincere

the law must be. This suggests that rule-makers may adopt the strictest

rules in criminal law—which is exactly what some scholars perceive. In

a classic paper, Professor Bill Stuntz argued that criminal liability has

broadened, leading to a “world in which the law on the books makes

everyone a felon.”

14

Our work relates to Stuntz’s. We generalize,

improve, and challenge his analysis.

15

As this discussion suggests, insincere evidence raises difficult

normative questions. It can unwind judicial protections, especially in

criminal law. It appears to require dishonesty by the state. (Actually,

lawmakers do not have to lie or deceive for insincerity to work this way,

but a wedge does open between laws and preferred behavior.) Insincere

evidence also raises practical problems. It may, for example,

unintentionally distort the behavior of scrupulously law-abiding citizens

and unsophisticated enforcement agents. For reasons like these, we do not

believe that insincere evidence strategies would work or should be

adopted in every setting. Our objective is not to champion or vilify

insincere evidence but to study and explain it.

The remainder of this Article proceeds as follows. Part I provides

background on enforcement and other literatures related to our work. Part

II develops our theory. Part III relates our theory to other concepts such

as over-inclusive rules, prophylactic rules, proxy crimes, and acoustic

separation. Part IV considers normative implications, placing our theory

in the context of evidence law, enforcement, deterrence, and truth. We

14

William J. Stuntz, The Pathological Politics of Criminal Law, 100 Mich. L. Rev. 505, 511

(2001).

15

See infra Section IV.E.

1122 Virginia Law Review [Vol. 105:1115

conclude with brief remarks on implications for other areas of law. A

technical appendix appears at the end.

II. LAW IN BOOKS, LAW IN ACTION

This Part briefly reviews the literature on enforcement. The main point

is that one can improve enforcement by raising penalties (the focus of

much scholarship) or lowering proof costs (the focus of relatively little

scholarship). One can lower proof costs by loosening the burden of

persuasion. Or—this is where our intervention begins—one can lower

proof costs by tightening rules. Readers familiar with the enforcement

literature can safely skip ahead.

Cyclists ride without helmets. Stores sell alcohol to minors. Truckers

drive longer shifts than legally allowed. City ordinances prohibit smoking

at bus stops, but smokers linger nearby. A quick glance at just about any

slice of life reveals a pervasive and important phenomenon: law mandates

one behavior but yields another. Sometimes the slippage is minor, as in

our bus stop example. Other times the stakes are high. Twenty years after

the Clean Water Act, 10,000 dischargers still lacked permits.

16

Around

the world, governments fail systematically to protect the rights enshrined

in their national constitutions.

17

The legal scholar Roscoe Pound captured

this regularity with a memorable phrase: there is a gap between the “law

in books” and the “law in action.”

18

16

See Daniel A. Farber, Taking Slippage Seriously: Noncompliance and Creative

Compliance in Environmental Law, 23 Harv. Envtl. L. Rev. 297, 304 (1999) (citing Robert

W. Adler et al., The Clean Water Act: 20 Years Later 151 (1993)).

17

See generally David S. Law & Mila Versteeg, Sham Constitutions, 101 Calif. L. Rev.

863, 863 (2013) (documenting the “global prevalence and severity of constitutional

noncompliance over the last three decades”). Perhaps constitutions get ignored because

political actors do not pay a price for violating law. See Frederick Schauer, The Political Risks

(If Any) of Breaking the Law, 4 J. Legal Analysis 83 (2012).

18

Pound, supra note 12; see also Eugen Ehrlich, Fundamental Principles of the Sociology

of Law, at xxvii (Walter L. Moll trans., 1936) (distinguishing legal propositions from “living”

law); Jason S. Johnston, Bayesian Fact-Finding and Efficiency: Toward an Economic Theory

of Liability Under Uncertainty, 61 S. Cal. L. Rev. 137, 139 (1987) (distinguishing “nominal”

and “administered” rules); Karl N. Llewellyn, A Realistic Jurisprudence—The Next Step, 30

Colum. L. Rev. 431, 439 n.9 (1930) (distinguishing “paper” rules from real or “working”

rules).

2019] Insincere Evidence 1123

Why do gaps arise? Prior scholarship has answers.

19

Consider the

Holmesian “bad man” deliberating over whether to break the law.

20

This

person is not deterred by any abstract force of nature or internal

commitment to justice. Instead, he weighs the expected benefits of

breaking the law (profits, revenge, the pleasure of listening to illegally-

downloaded music) against the expected costs (potential fines,

imprisonment, social opprobrium). If the benefits exceed the costs, the

bad man breaks the law. The law in books is whatever the statutes or

regulations say. The law in action is what the bad man is actually deterred

from doing.

21

The state can close the gap between the law in books and the law in

action by increasing the expected cost of breaking the law.

22

It has many

levers for doing so. It can raise the probability that a transgressor will be

caught in the act.

23

More cops usually means less crime. Alternatively,

the state can adopt heavier sanctions.

24

Want to reduce littering in parks?

Replace small fines with imprisonment.

19

For overviews of the literature on the economic theory of law enforcement, see Nuno

Garoupa, The Theory of Optimal Law Enforcement, 11 J. Econ. Survs. 267, 267–68 (1997);

A. Mitchell Polinsky & Steven Shavell, The Economic Theory of Public Enforcement of Law,

38 J. Econ. Lit. 45 (2000) [hereinafter Polinsky & Shavell, Economic Theory of Public

Enforcement of Law]; and A. Mitchell Polinsky & Steven Shavell, Public Enforcement of

Law, in 5 The New Palgrave: Dictionary of Economics 38 (Steven N. Durlauf & Lawrence E.

Blume eds., 2d ed. 2008). This literature presents the law and economics perspective on the

“gap.” For a philosophical perspective, see, for example, Heidi M. Hurd, Moral Combat 185–

202 (1999), and Larry Alexander, The Gap, 14 Harv. J.L. & Pub. Pol’y 695 (1991).

20

See Oliver Wendell Holmes, The Path of the Law, 10 Harv. L. Rev. 457, 459 (1897)

(“You can see very plainly that a bad man has as much reason as a good one for wishing to

avoid an encounter with the public force . . . . A man who cares nothing for an ethical rule

which is believed and practised by his neighbors is likely nevertheless to care a good deal to

avoid being made to pay money, and will want to keep out of jail if he can.”).

21

Id. (“If you want to know the law and nothing else, you must look at it as a bad man . . . .”).

22

See Becker, supra note 3, at 179–85.

23

See id. at 174.

24

See id. at 179–80. Relatedly, the government can affect perceptions of the sanction. See

Bert I. Huang, Shallow Signals, 126 Harv. L. Rev. 2227, 2253–57 (2013) (analyzing the failure

to publicize when one actor has permission to engage in behavior that is forbidden for other

actors). The state can also effectively raise sanctions by increasing the costs to defendants of

defending themselves. Cf. Maximilian A. Bulinski & J.J. Prescott, Online Case Resolution

Systems: Enhancing Access, Fairness, Accuracy, and Efficiency, 21 Mich. J. Race & L. 205,

208–10 (2016) (proposing solutions to the high cost of accessing the justice system); J.J.

Prescott, Improving Access to Justice in State Courts with Platform Technology, 70 Vand. L.

Rev. 1993 (2017) (discussing the sources of costs in the criminal justice system and evaluating

online platform technology as a structural reform to remedy the problem).

1124 Virginia Law Review [Vol. 105:1115

Professor Gary Becker developed the latter idea, arguing that a

combination of fewer police and severe sanctions will efficiently deter

bad men from breaking the law.

25

To illustrate, suppose that drivers will

stop at red lights only if the expected cost of running the light is $100 or

more. How can the state set the expected cost at $100? It can install a red-

light camera at every intersection, making the probability of getting

caught 100 percent, and set the fine at $100. Or it can secretly install one

red-light camera in every hundred intersections, making the probability

of getting caught one percent, and set the fine at $10,000.

26

The expected

cost of running the light is the same (1% × $10,000 = $100), but the latter

approach economizes on red-light cameras, saving the state money.

27

Becker’s prescription works in theory but faces challenges in practice.

Publicity and morals place an upper bound on sanctions. “Tough on

crime” makes for good politics. “Tough on trivial regulatory infractions”

does not. A street vendor in New York received $2,250 in fines for using

a table that was an inch too tall and two inches too close to a store

entrance.

28

A student was fined $675,000 for illegally downloading 30

songs.

29

Run-of-the-mill traffic violations in California cost $500.

30

Sanctions like these generate negative attention and political pressure,

making it hard for politicians to follow Becker’s severe-sanction advice.

Even if they could, severe sanctions might not deliver on the promise.

Becker’s approach fails when law-breakers exhibit certain cognitive

biases. Some people might treat the very low probability of punishment

25

See Becker, supra note 3, at 183–84.

26

For simplicity, we assume drivers cannot tell which intersections have cameras. Thus,

they assume the probability of getting caught at any given intersection equals one percent.

27

According to a 2001 study, a single red-light camera costs at least $50,000. Robert P.

Maccubbin, Barbara L. Staples & Arthur E. Salwin, Automated Enforcement of Traffic

Signals: A Literature Review, at ES-1 (2001), https://rosap.ntl.bts.gov/view/dot/3822

[https://perma.cc/8XTF-UW6B].

28

Sally Goldenberg, Street Vendor Selling Cellphone Cases Fined 2G Fine for Inches, N.Y.

Post (Oct. 8, 2012, 4:00 AM), http://nypost.com/2012/10/08/street-vendor-selling-cellphone-

cases-fined-2g-fine-for-inches/ [http://perma.cc/E5Q4-4Z45].

29

Denise Lavoie, Joel Tenenbaum Boston University Student Download Fine: Court Won’t

Reduce $675,000 Penalty, HuffPost College (May 21, 2012, 11:10 AM),

[http://perma.cc/45RG-8KLB].

30

Editorial, Relief from the High Cost of Traffic Tickets — For Some Californians, at Least,

L.A. Times, (Aug. 11, 2017), https://www.latimes.com/opinion/editorials/la-ed-fairer-traffic-

ticket-20170811-story.html [https://perma.cc/XD7J-9M8N] (“The state’s exorbitant traffic

fines are tough for anyone to pay, so laden as they are with add-on fees that inflate a $100

ticket to nearly $500 in the end.”).

2019] Insincere Evidence 1125

for breaking the law as a zero probability of punishment.

31

The threat of

a sanction, even a severe one, will not deter these people. Others might

heavily discount future punishments relative to current benefits.

32

For

them, today’s pleasure is a greater concern than next month’s pain. Again,

severe penalties will not deter these people, as long as there is a big

enough lag between the benefit of breaking the law and the punishment.

If the state cannot enact severe enough sanctions, or if those sanctions

fail to achieve deterrence, then another option remains: increase the odds

of detection. The state can detect violations more accurately and more

quickly by hiring more enforcement officers (cops, forest rangers, air-

quality regulators), purchasing more surveillance equipment (drones,

listening devices, testing kits), and paying more informants (street

criminals, whistleblowers, foreign agents). These investments increase

the probability of a law breaker getting caught. But this approach faces a

political and moral objection. All of these investments are costly. The

resources may be needed elsewhere. Spending another $1,000 on schools

may do more good for society than spending another $1,000 on red-light

cameras.

Given this tradeoff, slippage against the law in the books is inevitable.

In a sense, it is even desirable. As Professors Mitchell Polinsky and

Steven Shavell put it, “[o]ptimal enforcement tends to be characterized

by some degree of underdeterrence . . . because allowing some

underdeterrence conserves enforcement resources.”

33

In other words,

31

On the underweighting on low probability risks, see, for example, Greg Barron & Ido

Erev, Small Feedback-Based Decisions and Their Limited Correspondence to Description-

Based Decisions, 16 J. Behav. Decision Making 215, 223–24 (2003); Ralph Hertwig, Greg

Barron, Elke U. Weber & Ido Erev, Decisions from Experience and the Effect of Rare Events

in Risky Choice, 15 Psychol. Sci. 534, 537 (2004); and Dov Zohar & Ido Erev, On the

Difficulty of Promoting Workers’ Safety Behaviour: Overcoming the Underweighting of

Routine Risks, 7 Int’l J. Risk Assessment & Mgmt. 122, 125–26 (2007).

32

For critical reviews of models related to present-bias, see Shane Frederick, George

Loewenstein & Ted O’Donoghue, Time Discounting and Time Preference: A Critical Review,

40 J. Econ. Lit. 351 (2002); and Ariel Rubinstein, “Economics and Psychology”? The Case of

Hyperbolic Discounting, 44 Int’l Econ. Rev. 1207 (2003). For recent empirical work on this

topic, see Jess Benhabib, Alberto Bisin & Andrew Schotter, Present-Bias, Quasi-hyperbolic

Discounting, and Fixed Costs, 69 Games & Econ. Behav. 205 (2010).

33

Polinsky & Shavell, Economic Theory of Public Enforcement of Law, supra note 19, at

70. “Optimal” enforcement means social welfare-maximizing enforcement. The dominant

strand in the enforcement literature assumes that the state maximizes social welfare. See, e.g.,

Becker, supra note 3, at 207; Polinsky & Shavell, Economic Theory of Public Enforcement of

Law, supra note 19, at 49. A second line, growing from public choice theory, examines

enforcement when the state is self-interested. See, e.g., David Friedman, Why Not Hang Them

1126 Virginia Law Review [Vol. 105:1115

society is better off with some lawbreaking—but with money for roads

and schools—than with no lawbreaking but no money for other needs.

Of course, if enforcement were costless there would be no gap between

the law in the books and the law in action. Gaps only arise because the

costs of enforcing the law sometimes exceed the benefits. No one contests

that enforcement is costly.

34

But what if enforcement costs could be

reduced? The state could elicit better behavior and still have money for

schools and roads.

Policymakers have long sought to lower enforcement costs. Their ideas

have included technological innovations like red-light cameras, ankle

bracelets, and smoke stack monitors.

35

These innovations help, but

enforcement is still costly. In the following pages we describe a different

strategy with general applicability for reducing enforcement costs. The

strategy we consider does not arise out of technical innovation but instead

out of legal innovation regarding the content of law.

III. ENFORCEMENT AND INSINCERITY

This Part develops our theory. We begin by pinpointing an

underappreciated source of enforcement costs: a violation of law must be

proved before a penalty can be imposed. Like the costs of police, the costs

of proof lead to underenforcement of laws. We show how insincere rules

can reduce proof costs. Lawmakers can achieve better compliance by

misrepresenting the conditions under which a wrong has been committed.

In short, hard cases can be turned into easy cases. We discuss when this

All: The Virtues of Inefficient Punishment, 107 J. Pol. Econ. S259, S262 (1999); Nuno

Garoupa & Daniel Klerman, Optimal Law Enforcement with a Rent-Seeking Government, 4

Am. L. & Econ. Rev. 116, 116–17 (2002). In both cases, enforcement costs can limit

enforcement. Sometimes non-enforcement today can lead to deteriorating behavior tomorrow.

See Omri Ben-Shahar, The Erosion of Rights by Past Breach, 1 Am. L. & Econ. Rev. 190

(1999). This temporal dynamic complicates the cost-benefit analysis, but the same

fundamental tradeoff remains.

34

Cf. Douglass C. North, Nobel Prize Lecture: Economic Performance Through Time (Dec.

9, 1993), https://www.nobelprize.org/prizes/economic-sciences/1993/north/lecture/ [https://-

perma.cc/AJ7S-QNWL] (“Only under the conditions of costless bargaining will the actors

reach the solution that maximizes aggregate income regardless of the institutional

arrangements. When it is costly to transact then institutions matter. And it is costly to

transact.”).

35

See generally Byrne & Marx, supra note 3, at 25–28 (examining a wide range of new

technological innovations with applications in the areas of crime prevention generally and

crime control by police).

2019] Insincere Evidence 1127

mechanism works or fails, and we identify important implications for

burdens of persuasion.

A. The Costs of Proof

The ideas we develop have broad application, as we explain below.

36

To keep things as clear as possible, though, we begin by focusing on the

simple example raised in the Introduction. To repeat the setup, suppose

that lawmakers have decided drivers should travel no faster than 55 mph

on a particular stretch of highway. This speed strikes the optimal balance

of convenience, safety, and other factors. How might lawmakers go about

enacting and enforcing the speed limit?

A natural option would be to enact a 55 mph speed limit and deploy

cops to watch for speeding vehicles. Of course, there are many other

details to consider. How many police should be assigned to monitor this

highway? How severe should the punishment for speeding be? Others

have explored these questions and we will not discuss them.

37

We assume

that the state has already optimized these levers as best it can.

38

Our focus is on the proof process. Even when a violation of the speed

limit is committed and detected, it is not certain that a penalty will be

imposed. The driver can always dispute a ticket. And if the driver disputes

the ticket, the state cannot impose a penalty without first proving in court

that a violation has occurred.

Much goes into proving a violation of law, even in this simple setting.

Prosecutors, judges, clerks, and possibly jurors need to devote time and

resources to the case. The ticketing officer needs to take the stand to

testify. Time that these actors spend on traffic cases is time not spent on

other, more serious cases. Assuming the officer relied on something more

than subjective observation, details will need to be provided about how

the driver’s speed was measured. Was it measured by the officer’s

speedometer while pacing the driver down a length of road?

39

By a hand-

36

See infra Section III.D.

37

See supra note 19 (citing several good overviews of the existing literature).

38

For completeness, we could also assume that the state will re-optimize these levels to

account for any changes to the setup we introduce in the following pages. The point is that we

are abstracting from these standard enforcement considerations.

39

See David Brown, Fight Your Speeding Ticket: Determining Your Speed, Nolo, https://-

www.nolo.com/legal-encyclopedia/free-books/beat-ticket-book/chapter6-1.html [https://per-

ma.cc/6PT9-CDW4] (last visited May 14, 2019) (describing different forms of speed detection

and commenting on their comparative accuracies).

1128 Virginia Law Review [Vol. 105:1115

held radar or laser gun?

40

By aerial surveillance?

41

None of these

detection technologies is perfect.

42

Absent special laws to the contrary,

43

the state would need to prove the accuracy of whatever method the officer

used to estimate the driver’s speed.

44

In principle, this would include not

only the accuracy of the detection method (use of a radar gun) but also

the accuracy of any techniques used to ensure the accuracy of the

detection method (calibration of a radar gun by speedometer or tuning

fork), which themselves have different rates of error.

45

Even with every

piece of foundational testimony in place, uncertainty will remain.

46

Now consider the driver’s defense. The driver could take the stand to

testify—perhaps compellingly—that he was not speeding. The driver

might call a disinterested passenger to testify that the driver was obeying

the speed limit at the time of the stop. The driver might also point to

factors that can throw off the reading of a radar gun: detection angles,

40

Id.

41

Id.

42

See, e.g., Nat’l Highway Traffic Safety Admin., LIDAR Speed-Measuring Device Per-

formance Specifications, at 7, 10, 12, 17–28, (Mar. 2013), https://www.nhtsa.gov/sites/-

nhtsa.dot.gov/files/documents/809811-lidarspeedmeasuringdevice.pdf [https://perma.cc/2Z-

P8-W63D] (providing test procedures and tolerances for how much inaccuracy a LIDAR unit

may exhibit in different operating conditions).

43

E.g., Va. Code Ann. § 46.2-882 (2017) (providing that a radar gun reading “shall be

accepted as prima facie evidence” of the speed a vehicle was driving, and providing that “a

certificate, or a true copy thereof, showing the calibration or accuracy of . . . any tuning fork

employed in calibrating or testing the radar . . . and when and by whom the calibration was

made, shall be admissible [to address questions about calibration or accuracy]” as long as such

calibration or testing was conducted within the last six months). But cf. State v. Bonar, 319

N.E.2d 388, 389 (Ohio Ct. App. 1973) (finding that the State’s failure to present evidence of

a radar gun’s accuracy, while not rendering the evidence inadmissible, does entitle the

defendant to a directed verdict when the State’s only evidence was the radar gun reading).

44

See, e.g., Fed. R. Evid. 901 (requiring in federal courts, as in most state courts, that the

proponent of evidence must show that the evidence is what it purports to be, including, for

example, the basis for believing that a process or system—i.e., radar gun—produces an

accurate result); see also Fed. R. Evid. 104(b) (providing that when evidence is relevant only

upon the support of a foundational fact, evidence must be introduced sufficient to support a

finding that the foundational fact exists).

45

See, e.g., John Jendzurski & Nicholas G. Paulter, Calibration of Speed Enforcement

Down-the-Road Radars, 114 J. Res. Nat’l Inst. Standards & Tech. 137 (2009) (comparing the

accuracy of different techniques for calibrating radar guns).

46

See, e.g., Mobile Speed Cameras, BBC Inside Out — South West (Feb. 28, 2005),

http://www.bbc.co.uk/insideout/southwest/series7/speed-cameras.shtml [https://perma.cc/Q-

R29-R3WA] (illustrating some potential errors in speed detection and describing case studies

of individuals who successfully opposed tickets based on speed gun readings).

2019] Insincere Evidence 1129

intervening vehicles, and reflective surfaces;

47

wind or weather

conditions;

48

the officer’s shaky hands or inexperience.

49

If the proof process in even this simple example sounds like an ordeal,

imagine how much more involved it would be in other settings. In

antitrust practice, for example, the state’s proof that a merger could have

anticompetitive effects will often involve hundreds of hours of

depositions and live testimony, and thousands of pages of reports and

exhibits. In 1934, the trial of monopolization charges against Alcoa took

two years and three months from start to finish.

50

Securities fraud,

environmental regulations, tax fraud, banking regulations, discrimination

claims—in all these areas and many others, the costs of proof can be

substantial.

51

One takeaway from this discussion is that the state’s detection of a

violation of law does not end the costs it faces in seeking to enforce the

law. Proving a violation is as costly as the proof process makes it.

A second takeaway is that even the state’s decision to shoulder the cost

of proving a violation does not guarantee that a violator will actually be

punished. The judge or jury could always be unpersuaded by the state’s

evidence, in which case the lawbreaker walks away scot-free.

52

47

See, e.g. Myatt v. Commonwealth, 397 S.E.2d 275, 277 (Va. Ct. App. 1990) (discussing

expert testimony about “[t]he reflective qualities of the appellant's motor vehicle, the presence

of the tractor-trailer truck behind the appellant’s vehicle, the angle of the radar unit with

respect to the highway, and its height above the highway” as factors properly challenging a

radar unit’s accuracy on a particular occasion).

48

See, e.g., Brown, supra note 39 (discussing how wind and weather conditions might cause

spurious radar readings).

49

Cf. Decatur Electronics, GHD & SCOUT Handheld Directional Radar User’s Manual

§ 7.4 (2010), http://www.decatureurope.com/uploads/Brochures/GHD-SCOUT%20User%-

20Manual%206-10-10-E.pdf [https://perma.cc/B55Z-C2VF] (“Fan interference is the most

common form of interference that you are likely to experience [when trying to measure driving

speeds]. It is caused when the radar measures the speed of the vehicle blower fan.”); id. § 12

(“Q. Will my radar work while my vehicle is moving? A. No, the GHD and SCOUT radar

guns are a [sic] stationary only models, so your vehicle should be parked. You need to hold

the radar steady while operating it.”).

50

See Spencer Weber Waller, The Story of Alcoa: The Enduring Questions of Market

Power, Conduct, and Remedy in Monopolization Cases, in Antitrust Stories 121, 127–28

(Eleanor M. Fox & Daniel A. Crane eds., 2007).

51

See, e.g., Stagi v. Nat’l R.R. Passenger Corp., 880 F. Supp. 2d 564, 570 (E.D. Pa. 2012)

(class counsel spent nine years litigating Title VII class action); Carpenters Health & Welfare

Fund v. Coca-Cola Co., 587 F. Supp. 2d 1266, 1271 (N.D. Ga. 2008) (class counsel spent

eight years litigating securities fraud class action).

52

We ignore the possibility that the cost to the defendant of trial, even if he wins, may itself

act as a form of punishment and deterrent. Accounting for this possibility appears to

complicate matters without fundamentally changing our results.

1130 Virginia Law Review [Vol. 105:1115

Not all cases are equally difficult though. All else equal, the probability

of an innocent verdict would seem to shrink as the extent of the violation

grows.

53

This is easily illustrated. Suppose that a hypothetical driver has

been ticketed for going 90 mph in a 55-mph zone. The officer testifies

that the driver appeared to be going about this speed, and that the radar

gun confirmed it. Can the driver really undermine this evidence? Setting

aside the possibility of pure fabrication by the officer, uncertainty around

the measurement seems unlikely to win the day. The radar gun may be

imperfect, but not that imperfect. Proving that the driver was speeding

seems almost certain.

Now suppose the driver has been ticketed for going 56 mph in the 55-

mph zone. Even small uncertainties loom large. The state can call all the

experts it wants in trying to bolster the 56-mph reading. When considering

the many ways that the radar could possibly have been a mere 1 mph high,

the fact-finder could remain unpersuaded. The odds of the state winning

at trial now seem low—maybe no better than a coin flip.

This thought experiment shows that even when a violation of law is

detected, there may be good reasons for the state to decline to enforce. At

a minimum, it should not attempt to prove a violation unless the expected

benefits of doing so exceed the costs.

We can capture this with a simple equation. When the state chooses to

enforce the law, it pays a cost of C. This cost could be fixed, or it could

vary, increasing for harder cases (the state needs more evidence) and

decreasing for easier cases. The benefit of enforcement is B, which is not

fixed. B grows with the magnitude of the infraction. Intuitively, the state

gets more benefit from penalizing a 90-mph speeder than it does from

penalizing a 56-mph speeder. Proof is never certain; there is always a

chance that the state’s case falls apart. Thus, the benefit must be

discounted by the probability of conviction, p, which varies with the

extent of the violation, as discussed above. The state would not rationally

seek to enforce the law unless pB > C. Rearranging terms for clarity, the

state would not enforce unless B > C/p.

53

Holding all else equal means, among other things, holding fixed the amount of resources

that the defendant is willing to spend on fighting the charge or claim. We accept that—

depending on punishments and collateral implications—defendants might often feel more

compelled to fight larger alleged infractions. Our point is to identify a marginal property of

the system that applies regardless of what defendants choose to do with their resources.

2019] Insincere Evidence 1131

Figure 1 illustrates. The benefit line captures B, the social gain from

successfully penalizing a speeding driver at a violation of a certain size.

54

Punishing a driver going 85 yields a large benefit; punishing a driver

going 65 yields a small benefit; punishing a driver going 55 or slower

yields no benefit at all.

55

The cost line captures C/p. It illustrates the

opportunity cost of trying to prove a given violation of law. For large

violations, like speeds of 90 mph and above, the probability of victory is

almost certain (p approaches 1), so the line simply reflects C, the

resources that go into putting on witnesses and presenting evidence at

trial. For smaller violations, like speeds of 56 mph, the cost line reflects

these same expenses scaled up to account for the possibility that the state

could lose (C/p grows as p approaches 0). The intuition is that as the

probability of conviction shrinks, the social resources expended on

prosecution are less likely to yield a benefit.

56

Resources wasted on a loss

could have been spent elsewhere. Thus, the opportunity cost of attempting

to punish a lawbreaker grows as the difficulty of the case rises.

54

Nothing turns on this being the only value that the state perceives from enforcing the law.

Fines, for example, may be an important source of revenue for the state. And the public at

large may derive some form of deontological value from seeing popular laws enforced. These

additional benefits could be factored into the following analysis with few real changes in result

or prediction. Note also that while the benefit line in Figure 1 originates at 55 mph—implying

that deterring drivers from going 54 mph provides no benefit to society—nothing in our

analysis is particularly sensitive to this assumption. Finally, we assign the benefit line a

specific shape in Figure 1, starting at zero at the legal limit. Nothing turns on the shape, as

long as it slopes upward.

55

Consistent with this description, the benefit line slopes upward. One could characterize

the benefits of enforcement in different ways, and the line could take different shapes.

However, this is not critical to our results. As long as the benefit line is not vertical, insincere

rules can improve enforcement.

56

Recall that we assume that the state gets no deterrence benefit from a loss at trial. See

supra note 53. This is not critical to our results.

1132 Virginia Law Review [Vol. 105:1115

As Figure 1 shows, the state should not even try to prove small

violations of the speed limit. The cost of attempting to enforce the law

against a driver clocked at 60 mph exceeds the benefit. The speed at which

the lines intersect is the breakeven point. This does not occur until 65

mph. It is irrational for the state to seek to enforce the law against drivers

going less than 65 mph.

Recall Justice Holmes’s suggestion that one should consider law from

the perspective of the bad man.

57

Figure 1 shows that being clocked at

speeds above the legal limit does not necessarily carry with it even the

threat of punishment. Correctly observing or intuiting this, the bad man

realizes that he can go as fast as 65 mph without fear of any punishment

at all, whether his speeding is detected or not. The speed limit in the books

is 55 mph, but the speed limit in action is 65 mph.

58

57

See supra notes 20–21 and accompanying text.

58

For ease of exposition, the following discussion presumes that the bad man does drive 65

mph under this insincere rule. That is not necessarily right. The bad man may go faster than

65 mph if he thinks that he will not be detected, or if he values speeding more than he fears

the potential punishment, or if his risk preferences make it rational for him to further violate

the law. These details complicate analysis without adding or clarifying anything important, so

we omit them in our simple model. Still, the careful reader may keep in mind that while

Benefit

+

0

Figure 1. Social Cost and Benefit of Trying to Prove a Wrong

Value

Speed

55 65

45

Sincere

speed limit

Law in

action

Cost

2019] Insincere Evidence 1133

Figure 1 illustrates a previously underexplored source of gaps between

law in books and law in action. While this is similar in some respects to

the rational slippage that results from a decision to spend less on police

so that resources will be available for schools and roads, it is actually a

functionally distinct and additive source of slippage that results from the

cost of the proof process itself. Being connected to the proof process gives

this source of slippage some concerning properties but also opens up

novel avenues for mitigation.

To start with the concerning properties, consider what happens to

enforcement as the burden of persuasion rises. Speeding is typically a

civil offense, so our example implicitly assumes that the state’s burden of

persuasion is preponderance of the evidence. In criminal cases, due

process demands that the state prove a violation of law beyond a

reasonable doubt.

59

Suppose we were to change the burden of persuasion

in our driving example, raising it from proof by a preponderance of the

evidence to proof beyond a reasonable doubt.

60

For any given size of

infraction, the state has a weaker case under the reasonable doubt standard

than it does under the preponderance standard. Graphically, this equates

to an outward shift of the cost curve, as Figure 2 illustrates.

61

Figure 1 shows the law imposes no incentive for the bad man to go slower than 65 mph, it is

possible that the bad man would go faster than this in practice.

59

See In re Winship, 397 U.S. 358, 364 (1970) (holding “that the Due Process Clause

protects the accused against conviction except upon proof beyond a reasonable doubt of every

fact necessary to constitute the crime with which he is charged”).

60

See Sean P. Sullivan, A Likelihood Story: The Theory of Legal Fact-Finding, 90 U. Colo.

L. Rev. 1, 18–22 (2019) (commenting on the interpretation of these intermediate standards).

61

Recall that the cost curve is defined as

!"#

. The probability of conviction is everywhere

lower under the higher burden of persuasion. Since

#

is everywhere lower, the cost curve is

everywhere higher under the higher burden of persuasion.

1134 Virginia Law Review [Vol. 105:1115

The breakeven point at which the state might rationally start trying to

enforce the law has shifted rightward. The speed limit remains 55 mph,

but under the higher burden of persuasion, drivers can now go as fast as

70 mph with impunity.

62

The higher the burden of persuasion, the greater

the violation of law must be before it will be rational for the state to

enforce. That is, higher burdens of persuasion lead to greater proof costs,

in turn leading to greater gaps between the law in books and the law in

action.

The point is not that higher burdens of persuasion are bad. The received

wisdom is that the social cost of proving guilt beyond a reasonable doubt

(including the cost of letting guilty people go free) is justified by the

benefit of avoiding false positives (the benefit of fewer innocent people

in jail).

63

This false positive versus false negative tradeoff is well

62

These numbers are stylized. In reality, a switch to the reasonable doubt standard could

result in much greater slippage than this.

63

See, e.g., Addington v. Texas, 441 U.S. 418, 423–24 (1979) (“In a criminal case . . . the

interests of the defendant are of such magnitude that . . . they have been protected by standards

of proof designed to exclude as nearly as possible the likelihood of an erroneous [conviction].

In the administration of criminal justice, our society imposes almost the entire risk of error

upon itself.”); In re Winship, 397 U.S. at 372 (Harlan, J., concurring) (“I view the requirement

of proof beyond a reasonable doubt in a criminal case as bottomed on a fundamental value

Benefit

+

0

Figure 2. Greater Slippage under Higher Burdens of Persuasion

Value

Speed

55 65

45

Cost

Sincere

speed limit

Law in

action

Higher

burden

70

Law in

action

2019] Insincere Evidence 1135

known.

64

What Figure 2 illustrates is not well known. As the burden of

persuasion rises, the band of marginal violations that the state will not

enforce also expands. The cost of raising the burden of persuasion is not

just that more guilty people go free. The cost is also that all bad men are

empowered to violate the law a little more extensively than they could

before.

The discussion up to this point can be summarized succinctly:

underdeterrence is inevitable when proof is costly. The reason is not any

of the traditional sources explored in the literature like monitoring costs

or inadequacy of the threatened sanction. Rather, this underdeterrence

traces to the costs of proving a violation of law. Proof is expensive,

particularly when the violation is small and the burden of persuasion is

heavy.

If proof costs frustrate enforcement, one wonders what lawmakers

could do to combat the problem.

65

Committing additional resources to

fund the litigation of small violations is not the answer. This would be

throwing money away on a gamble for which the costs have already been

shown to outweigh the benefits. What lawmakers would ideally do is

lower the cost of proof. How might this be achieved?

One idea might be to reduce the burden of persuasion. Figure 2 shows

that raising the burden of persuasion increases the cost of proof, thereby

increasing the gap between the law in books and the law in action. If

lawmakers could lower the burden of persuasion, they could shrink this

gap. The idea of varying the burden of persuasion to improve enforcement

has received considerable scholarly attention.

66

But while the idea works

determination of our society that it is far worse to convict an innocent man than to let a guilty

man go free.”); 2 McCormick on Evidence § 341 (Kenneth S. Broun ed., 7th ed. 2013)

(“Society has judged that it is significantly worse for an innocent person to be found guilty of

a crime than for a guilty person to go free.”).

64

See generally Larry Laudan, Truth, Error, and Criminal Law: An Essay in Legal

Epistemology (2006) (evaluating “legal errors” and strategies for their reduction).

65

The idea that legislators should act in conscious consideration of the implications of the

proof process and rules of evidence is not novel. See, e.g., 6 Jeremy Bentham, The Works of

Jeremy Bentham, chs. 3–4 (London, Simpkin, Marshall, & Co. 1843) (discussing the ends of

a legal system of proof and the duties of the legislator in relation to proof and evidence

problems).

66

See, e.g., Michael L. Davis, The Value of Truth and the Optimal Standard of Proof in

Legal Disputes, 10 J.L. Econ. & Org. 343, 344 (1994); Louis Kaplow, Likelihood Ratio Tests

and Legal Decision Rules, 16 Am. L. & Econ. Rev. 1, 35 (2014); Louis Kaplow, Burden of

Proof, 121 Yale L.J. 738, 815 (2012); Louis Kaplow, On the Optimal Burden of Proof, 119 J.

Pol. Econ. 1104, 1105 (2011); see also James Andreoni, Reasonable Doubt and the Optimal

Magnitude of Fines: Should the Penalty Fit the Crime?, 22 RAND J. Econ. 385, 393 (1991)

1136 Virginia Law Review [Vol. 105:1115

in theory, it has little practical potential. The justice system has long

encompassed roughly three burdens of persuasion,

67

and there is little

appetite for introducing more.

68

Even if there were such an appetite, the

constitutional mandate of proof beyond a reasonable doubt in criminal

cases would frustrate efforts to vary the burden in important

circumstances.

69

And even without these hurdles, it might simply be too

difficult to articulate burdens of persuasion with any greater granularity.

Even the existing burdens of persuasion seem to draw distinctions finer

than the legal system can handle.

70

Fine tuning the burden of persuasion is not the answer. Fortunately,

this does not end the analysis. There is another strategy that lawmakers

could use.

B. The Benefits of Insincerity

The state can lower the cost of proof, and thereby improve compliance,

by redefining the content of a wrong. Recall that lawmakers believe the

optimal speed is 55 mph. What if they do not adopt a sincere speed limit

of 55 mph but instead an insincere speed limit of 45 mph? Previously the

probability of convicting a driver clocked at 60 mph was relatively low,

so the opportunity cost of ticketing such a driver was high. With a 45-

mph speed limit, the probability of convicting a driver clocked at 60 mph

(discussing possible interdependence between the severity of a crime and fact-finders’

subjective burden of persuasion); Jack B. Weinstein, Some Difficulties in Devising Rules for

Determining Truth in Judicial Trials, 66 Colum. L. Rev. 223, 236 (1966) (commenting that,

while affording criminal defendants the benefits of a high burden of persuasion is a “deliberate

choice . . . based upon deeply held moral values and ethical judgments[,]” in the case of a

“particularly dangerous crime indulged in by relatively small numbers of the population and

a relatively small number of suspects, a perfectly rational argument could be developed for

conviction on the least shadow of a doubt.”).

67

See 2 McCormick on Evidence, supra note 63, §§ 339–41 (summarizing and discussing

the common burdens of persuasion); Sullivan, supra note 60, at 18–19 (same).

68

Cf. Michelson v. United States, 335 U.S. 469, 486 (1948) (explaining that while a

significant part of the law of evidence is archaic, paradoxical, and irrational, it should not be

disturbed because “somehow it has proved a workable even if clumsy system . . . . [and so to]

pull one misshapen stone out of the grotesque structure is more likely simply to upset its

present balance between adverse interests than to establish a rational edifice”).

69

The clear-and-convincing evidence standard, for example, has long been criticized as

vague and inconsistent in its application across proceedings. See Sullivan, supra note 60, at

19.

70

Cf. 2 McCormick on Evidence, supra note 63, § 340 (commenting that “[n]o high degree

of precision can be attained” through things like alternative articulations of intermediate

burdens of persuasion).

2019] Insincere Evidence 1137

is higher, making the opportunity cost of enforcing the law lower. In short,

the state can now enforce the speed limit more aggressively than before.

Figure 3 illustrates. The optimal speed remains 55 mph. Thus, the

benefit line does not change position. What changes is the cost curve. It

shifts inward. This captures the fact that some cases which were

previously hard have now become easy. A 1-mph violation has become

an easier-to-prove 11-mph violation. A 5-mph violation has become a 15-

mph violation, and so on. For any given degree of speeding, the

probability of successful proof has gone up. Thus, the cost of proving

every violation has gone down.

The shift in the cost curve changes the breakeven point at which

enforcement becomes worthwhile. As illustrated in Figure 3, when the

speed limit is set at the sincere 55 mph level, it is irrational for the state

to try to prove violations of the speed limit by drivers clocked at anything

less than 65 mph. When the speed limit is reduced to the insincere level

of 45 mph, it becomes rational to prove violations at a speed of 60 mph.

The gap between the law in the books and the law in action looks even

Benefit

+

0

Figure 3. The Insincere Rule

Value

Speed

55 65

45

Insincere

speed limit

Law in

action

60

Cost

1138 Virginia Law Review [Vol. 105:1115

worse than before: previously it was 10 mph,

71

now it is 15 mph.

72

But

compliance with the sincere (55 mph) benchmark has actually improved:

drivers travel 60 mph instead of 65 mph.

73

We want to emphasize that the state does not achieve the mandate of

the insincere rule. Drivers still violate the posted speed limit. But the state

does not care. It focuses on what the bad man prioritizes: the point at

which enforcement becomes rational for the state. This is what determines

the law in action. By reducing the cost of proof, the insincere rule gives

the state a credible threat of punishment against some infractions for

which it was previously helpless. Consequently, the bad man slows down.

C. Interpretation and Implications

Our title, Insincere Evidence, captures the mechanism by which

insincerity reduces proof costs. An insincere rule gives the state a block

of “free” or “artificial” evidence that can be used to help prove a violation

of law. From a proof-cost perspective, adopting an insincerely low speed

limit is the functional equivalent of the state secretly giving police officers

rigged radar guns that tack an extra 10 mph onto every reading.

In retrospect, this property is obvious. The effect of moving the speed

limit from the sincere (55 mph) level to the insincere (45 mph) level is

not to change the speed that a 65-mph driver was actually going. The

effect is simply to turn what had been a 10-mph infraction into what is

now a 20-mph infraction. And, from a proof-cost perspective, that is no

different than simply adding 10 mph onto whatever speed the radar gun

would have otherwise shown.

In respects other than proof-cost, however, an insincere rule is quite

different from a rigged radar gun. Indeed, it may out-perform the rigged

radar gun in practice. Judges and jurors are unlikely to convict if they

learn that the radar gun has been manipulated. An insincere rule could

continue to function even if the fact-finder doubted its sincerity.

The difference relates to the location of the insincerity. With a rigged

radar gun, insincerity is about the facts. The finding of legally operative

facts is the province of the fact-finder (judge or jury). The fact-finder

71

Under the sincere speed limit of 55 mph, Figure 3 indicates a breakeven point of 65 mph.

Thus, the gap between the law in action and the law in the books is 65 – 55 = 10 mph.

72

Under the insincere speed limit of 45 mph, Figure 3 indicates a breakeven point of 60

mph. Thus, the gap between the law in action and the law in the books is 60 – 45 = 15 mph.

73

As discussed previously, we assume for ease of exposition that drivers go as fast as they

can without triggering enforcement. See supra note 58.

2019] Insincere Evidence 1139

would not only be entitled to reject falsified evidence but probably

required to do so. By contrast, with an insincere rule, insincerity is about

the law. The statement of legal standards is the province of lawmakers,

not fact-finders. In a legal proceeding, under the charge of deciding

whether the evidence shows that the posted speed limit has been violated,

the fact-finder might well answer in the affirmative even if she personally

doubts the optimality of that speed limit.

We should emphasize this point. The rule-maker does not have to fool

the judge, the jury, the regulated parties, or anyone else in order for the

insincere rule to reduce the cost of proof. Everyone can understand and

agree that 55 mph is the optimal speed, even though the speed limit is 45

mph. The logic works as long as the judge or jury applies the law as

written: weighing the evidence against the posted speed limit. As a matter

of principle, this is how the legal system usually assumes that judges and

jurors will behave. As a matter of practice, judges and juries might often

behave this way. They might not know the optimal speed, so they just

enforce the law as written, or they might understand that enforcing the

law as written will draw actual speeds closer to the optimum. The logic

does not extend forever. Suppose lawmakers set the speed limit at 1 mph.

Officers might refuse to enforce such an absurd rule. Jury nullification

might result. But for the type of marginal shading of legal standards that

we discuss in this Article, abject refusals to enforce the law seem unlikely.

We have connected insincere evidence to artificial evidence. Now we

connect it to burdens of persuasion. Just as insincere rules and rigged

radar guns have the same theoretical effect on deterrence and

enforcement, there is a functional equivalence between adopting an

insincere rule and lightening the state’s burden of persuasion.

Figure 4 illustrates. To understand the figure, consider a series of

events. To begin, the state’s burden is preponderance of the evidence.

This corresponds to the solid cost curve which intersects the benefit curve

at 65 mph. Now suppose that the state is suddenly required to meet the

higher burden of proof beyond a reasonable doubt. This change has the

effect of shifting the cost curve to the right, as illustrated by the dashed,

black cost curve which intersects the benefit curve at 70 mph. The gap

between the law in the books and the law in action has widened. Finally,

suppose that lawmakers react to this higher burden of persuasion—and

the weaker enforcement achieved under it—by reducing the speed limit

to an insincerely low level. As illustrated by the dashed, gray cost line,

the insincere speed limit mitigates the increase in proof costs created by

1140 Virginia Law Review [Vol. 105:1115

the higher burden of persuasion. With a little bit more insincerity, this

offset would have returned the cost curve to the initial preponderance-of-

the-evidence level.

This may seem alarming. Consider criminal defendants and their

constitutional right to the protection afforded by proof beyond a

reasonable doubt. One implication of Figure 4 is that the benefit of this

protection may be weakened, or even eliminated, by the simple act of

more strictly defining what constitutes a wrong. While this is an accurate

statement, any alarm arising from it owes only to a naïve view of the

protection that a burden of persuasion affords. Figure 4 does not show

that insincere rules threaten the protection of the reasonable doubt

standard; it shows that the reasonable doubt standard is not—by itself—

much protection. The benefit enjoyed by a defendant is a function of both

the burden of persuasion and the definition of a wrong.

74

To the extent

that the Constitution constrains only one of these levers, its protection has

never been more than what Figure 4 suggests it to be.

74

See Richard H. McAdams, The Political Economy of Criminal Law and Procedure: The

Pessimists’ View, in Criminal Law Conversations 517, 519–20 (Paul H. Robinson, Stephen

P. Garvey & Kimberly Kessler Ferzan eds., 2009) (arguing that the state can make violations

of law easier to prove by lightening the burden of persuasion or by changing the elements of

the crime).

Benefit

+

0

Figure 4. Equivalence of Insincere Rule and Reduced Burden of Persuasion

Value

Speed

55 65

45

Cost

Insincere

rule

Higher

burden

70

2019] Insincere Evidence 1141

There is a more constructive way to think about the equivalence of

insincere evidence and burdens of persuasion. As noted, scholars have

long discussed manipulation of burdens of persuasion as a way of

achieving more efficient deterrence outcomes. Burdens of persuasion are

not susceptible to fine tuning, and thus have failed to deliver on this idea.

But many substantive legal thresholds are susceptible to detailed

refinement. Insincere evidence may be a practical way to unlock the long-

sought promise of variable burdens of persuasion.

D. Generalizations and Applications

We have explained our idea using a simple speed limit example. But

the idea is not limited to this context. Insincere rules function the same

way in any setting that meets the following three conditions: (1) the cost

of proof decreases as the magnitude of the legal violation grows; (2) the

rule-maker can adjust the content of the law in a way that converts

relatively minor violations of law into relatively major violations; and (3)

adjudicators apply the law as written. To emphasize the generality of

these conditions, we highlight several that do not appear on the list.

Deceit. As explained above, the rule-maker does not have to fool

anyone. Judges, jurors, the public—everyone can know that the law in

books is insincere. This does not matter. As long as the judge or jury

weighs the evidence against the stated rule—the actual, duly-enacted

law—then the cost of proof goes down. In many settings it seems

reasonable to assume that judges and jurors will often behave this way.

They may have no strong sense of what the sincere law would be (what

are the optimal mercury emissions from a power plant?), in which case

they simply apply the law in books. Or they may prioritize the law in

action, not the law in books, and realize that enforcing the insincere law

in books will improve the law in action.

Knowledge. Similarly, the functionality of insincere evidence does not

require the public to know the law is insincere. In some cases, regulated

parties may mistakenly believe that the law is sincere when it is insincere.

This does not matter. Think again of the speed limit example. If bad men

observe a posted speed limit of 45 mph, and if they learn from experience

or intuition that they can go as fast as 60 mph with impunity, they will go

60 mph. It does not matter whether they believe that the 45-mph speed

limit is sincere or not. The potential for state enforcers to become

1142 Virginia Law Review [Vol. 105:1115

confused about the sincerity of law is slightly more problematic, a point

we address in detail below.

75

Irrationality. Some people might imagine that insincere rules improve

behavior through a psychological mechanism. Judge Calabresi has argued

that “use of . . . ‘technically incorrect’ language”—the First Amendment

says Congress shall not abridge speech, but occasionally it does—“can

bring us closer to the desired result than would use of more precise

language.”

76

He attributes this to a kind of psychological bias: “If we

admit that the state can regulate religion, we are psychologically . . . more

likely to allow such regulation than if we say that there can be no

regulation of religion and then from time to time” regulate it anyway.

77

Our idea does not hinge on psychology. Insincere rules can lower proof

costs, and therefore improve behavior, even if everyone is strictly rational,

meaning they mechanically weigh the personal benefit of violating the

rule against the pain of punishment.

Objective uncertainty. Insincere rules do not require objective

uncertainty about the facts to function as we describe. In our speeding

example, we assume that radar guns measure speeds with error. Those

errors create evidentiary uncertainty. Insincere speed limits overcome that

uncertainty by turning close calls into sure things. But objective

uncertainty is not required. Perceived uncertainty works just as well. If

judges or jurors believe that radar guns measure with error, then they will

not convict drivers clocked at one mile-per-hour over the speed limit—

even if radar guns are actually flawless. They will, however, convict

drivers clocked at 11 mph over the limit. Insincere rules can overcome

perceived as well as real measurement errors.

Deterrence. Like much of the enforcement literature,

78

we have

assumed that the goal of enforcement is deterrence. In reality, rule-makers

might have other (or additional) objectives, like retribution,

incapacitation, or some notion of corrective justice. Insincere rules can

benefit rule-makers even when deterrence is not their objective. Rule-

makers can further multiple goals by drawing the behavior of regulated

parties toward their ideal.

Precisely scalar legal standards. Our speed limit example is scalar: the

limit can move up or down continuously, as though set by a dial.

75

See infra Section III.E (including a discussion of credulous enforcement officers).

76

Guido Calabresi, A Common Law for the Age of Statutes 173 (1982).

77

Id.

78

See supra Part II.

2019] Insincere Evidence 1143

Moreover, the speed limit is precise in the sense that most everyone

agrees, conceptually if not empirically, on what constitutes a violation of

the posted speed limit. Laws addressing countless subjects have this

precisely scalar feature and are good candidates for insincere rules: blood

alcohol content while driving; emissions of mercury, sulfur, carbon, or

other pollutants; the length, weight, and number of fish caught; the time

of day when hunters can hunt and marchers can march; decibel limits at

airports and nightclubs; the number of milligrams of acetaminophen or

other ingredients in each dose of medicine; distances between guns and

schools, protesters and abortion clinics, and campaigners and voting

booths. The list goes on.

Yet the logic of insincerity reaches beyond this broad class of laws.

The underlying variable does not have to be precise. Consider a regulation

requiring that facilities storing prescription drugs have “adequate

lighting.”

79

“Adequate” is vague; no one can say what exactly adequate

means, so no one can say for sure when lighting is inadequate. Thus, a

claim that a facility has inadequate lighting might fail in court because of

differing opinions. To solve this problem, the state can require that

facilities have “excellent” lighting. Like “adequate,” the term “excellent”

is vague. Nevertheless, it helps. Evidence that a facility is somewhat

dimly-lit may fail to convince a judge when the standard is “adequate”

but easily convince a judge if the standard is “excellent.” Foreseeing the

ease of proof, the facility will install better lighting if the standard

switches from “adequate” to “excellent.”

The previous example is straightforward but mundane. An example

with greater economic heft is the lack of a de minimis standard in U.S.

antitrust law. In many respects, U.S. and E.U. competition policy is

similar.

80

But unlike the E.U., which provides a safe harbor for small

violations of European competition laws,

81

U.S. antitrust law has never

79

See, e.g., 21 C.F.R. § 205.50 (2019).

80

See generally Margaret Bloom, The U.S. and EU Move Towards Substantial Antitrust

Convergence on Consumer Welfare Based Enforcement, 19 Antitrust 18 (2005) (discussing

the perception and reality of convergence between the U.S. and E.U. on a consumer welfare

standard in antitrust); Larry Fullerton & Megan Alvarez, Convergence in International Merger

Control, 26 Antitrust 20 (2012) (discussing the perception and reality of international

convergence in merger law).

81

See 2014 O.J. (C 291) 1 (describing conditions under which “agreements of minor

importance . . . do not appreciably restrict competition under Article 101(1) of the Treaty on

the Functioning of the European Union”).

1144 Virginia Law Review [Vol. 105:1115

evolved a clear minimum injury requirement.

82

In both regimes, small

antitrust violations are likely to go unpunished as too minor to warrant the

cost of enforcement. But, due to the lack of a de minimis standard,

moderate violations may be easier to prove in the U.S. than they are in the

E.U. If so, compliance with the antitrust laws may be stronger in the U.S

than it is in Europe.

Or consider an even more fundamental example. Some scholars argue

that the reasonable person standard in tort demands a higher level of care

than most people are capable of exercising on a regular basis.

83

This is

consistent with an insincere evidence strategy. The cost of proving

liability makes small acts of negligence too costly to prove, leading to

under-deterrence. The system corrects by requiring an insincerely strict

standard of care. Small acts of negligence that would be hard to prove

under a more subjective reasonable person standard may be easy to prove

under an artificially demanding objective standard.

The logic of insincerity extends further yet. It does not require even a

vaguely scalar variable. It can also work with “lumpy” variables, like

elements of a crime. To make a violation of law easier to prove, subtract

an element from the offense. We explain and comment further on this idea

below.

84

In sum, the conditions needed for insincere evidence to lower the cost

of proof are simple and general. This means insincere rules can be

deployed across the legal system. Are they? Professor Stuntz, to whom

we will return,

85

argues that criminal liability has broadened, leading to a

“world in which the law on the books makes everyone a felon.”

86

Professor Farber characterizes environmental standards as “threat points

82

U.S. law does have qualitative standards for the requisite degree of injury. See, e.g.,

Clayton Act, ch. 323, § 7, 38 Stat. 730, 731–32 (1914) (current version at 15 U.S.C. § 18

(2012)) (prohibiting mergers, the effects of which “may be to substantially lessen

competition . . . , or tend to create a monopoly”).

83

See, e.g., Fleming James, Jr., The Qualities of the Reasonable Man in Negligence Cases,

16 Mo. L. Rev. 1, 1–2 (1951) (“[M]any of the actor’s shortcomings such as awkwardness,

faulty perception, or poor judgment, are not taken into account [in the reasonable man

standard] if they fall below the general level of the community. This means that individuals

are often held guilty of legal fault for failing to live up to a standard which as a matter of fact

they cannot meet.”); see also Mark Grady, Res Ipsa Loquitur and Compliance Error, 142 U.

Pa. L. Rev. 887, 900 (1994) (“It is impossible to drive a car for any period of time without

missing a required precaution.”).

84

See infra Section IV.E (discussing adding and subtracting elements and the connection

between our theory of insincere rules and the work of Professor Stuntz).

85

See infra Part IV.E.

86

Stuntz, supra note 14, at 510–11.

2019] Insincere Evidence 1145

in negotiation,” arguing that “the criticism that regulatory standards are

too harsh loses some of its force, once it is recognized that the standards

are often only partially implemented.”

87

In principle, the IRS requires

extensive recordkeeping by some taxpayers, but in practice those

taxpayers have “been held to a somewhat less rigorous standard.”

88

A

New Jersey legislator proposed to double fines for speeding, and then,

when his proposal failed, introduced a bill to reduce speed limits.

89

All of

these examples and many others are consistent with the use of insincere

rules to generate insincere evidence.

E. The Limits of Insincere Evidence

This may sound too good to be true. The strategy of mitigating proof

costs through insincere evidence exhibits all the benefits of theft over

honest toil.

90

It drives down the cost of proving a violation of law, making

it easier for the state to punish transgressors. In so doing, it increases the

state’s ability to deter violations of law, and more efficiently incentivizes

compliance with laws. And it does all of this at seemingly no cost to the

state or society. No new police need to be hired. No new penalties need

to be administered. At no greater cost than the ink needed to change the

law, social benefits are unlocked.

There are indeed benefits to an insincere rule. But there are also

situations in which insincerity is an ineffective strategy. There are also

situations in which insincerity may backfire. We explore several of these

below.

To begin, the benefits of adopting an insincere rule depend on how

much the possibility of losing at trial factors into enforcement decisions.

Figures 5 illustrate this point. Unlike the figures above, this figure depicts

a cost curve that is already becoming flat when it intersects the benefit

curve. Shifting this cost curve to the left hardly changes the intersection

point: the law in action is 65 mph under the sincere rule and about 64 mph

87

Farber, supra note 16, at 315–16.

88

Stephen A. Zorn, The Federal Income Tax Treatment of Gambling: Fairness or Obsolete

Moralism?, 49 Tax Law. 1, 48 (1995).

89

Christopher Baxter, N.J. Assembly Panel Clears Bill Allowing Lower Speed Limits on

Quiet Residential Streets, NJ.com (May 13, 2013), https://www.nj.com/politics/2013/-

05/nj_assembly_panel_clears_bill.html [https://perma.cc/C5KE-H3S6].

90

Bertrand Russell, Introduction to Mathematical Philosophy 71 (1919) (“The method of

‘postulating’ what we want has many advantages; they are the same as the advantages of theft

over honest toil.”).

1146 Virginia Law Review [Vol. 105:1115

under the insincere rule. Intuitively, the fixed costs of proof (things like

the cost of preparing for trial) are the main drivers of slippage against the

posted speed limit. Since adopting an insincere rule does not change any

of these costs, its adoption has little effect on behavior. To restate that

point, the benefit of the insincere rule grows from the artificial evidence

that it gives the state to make its case. If the state does not need that

evidence because the violation is already easy to prove, then the adoption

of an insincere rule does not change behavior substantially.

One conclusion that seems to follow is that insincere rules may provide

the most benefit when applied to causes and crimes that require the state

to meet high burdens of persuasion. The higher the burden, the greater the

possibility of a loss. This is probably right, but the proposition could be

taken too far. With different assumptions about the shapes of the cost and

benefit curves, the effect of an insincere rule could be magnified, even

under the preponderance standard. And even if the insincere rule does not

imply as large a benefit under the preponderance standard as it does under

higher burdens of persuasion, its effect is not zero. To the extent that

insincere rules improve compliance without creating social costs, the

approach could be attractive.

Law in

action

Benefit

+

0

Value

Speed

55 65

45

Cost

Speed limit

Figure 5. Comparative Benefit of Insincere Rule

2019] Insincere Evidence 1147

But do insincere rules really have no social costs? In practice, there are

at least two actors whose behavior could be distorted by insincere rules to

the detriment of society: “good men” and credulous law enforcement

agents.

To start with the first and more intuitive of these costs, everything thus

far has assumed a society composed entirely of Holmesian “bad men.”

This assumption focuses analysis and is common in the enforcement

literature. In reality, however, there could be “good men” who follow the

law more-or-less to the letter.

91

In the language of economics, these

people have a “taste” for obeying the law irrespective of the potential

threat of sanctions.

The existence and prevalence of such “good men” is hardly obvious. A

person who happens to prefer acting as the law commands might in a

sense comply regardless of the threat of enforcement. But when personal

preference and the law diverge, compliance seems unlikely without the

backing of a credible threat of sanction.

92

The sale and use of marijuana

is illegal under federal law,

93

but enforcement of the federal law is lax and

marijuana dispensaries operate openly in many states. The unauthorized

use of copyrighted music and movies is illegal under both federal law and

international treaty, yet the ease of downloading or streaming music and

movies, and the difficulty of tracking such acts, makes digital piracy a

perennial problem.

94

Federal laws provide standards for the minimally

humane treatment of animals,

95

but these laws are poorly enforced and

acts of cruelty abound.

96

Without strict sanctions for non-appearance, it

91

See generally Tom Tyler, Why People Obey the Law (2006) (exploring the variables that

govern citizen compliance with the law and possible threats to such compliance); Robert

Cooter, Do Good Laws Make Good Citizens? An Economic Analysis of Internalized Norms,

86 Va. L. Rev. 1577, 1579 (2000) (arguing that some individuals internalize values that give

them a “taste” for following laws).

92

Frederick Schauer, The Force of Law 65 (2015) (“Indeed, when we turn from

experimental research to data on actual legal compliance, we find substantial support for the

hypothesis that unenforced law that does not track people’s law-independent

preferences . . . is often ineffective.”).

93

See United States v. Pickard, 100 F. Supp. 3d 981, 1006–09 (E.D. Cal. 2015) (surveying

arguments and opinions concerning the classification of marijuana as a Schedule I drug).

94

See, e.g., George E. Higgins, Digital Piracy, Self-Control Theory, and Rational Choice:

An Examination of the Role of Value, 1 Int’l J. Cyber Criminology 33, 33–35 (2007)

(discussing the prevalence and illegality of digital content piracy).

95

E.g., Humane Slaughter Act, 7 U.S.C. §§ 1901–1907 (2012); Twenty-Eight Hour Act, 49

U.S.C. § 80502 (2012); Animal Welfare Act, 7 U.S.C. §§ 2131–2159 (2012).

96

See, e.g., Jo Warrick, They Die Piece by Piece, Wash. Post (Apr. 10, 2001),

https://www.washingtonpost.com/archive/politics/2001/04/10/they-die-piece-by-

1148 Virginia Law Review [Vol. 105:1115

seems that as few as twenty percent of people will even respond to a

summons for jury duty.

97

Assuming, however, that scrupulously law-abiding good men did exist

in large numbers, they could be a problem for the insincere evidence

strategy. To see why, consider how an insincere speed limit would work

in a society composed of both good and bad men. The insincere limit will

cause everyone to slow down. Bad men might go 60 mph instead of 65

mph, which is good. But good men will go 45 mph instead of 55 mph,

which is bad.

98

The benefit of the insincere rule in correcting the behavior

of the bad men is offset—and possibly overtaken—by the distorting effect

of the insincere rule on the behavior of the good men.

99

This is an important caveat, but it does not change the fundamental

availability of insincerity as a cost mitigation strategy. In a society

composed of both good and bad men, some insincerity may still be

efficient if the cost-mitigation benefits outweigh the distorting effects. As

the proportion of good men in the population declines, the cost of

distorting their behavior shrinks and the benefit of insincerity grows.

100

A second potential cost of generating insincere evidence lies in the risk

of overenforcement. The intuition runs as follows. Suppose, as before,

that lawmakers want drivers to go 55 mph and thus adopt an insincere

speed limit of 45 mph. What happens if law enforcement agents conclude

that driving faster than 45 mph is socially harmful? This conclusion would

piece/f172dd3c-0383-49f8-b6d8-347e04b68da1/?noredirect=on&utm_term=.e344064ae47b

[https://perma.cc/AKZ2-CUTU] (reporting the failure of the USDA to take substantive

remedial actions against even extreme violations of the Humane Slaughter Act); ASPCA,

USDA Enforcement of Animal Welfare Act Hits a New Low (Aug. 10, 2018),

https://www.aspca.org/news/usda-enforcement-animal-welfare-act-hits-new-low [https://per-

ma.cc/TQ3A-FC9E] (reporting a sharp decline in USDA’s already lax enforcement of the

Animal Welfare Act without reason to expect that violations of this law have become less

common).

97

See Schauer, supra note 92, at 65–66.

98

This is bad not only because good men go too slow, but because the speed differential

between good and bad men increases from 10 mph under the sincere rule to 15 mph under the

insincere rule. Higher speed differentials tend to cause more accidents. We thank Judge

Easterbrook for a conversation on this point.

99

To be clear, slowing down good men is costly in the following sense. They would prefer

to go faster, yet law slows them down (they pay a cost) that society does not recoup through

an offsetting benefit. In our setup, there is no gain associated with pushing drivers to go 45

mph.

100

We assume that the numbers of good and bad men in society are exogenous. In fact, the

numbers could be endogenous, meaning they change as rules become increasingly insincere.

This is an interesting possibility, but it does not fundamentally alter our analysis.

2019] Insincere Evidence 1149

be wrong: lawmakers think that driving faster than 55 mph is harmful, not

driving between 45 and 55 mph. But if these credulous law enforcers fail

to realize this, and if they act on their erroneous conclusions, then they

will seek to enforce the speed limit at speeds lower than what lawmakers

want.

Figure 6 illustrates. Lawmakers want drivers to go 55 mph and have

adopted an insincere speed limit of 45 mph to better achieve this end.

While the solid benefit line captures the true social benefits of enforcing

the law, credulous agents of the state think that 45 mph is the sincerely

optimal speed, and thus act as though the dashed benefit line reflects the

social benefits of enforcing the law. These credulous agents will try to

prove violations at lower speeds than they would if acting from the sincere

benefit curve. They begin enforcing the law at 55 mph. This is bad.

Punishing drivers for speeds around 55 mph entails greater social cost

than benefit. These credulous agents are over-enforcing the law.

Again, this is an important caveat, but it does not doom the strategy.

The problem does not arise at all if agents of the state focus on the correct

benefit curve. They might be directed to this curve by lawmakers or be

sophisticated enough to intuit its location. Even if the agents remain

Benefit

+

0

Value

Speed

55 65

45 60

Ins.

rule

Figure 6. Overenforcement

Perc eived

benefit

Speed limit Law in

action

Cost

1150 Virginia Law Review [Vol. 105:1115

credulous, lawmakers may have levers for limiting over-enforcement,

such as budgetary limitations or department policies. Consider, for

example, a Pennsylvania law stating that speeding drivers shall not be

convicted unless “the speed recorded is six or more miles per hour in

excess of the legal speed limit.”

101

Finally, even if the agents are

hopelessly credulous and lawmakers are unable to correct their behavior,

there may still be instances in which the adoption of an insincere rule

confers more benefits (in terms of reduced proof costs) than costs (in

terms of over-enforcement).

102

This discussion of agents provides an opportunity to address a concern

about discrimination in enforcement. If the speed limit is sincerely set at

55 mph, officers have legal authority to stop drivers going faster than 55

mph, including drivers going only, say, 56 mph. It would not be cost-

justified to ticket a 56-mph driver, but an officer could still do so to harass

certain drivers, like racial minorities. A reasonable apprehension is that

insincere rules could exacerbate this problem. Instead of having legal

authority to stop anyone exceeding the sincere 55 mph limit, officers

could now stop anyone exceeding the insincere 45 mph limit.

We share the overarching concern about discrimination in law

enforcement. However, insincere rules do not necessarily make this

problem worse. Discrimination may happen regardless of the content of

law. (When discriminatory police look to harass someone, do they care if

the speed limit is 35 or 45?) But suppose this is wrong and discrimination

does relate to the content of law. In other words, you must violate the

posted rule before the cops can harass you. Even in that scenario,

insincerity would only increase the potential for discriminatory

enforcement if it increased the number of people who violate the posted

rule. Our analysis precludes that possibility. We assume that everyone

violates the law to the extent they can, regardless of the rule. The same

number of people who speed under a sincere speed limit will speed under

an insincere speed limit. Thus, the insincere rule does not increase the

number of people subject to discriminatory enforcement.

101

75 Pa. Cons. Stat. § 3368(c)(4) (2018). This law does not eliminate hard cases. If the

posted speed limit is 55 mph, and if an officer clocks a driver at 62 mph, the officer might

have trouble proving that the driver exceeded the speed limit by at least six miles per hour. To

avoid this, officers might wait until drivers exceed the limit by, say, 10 mph. This is consistent

with our argument. The Pennsylvania rule, coupled with officers’ own incentives, mitigates

overenforcement.

102

The net benefit of an insincere rule with credulous enforcers depends on assumptions

about the shape and positions of the social cost and benefit curves, among other things.

2019] Insincere Evidence 1151

Of course, our analysis might be wrong. In reality, some people might

violate an insincere rule who would not have violated a sincere rule.

103

If

they do, and if enforcement agents discriminate, then insincere rules

increase the number of people subject to discriminatory enforcement. If

there were no way to counter that discrimination (e.g., better training or

agent screening), then we face a tradeoff. Insincere rules create benefits

by improving the behavior of regulated parties, and they create costs by

exacerbating discrimination by enforcers.

104

To optimize this tradeoff,

rule-makers can tinker with the rule. Perhaps it should be less insincere

than otherwise—for example, 50 mph instead of 45 mph. Perhaps it

should be sincere. Or perhaps it should be insincere in the other direction.

If the state sets the speed limit at 75 mph, drivers go too fast, which is

bad. But, because few people exceed the rule, discrimination by officers

is limited, and that is good.

105

F. Summary

The purpose of this discussion has been to explain how insincere rules

can reduce proof costs and thus improve compliance with law. We can

summarize the argument in short order. Enforcement costs cause

underdeterrence. Proving a violation of law is an underexplored but

potentially important source of enforcement costs. Anything that lowers

the costs of proof lowers enforcement costs, and thus improves behavior.

Insincere rules lower proof costs by turning minor violations (ordinarily

cost-prohibitive to prove in court) into major violations. This use of

insincere rules to generate insincere evidence will generally not lead to

fully optimal behavior, and is subject to potential costs and limitations.

The strategy is robust, however, and seems suited to modest deployment

in many settings.

103

Perhaps, for example, driving preferences are heterogeneous and distributed in such a

way that more drivers will elect to violate a lower speed limit than will elect to violate a higher

speed limit.

104

As is conventional in economics, we assume that all benefits and costs, including costs

of discrimination, are commensurable. In the grand social welfare function, better enforcement

and worse discrimination can be traded off against one another. We do not take a position on

how much weight to ascribe to these offsetting factors, but we share what we suspect is a

common view: it would take a lot in terms of improved enforcement to compensate for even

a little more discrimination.

105

Cf. Dhammika Dharmapala, Nuno Garoupa & Richard H. McAdams, Punitive Police?

Agency Costs, Law Enforcement, and Criminal Procedure, 45 J. Legal Stud. 105, 108 (2016)

(explaining that criminal procedure protections can offset excessive enforcement zeal).

1152 Virginia Law Review [Vol. 105:1115

IV. DISTINGUISHING INSINCERE EVIDENCE

Besides us, scholars have not used the language of insincere evidence

to consider the relationship between the costs of proof and the efficiency

of enforcement. Scholars have, however, described concepts that relate to

our argument in various ways. These include conduct rules, decision

rules, over-inclusive rules, and the like. This Part distinguishes our focus.

The key insight is that lawmakers attempting to implement an insincere

evidence strategy do not intend or expect regulated parties to actually

obey the letter of an insincere rule. This insight puts distance between our

ideas and many others, and it makes all of them clearer.

A. Gilbert on Insincere Rules

In 2015, one of us (Gilbert) published a paper titled Insincere Rules.

That paper argued that lawmakers could improve behavior with

artificially demanding rules. To adapt one of the paper’s examples,

suppose lawmakers want factories to emit no more than 60 units of

pollution. An insincere rule might cap emissions at 50 units with the

expectation that factories would rationally choose to emit 60 units, the

actual amount that lawmakers wanted in the first place.

Gilbert’s paper introduced the term insincere rules, and the conclusion

of that paper—“By making the law in books wrong, lawmakers can get

the law in action right”

106

—resembles ours here. What makes the present

Article different is the context in which insincerity is considered. Our

focus on the proof process differs in fundamental ways from the two

channels that Gilbert initially investigated: (1) the use of insincerity as a

means of increasing sanctions, and (2) the use of insincerity to deceive

bad men.

Take the first channel. As explained, one way for lawmakers to

improve compliance is by increasing penalties. Gilbert showed that when

lawmakers cannot increase penalties directly—because of public outcry

against very high fines, for example—they may be able to increase

penalties indirectly by tightening rules. If a fine of $1,000 for every unit

of pollution above 60 is unpopular, lawmakers might replace it with a fine

of $500 for every unit of pollution above 50. A factory emitting 70 units

of pollution pays the same price either way ($1,000 × 10 = $10,000 is the

same as $500 × 20 = $10,000), but the latter strategy does not appear to

106

Gilbert, supra note 4, at 2187.

2019] Insincere Evidence 1153

involve as severe a penalty, and thus may be more politically feasible.

107

In short, insincere rules can act as penalty-enhancements. We discussed

penalties previously,

108

and we assume that they have already been

optimized for purposes of this Article.

109

Now consider Gilbert’s second channel, which involved deception.

Suppose the pollution law did not involve a fine but an injunction: if a

factory emits more than the pollution cap, the factory will be enjoined (via

a monitor) to comply with the legal limit moving forward. Insincerity may

allow enforcement agents to appear more aggressive than they really are.

If agents want pollution capped at 60 units and lawmakers set the law at

that level (i.e., the rule is sincere), factory owners may exploit detection

problems and other frictions to emit 70 units of pollution. Now suppose

lawmakers set the cap at 50 units (the insincere rule) but factory owners

believe this is what enforcement agents want. The same exploitation of

detection problems may lead them to emit 60 units of pollution—just what

the enforcement agents actually want. Unlike our focus in this Article,

Gilbert’s channel requires deception. Insincerity does not work, in this

way, unless factory owners are duped into thinking that the insincere rule

is sincere. Our approach does not require deception. Even if the bad man

realizes that lawmakers have adopted an insincere rule, the reduction in

proof costs remains the same.

In sum, Gilbert introduced insincere rules but studied only two

mechanisms by which they might elicit better behavior. Neither of those

mechanisms overlaps with the channel we develop here. Our focus is on

the use of insincerity as a means of lowering proof costs, which differs

fundamentally from what Gilbert developed in his earlier work. Insincere

evidence thus adds to the mechanisms Gilbert studied, and is, we suspect,

more common in practice.

B. Conduct and Decision Rules

Professor Meir Dan-Cohen distinguishes “conduct” rules, which direct

the public on how to behave, from “decision” rules, which direct officials

107

To express the idea in a different way, the insincere rule takes a particular behavior that

used to be classified as something akin to a misdemeanor and reclassifies it into something

akin to a felony. Because felonies carry larger fines, the insincere rule enhances punishment.

It does so without changing the fines for either misdemeanors or felonies.

108

See supra notes 24–32 and accompanying text (discussing the deterrence effects of strict

penalties).

109

See supra note 38 and accompanying text (discussing this maintained assumption).

1154 Virginia Law Review [Vol. 105:1115

on how to treat people who violate the law.

110

For example, a conduct rule

forbids theft, while a decision rule instructs judges to punish theft unless

it is committed under circumstances like duress. Dan-Cohen argues that

with “acoustic separation,” members of the public do not perceive

decision rules, just conduct rules, which improves behavior and

outcomes.

111

People commit fewer thefts because they do not realize that

duress is a defense. Yet when thefts do occur, judges have the flexibility

to excuse them. Acoustic separation “permits the law to maintain higher

degrees of both deterrence and leniency than could otherwise coexist.”

112

Lawmakers might try to achieve acoustic separation, though Dan-

Cohen did not push this possibility.

113

Suppose they did. Would this

approximate an insincere rule?

The answer is no. By Dan-Cohen’s account, lawmakers expect the

public to obey the conduct rule. People should act as though theft is

forbidden. In contrast, lawmakers do not expect the public to obey an

insincere rule: they adopt the insincere rule with the expectation that

people will violate it. In the driving example, lawmakers adopt a speed

limit of 45 mph to elicit a speed like 60 mph. If people strictly obey the

speed limit—as we discussed in the case of too many “good men,” for

example

114

—the insincere rule will not function as expected.

Indeed, one way to think about insincere evidence is that it represents

a strategic response to the failure of acoustic separation. Insincere rules

presuppose that people have discovered the decision rule. In our running

example, the sincere speed limit (55 mph) is the conduct rule, and the

110

Meir Dan-Cohen, Decision Rules and Conduct Rules: On Acoustic Separation in

Criminal Law, 97 Harv. L. Rev. 625 (1984). Dan-Cohen traces his work to Bentham. See id.

at 626; see also Gerald J. Postema, Bentham and the Common Law Tradition 408, 448–52

(1986) (explaining Bentham’s distinction between the law directing social interaction and the

law providing parameters for adjudication).

111

See Dan-Cohen, supra note 110, at 630–34.

112

Id. at 665.

113

Id. at 635 (“[A]ctual legal systems may in fact avail themselves of the benefits of acoustic

separation by engaging in ‘selective transmission’—that is, the transmission of different

normative messages to officials and to the general public, respectively. . . . I shall refer to these

techniques as strategies of selective transmission. The term ‘strategies’ calls for an

explanation. My use of the term should not be understood to connote deliberate, purposeful

human action. Imputing to the law strategies of selective transmission does not, therefore,

imply a conspiracy view of lawmaking in which legislators, judges, and other decisionmakers

plot strategies for segregating their normative communications more effectively. Instead,

strategies of selective transmission may be . . . strategies without a strategist . . . .” (footnotes

omitted)).

114

See supra Section III.E.

2019] Insincere Evidence 1155

speed limit coupled with the enforcement strategy (stop people going 65

mph or faster) is the decision rule. Our starting point is the assumption

that people know the decision rule, so they choose to drive 65 mph.

Changing the conduct rule, which everyone ignores, from 55 mph to 45

mph changes the decision rule, which everyone heeds, from 65 mph to 60

mph. Where acoustic separation has failed, insincere rules may succeed.

C. Over-Inclusive Rules

Rules are generally over- and under-inclusive when assessed against

their purpose.

115

Consider the 26th Amendment, which enfranchises

Americans aged 18 years or older.

116

Assume that the purpose of the age

requirement is to confine suffrage to those with sufficient maturity and

knowledge to cast votes in a socially responsible way.

117

The age

requirement is both over- and under-inclusive given this purpose. The

requirement is over-inclusive because it forbids too much: some 17-year-

olds are mature, knowledgeable, and responsible, yet they cannot vote.

118

And it is under-inclusive because it permits too much: some 18-year-olds

are not mature, knowledgeable, or responsible, yet they can vote. Good

rules balance the costs of over- and under-inclusiveness.

119

One might wonder if insincere rules have these features. Suppose

lawmakers want drivers to go 55 mph and adopt an insincere speed limit

of 45 mph. Is this simply an over-inclusive rule—one that prohibits

speeds (from 45 to 55 mph) that do not actually concern lawmakers at all?

The answer is no. Over-inclusiveness harms society. In the voting

context, if lawmakers want every responsible person to vote, they are

unhappy when their rule disenfranchises responsible 17-year-olds. This

does not mean they will change the rule. Perhaps the 18-year requirement

115

For a discussion, see Frederick Schauer, Playing by the Rules 31–34 (1991).

116

U.S. Const. amend. XXVI.

117

Some may dispute this. In the United States, the amendment was tied to the draft. See,

e.g., Jenny Diamond Cheng, Voting Rights for Millennials: Breathing New Life into the

Twenty-Sixth Amendment, 67 Syracuse L. Rev. 653, 668 (2017) (“This initial burst of interest

in eighteen-year-old voting lasted about five years and largely revolved around the perceived

injustice of drafting soldiers who were considered too young to vote.”). This does not matter

for our purposes.

118

And their vote today may affect whether they get conscripted at age 18.

119

“Good” rules do other things too. See generally Schauer, supra note 115, at 135–67; Cass

R. Sunstein, Problems with Rules, 83 Calif. L. Rev. 953, 971–77 (1995) (arguing that rules

minimize informational costs, reduce the likelihood of bias, embolden and constrain decision-

makers in particular cases, and promote predictability).

1156 Virginia Law Review [Vol. 105:1115

strikes the best balance between the costs of under- and over-

inclusiveness. But lawmakers, and socially conscious citizens, still lament

the over-inclusiveness. By contrast, insincerity does not necessarily harm

society. Everyone could be happy to adopt a 45 mph speed limit—stricter

than sincere social preference—because it elicits better behavior by

regulated parties.

Here is another way to see the distinction. Lawmakers want and expect

people to obey over-inclusive rules. In the voting example, lawmakers

want everyone younger than 18 to stay away from the polls, regardless of

whether they are responsible. In contrast, lawmakers do not want or

expect people to obey insincere rules. They adopt insincere rules with the

hope and expectation that everyone will violate them.

D. Prophylactic Rules

Scholars use the term “prophylactic rules” in different ways. The

standard definition runs as follows: prophylactic rules are judge-made

rules that overprotect constitutional rights.

120

The classic example is the

Miranda warning.

121

If police coerce a suspect into confessing, the Fifth

Amendment’s privilege against self-incrimination prevents them from

using the confession in court. In Miranda v. Arizona, the U.S. Supreme

Court went further in holding that, for many confessions to be admissible,

police must have told suspects that they have the right to silence and

counsel, and the suspects must have voluntarily waived these rights.

122

The Miranda warning overprotects the privilege against self-

incrimination. It provides more protection for criminal suspects than the

Fifth Amendment requires.

120

Evan H. Caminker, Miranda and Some Puzzles of “Prophylactic” Rules, 70 U. Cin. L.

Rev. 1, 1 (2001). For other definitions, see id. at n.2; Brian K. Landsberg, Safeguarding

Constitutional Rights: The Uses and Limits of Prophylactic Rules, 66 Tenn. L. Rev. 925, 926–

30 (1999).

121

See Caminker, supra note 120; see also Mitchell N. Berman, Constitutional Decision

Rules, 90 Va. L. Rev. 1, 7, 13–14 (2004) (discussing the “battle . . . on the question of whether

Miranda announced a ‘prophylactic rule’”).

122

See, e.g., Miranda v. Arizona, 384 U.S. 436, 444 (1966) (“The defendant may waive

effectuation of these rights, provided the waiver is made voluntarily, knowingly and

intelligently.”); Moran v. Burbine, 475 U.S. 412, 421 (1986) (elaborating that a waiver of

Miranda rights “must have been voluntary in the sense that it was the product of a free and

deliberate choice rather than intimidation, coercion, or deception” and that it also “must have

been made with a full awareness of both the nature of the right being abandoned and the

consequences of the decision to abandon it”).

2019] Insincere Evidence 1157

Scholars disagree on whether prophylactic rules are common or rare,

legitimate or illegitimate.

123

Prophylactic rules also have a converse

(albeit one without a name): rules that under-enforce constitutional

rights.

124

These details do not concern us; our interest is in showing that neither

prophylactic rules nor their converse are the same as insincere rules. To

see the difference, consider a question: why do courts adopt prophylactic

rules? Professor Evan Caminker provides an explanation in the Fifth

Amendment context:

[A] seemingly straightforward, case-by-case inquiry into whether a

constitutional norm has been transgressed will result in what might be

called “adjudication errors,” meaning the production of false-negatives

and false-positives. . . . [S]ometimes, the Court will conclude that the

likelihood of false-negatives is unacceptably high; in other words, the

direct doctrinal inquiry actually proves to be insufficiently protective of

the constitutional values at stake given the persistence of

unconstitutional conduct. I believe this is the best, and a fully sufficient,

explanation for and justification of Miranda’s so-called prophylactic

rule . . . .

125

This logic is familiar from above. Rules, including constitutional

doctrines, are both over- and under-inclusive. The precursor to Miranda

was a totality-of-the-circumstances test. Courts considered, on a case-by-

case basis, whether a suspect’s confession was coerced. In practice, this

was imperfect.

126

Some coerced confessions were admitted (the approach

was under-inclusive); some uncoerced confessions were excluded (the

approach was over-inclusive). In Miranda, the Court concluded that the

balance was off. The problem of under-inclusiveness was too great and

the costs of mistakenly admitting coerced confessions were too high. The

123

Compare Joseph D. Grano, Prophylactic Rules in Criminal Procedure: A Question of

Article III Legitimacy, 80 Nw. U. L. Rev. 100, 106–11 (1985) (arguing that some prophylactic

rules like the Miranda warning are illegitimate), with David A. Strauss, The Ubiquity of

Prophylactic Rules, 55 U. Chi. L. Rev. 190, 190 (1988) (arguing that prophylactic rules “are

not exceptional measures of questionable legitimacy but are a central and necessary feature of

constitutional law”).

124

Lawrence Gene Sager, Fair Measure: The Legal Status of Underenforced Constitutional

Norms, 91 Harv. L. Rev. 1212, 1213 (1978). According to Professor Mitchell Berman, “the

debate over prophylactic rules is parasitic upon a more fundamental contest over the logical

structure of constitutional adjudication.” Berman, supra note 121, at 50.

125

Caminker, supra note 120, at 9.

126

Berman, supra note 121, at 127.

1158 Virginia Law Review [Vol. 105:1115

solution lay in rebalancing the rule. The Miranda warning overprotects

the Fifth Amendment in the sense that it is self-evidently over-inclusive.

This has the undesirable effect of excluding many uncoerced confessions

from evidence, but may still strike the best balance of under- and over-

inclusion costs.

As this discussion shows, prophylactic rules are akin to over-inclusive

rules. The distinction between insincere rules and prophylactic rules is

thus the same as above. Lawmakers lament the over-inclusiveness of

prophylactic rules, but they expect people to follow them. Lawmakers

welcome the over-inclusiveness of insincere rules, and they do not expect

people to follow them.

E. Pathological Laws

Insincere rules are distinct from many common legal concepts, but they

resemble one important idea in criminal law. In 2001, Professor William

Stuntz published a paper called The Pathological Politics of Criminal

Law. Consider this quote from his article:

Suppose a given criminal statute contains elements ABC; suppose

further that C is hard to prove, but prosecutors believe they know when

it exists. Legislatures can make it easier to convict offenders by adding

new crime AB, leaving it to prosecutors to decide when C is present and

when it is not.

127

Stuntz characterizes this scenario in different ways. In one, he explains

the replacement of crime ABC with crime AB in terms of enforcement

costs: “Substituting an easy-to-prove crime for one that is harder to

establish obviously makes criminal litigation cheaper for the

government.”

128

We would put it like this: Stuntz imagines a multi-

element, binary, insincere rule.

We can unpack this phrase. Our running example, speed limits,

involves a single-element offense. Whether a driver breaks the law

depends on a single variable. Imagine a multi-element offense instead.

Suppose a motorcyclist engages in reckless driving when he (1) drives

faster than 55 mph and (2) raises his front tire more than five inches off

the ground. As we have explained, the cost of proof will prevent perfect

enforcement. The law in books is 55 mph/five inches, but the law in action

127

Stuntz, supra note 14, at 519.

128

Id. at 520.

2019] Insincere Evidence 1159

is, say, 65 mph/eight inches. Lawmakers could benefit from an insincere

rule. They could make the law in books, say, 45 mph/two inches. This

might improve the law in action. Because the law involves more than one

element, and because the threshold for each element is insincere, this is a

multi-element, insincere rule. The logic behind the rule—how it elicits

better behavior—matches the logic from above. The only difference is we

have two elements instead of one.

Consider another distinction: elements and thresholds. Elements of an

offense can be added or subtracted. Reckless driving can have two

elements (speed and wheelies) or just one element (speed). Once the

elements are determined, each gets a threshold.

129

Driving turns into

speeding when it surpasses the speed limit. Likewise, lawful wheelies

become unlawful at some threshold—say, five inches off the ground.

When a threshold reaches zero, its element disappears.

130

To illustrate,

suppose again that a motorcyclist engages in reckless driving when he (1)

drives faster than 55 and (2) raises his front tire more than five inches.

This is the law in books. The law in action is 65 mph/eight inches. To

shrink this gap, the state adopts an insincere rule. It lowers the wheelie

threshold—from five inches to three inches and so on. When the threshold

reaches zero, the wheelie element evaporates. Now reckless driving is a

one-element crime rather than a two-element crime. It depends on speed

only.

Given this logic, a lawmaker calibrating an insincere rule can approach

the wheelie threshold in two ways. She can treat it as continuous,

dropping the threshold from five inches to four, two, or whatever will

elicit the best behavior. Or she can treat it as binary: the threshold stays

at five inches, or it drops to zero, meaning the element goes away. There

is no choice between.

Recall our description of Stuntz: he gestures at a multi-element, binary,

insincere rule. Now the phrase makes sense. The crimes ABC and AB are

multi-element. The crime AB is insincere. The state does not want to

punish AB; it criminalizes AB only to lower the cost of enforcing ABC.

The insincere rule is binary. The state does not lower the threshold on

element C or even contemplate doing so. It simply eliminates C.

Our work builds on Stuntz. He offered a compelling example which,

under one of his characterizations, represents an insincere rule. We

129

We assume continuous variables.

130

Equivalently, when the threshold reaches zero, the element remains, but proof of the

element becomes certain and costless.

1160 Virginia Law Review [Vol. 105:1115

improve and generalize the idea. We show that insincerity works even for

one element-crimes; adding elements complicates matters without

changing the logic. We show that over-inclusiveness is a separate matter

(this line is blurry in Stuntz’s article). We show that the binary approach

Stuntz contemplated is actually an extreme and rather clumsy example of

an insincere rule. A continuous approach that fine-tunes the rule can work

better and might be more common (more on this below). We connect the

substance of the rule to the burden of persuasion in a clearer and fuller

way. Finally, we show that insincere rules can operate across enforcement

settings, not just in criminal law.

Stuntz studied “pathological politics.” The details of his argument are

beyond our scope, but here is a short summary. Legislators make more-

or-less everything a crime, and then prosecutors pick and choose whom

to prosecute among the law’s many offenders. Since more-or-less

everything is illegal, proving that any particular person has violated the

law—some law, whether minor or serious—is easy. The threat of

conviction induces guilty pleas from guilty and innocent offenders alike.

Prosecutors’ discretionary choices about enforcement determine the law

in action. Instead of legislators making law and judges adjudicating it,

prosecutors do all of the work. “Criminal law,” Stuntz wrote, is “not law

at all, but a veil that hides a system that allocates criminal punishment

discretionarily.”

131

We share Stuntz’s normative concerns with overcriminalization, and

we do not refute the descriptive accuracy of some of his claims.

Furthermore, we recognize that insincere rules could potentially

exacerbate discriminatory enforcement, as discussed above.

132

Nevertheless, we can extract insincere rules from Stuntz’s dark milieu.

Enforcement agents might exercise discretion judiciously. When they

do, insincere rules can improve rather than worsen the justice system. To

restate that point, insincere rules work especially well in systems without

discrimination in enforcement. Furthermore, our analysis pinpoints and

solves a tension in Stuntz’s analysis. He assumes the state cannot or does

not enforce crime ABC, while it can and does enforce AB—even though

AB should not be a crime. To make sense of this, Stuntz must assume that

the state overcorrects. If ABC is too costly to enforce, then AB is too

cheap. If this is what Stuntz has in mind, then the problem is not

131

Stuntz, supra note 14, at 599.

132

See supra Section III.E.

2019] Insincere Evidence 1161

insincerity per se but fine-tuning. The trick is to calibrate. Whenever

possible, the state should set the insincere rule so that the point at which

prosecutors find it worthwhile to enforce matches the point at which the

conduct becomes socially harmful. Perhaps because he focuses on the

binary case only, Stuntz does not consider this possibility.

F. Proxy Crimes

Proxy crimes are laws prohibiting behavior that is not itself harmful,

but that is associated with behavior that is harmful.

133

To give examples,

laws prohibiting open containers of alcohol in vehicles

134

and laws

prohibiting possession of burglar’s tools

135

are proxy crimes. These acts

are not harmful by themselves, but they correlate with the harmful acts of

drunk driving and burgling.

136

To analyze proxy crimes, we focus on the law forbidding open

containers of alcohol in vehicles. One way to understand this proxy crime

is as follows: the law always means what is says. It always forbids open

containers of alcohol in cars. That is what the rule-makers expect and

intend, on the theory that open containers very likely lead to harm. On

this understanding, “proxy crime” is just another term for a deliberately

over-inclusive or prophylactic rule. This kind of proxy crime differs from

insincere rules for the usual reasons: lawmakers want and expect

regulated parties to violate insincere rules. They do not want or expect

people to violate, to any degree, the prohibition on open containers in

vehicles.

There is another way to understand the proxy crime. Rule-makers do

not intend to forbid all open containers of alcohol in vehicles. In fact, they

do not care about open containers in general, they just want to forbid

drunk driving. They only expect officers to enforce the law against drunk

drivers. By eliminating some pesky elements (the vehicle was moving,

the driver was drunk), the proxy crime makes the proof process easier. On

133

See Frederick Schauer, Bentham on Presumed Offences, 23 Utilitas 363 (2011); see also

Larry Alexander & Kimberly Kessler Ferzan, Reflections on Crime and Culpability: Problems

and Puzzles 83–92 (2018) (exploring the justifications for proxy crimes and concluding that

they “promote only mischief”); Richard H. McAdams, The Political Economy of Entrapment,

96 J. Crim. L. & Criminology 107, 156–64 (2005) (discussing how enforcement of proxy

crimes is geared towards prevention of crimes that have great social detriment).

134

McAdams, supra note 133, at 160–61.

135

Larry Alexander & Kimberly Kessler Ferzan, Crime and Culpability: A Theory of

Criminal Law 309 (2009).

136

McAdams, supra note 133, at 160–61.

1162 Virginia Law Review [Vol. 105:1115

this understanding, the “proxy crime” is something like a binary, multi-

element, insincere rule, á la Stuntz.

137

One contribution of this Article is to divide proxy crimes into two

concepts, one being conventional proxy crimes, and the other being

insincere rules. These different concepts are easily conflated, but we can

pull them apart.

Proxy crimes are often criticized for penalizing conduct that is not

blameworthy, like driving home sober with a stoppered bottle of wine.

Under the conventional conception of proxy crimes, this criticism is

inevitable. Conventional proxy crimes are deliberately over-inclusive.

But under the insincere conception of proxy crimes, this criticism is not

inevitable. If the law in action penalizes blameless conduct, the insincere

rule needs fine-tuning.

V. NORMATIVE IMPLICATIONS: COSTS, TRUTH, AND TRIALS

Insincere evidence reduces proof costs, thereby increasing the state’s

ability to punish lawbreaking, thereby improving compliance with law.

The strategy does not always work well. Remember those “good men”

and credulous enforcers? But it can work very well. And because of this,

we believe it could be common in practice.

Assuming it is used, is it laudable? Better compliance with law is a

good thing, but the way insincere evidence achieves better compliance

may raise some concerns. One involves prosecutorial discretion, which

we addressed above.

138

This Part addresses other normative concerns.

First, we consider whether the justifiability of insincere evidence depends

on the source of proof costs. Second, we consider whether insincere

evidence is too much of an affront to truth. Third, we consider whether

cost reduction is a sufficient justification for introducing insincerity into

legal standards. In every case, we argue that the categorical rejection of

insincere evidence is unwarranted.

137

Stuntz did not characterize his analysis as involving proxy crimes. In fact, he never used

the word “proxy” in his article. Others understand his analysis in terms of proxies. See

McAdams, supra note 74, at 524 (“In Stuntz’s terms, selling drugs outside undercover

operations is crime ABC, while selling drugs in undercover operations is DEF; we punish the

otherwise harmless behavior DEF because we think it is correlated with ABC. . . . [T]he

undercover offense usually does not prove an external offense beyond a reasonable doubt. Yet

that turns out to be perfectly ordinary under the proxy approach.”).

138

See supra Section IV.E.

2019] Insincere Evidence 1163

A. Are There Better Ways to Reduce Costs?

We have assumed that the costs of proof spring from straightforward

sources: officers have to testify, lawyers have to draft pleadings, and so

on. To some extent these costs are inevitable. They can, however, be

modulated by the state and other forces. This raises questions about what

kinds of proof costs the state should be permitted to offset with insincere

rules.

Suppose there are two kinds of proof costs: unavoidable costs and

avoidable costs. Putting officers on the stand and having lawyers draft

critical documents are unavoidable costs. Making officers wait for hours

at the courthouse and making lawyers draft formalistic and redundant

documents are avoidable costs. While both costs have the same effect—a

gap arises between the law in books and the law in action—the

appropriate strategy for responding to those costs might differ. Insincere

rules might be a justifiable only in response to unavoidable costs. In the

case of avoidable costs, we might prefer the state to improve the

enforcement process by shortening wait times and streamlining pleadings.

It should not simply enact a stricter rule.

We are sympathetic to this argument. The case for insincerity seems

weaker when proof costs can be mitigated in other, straightforward ways.

The point should not, however, be taken too far.

We have assumed, up to this point, that lawmakers act as benevolent

and informed shepherds motivated by the desire to improve efficiency and

social order under the law. This assumption—which we make in common

with nearly all of the law and economics literature on enforcement

139

—

largely obviates the source-of-cost issue. A benevolent lawmaker

optimizes all proof costs, not just those reachable by an insincere rule. Put

another way, we have assumed that lawmakers will only enact an

insincere rule if the benefits outweigh the costs. The benefits, in this

framework, will be low if the state can costlessly achieve the same results

through streamlined pleadings and better docketing. Conversely, if these

reforms are themselves costly to implement, the argument against

insincere rules becomes less compelling. The seemingly sharp distinction

between avoidable and unavoidable costs can thus be seen as simply

139

See, e.g., Garoupa & Klerman, supra note 33, at 116 (“Previous analyses [in the law and

economics literature on enforcement] have assumed that the government aimed to maximize

social welfare.”); id. at 117 (“In the law enforcement literature, the optimal policy is usually

derived by maximizing social welfare, which is assumed to be the government’s objective

function.”).

1164 Virginia Law Review [Vol. 105:1115

another part of the cost-benefit analysis surrounding the adoption of an

insincere rule.

B. Is This an Affront to Truth?

Another potential concern is that the production of insincere evidence

rests on the introduction of systematic falsehoods into law. The strategy

reduces proof costs by substituting a disingenuous fact-finding exercise

for the genuine questions around which trial would otherwise revolve.

Rather than proving the fact of the sincere infraction (whether the

defendant violated the speed limit by 1 mph), trial is made to focus on the

insincere infraction (whether the defendant violated the speed limit by 11

mph). This is in more than a little tension with a core legal value: truth.

It is often said that trials are about nothing but the search for truth. This

applies equally to the institution of trial

140

and the rules that govern it.

141

At its most extreme, the Supreme Court has said that “[t]he fundamental

basis upon which all rules of evidence must rest—if they are to rest upon

reason—is their adaptation to the successful development of the truth.”

142

In a similar vein, it has referred to the “very nature” of trial as “a search

for truth.”

143

140

See David W. Peck, The Complement of Court and Counsel 9 (1954) (“The object of a

lawsuit is to get at the truth and arrive at the right result. That is the sole objective of the judge,

and counsel should never lose sight of that objective in thinking that the end purpose is to win

for his side.”); Marvin E. Frankel, Search for Truth: An Umpireal View, 123 U. Pa. L. Rev.

1031, 1033 (1975) (“Trials occur because there are questions of fact. In principle, the

paramount objective is the truth.”); Charles R. Nesson, Incentives to Spoliate Evidence in

Civil Litigation: The Need for Vigorous Judicial Action, 13 Cardozo L. Rev. 793, 793 (1991)

(describing “the ideal of the trial as a search for truth”); Robert S. Summers, Formal Legal

Truth and Substantive Truth in Judicial Fact-Finding–Their Justified Divergence in Some

Particular Cases, 18 Law & Phil. 497, 497 (1999) (“Some natural scientists, some social

scientists, some philosophers, and many others regularly assume that truth finding is the only

important function of trial court procedures and the rules of evidence.”).

141

See Chris William Sanchirico, Evidence Tampering, 53 Duke L.J. 1215, 1220 (2004)

(“Most analyses of evidence law take litigation’s prime object to be the discovery of truth

about past events.”); William Twining, Evidence and Legal Theory, 47 Mod. L. Rev. 261, 272

(1984) (“There is undoubtedly a dominant underlying theory of evidence in adjudication, in

which the central notions are truth, reason and justice under the law.”); Jack B. Weinstein,

supra note 66, at 246 (“In case of conflict, the court’s truth-finding function should receive

primary emphasis except when a constitutional limitation requires subservience to some

extrinsic public policy.”).

142

Funk v. United States, 290 U.S. 371, 381 (1933).

143

Nix v. Whiteside, 475 U.S. 157, 166 (1986); see also id. at 174 (describing the lawyer as

an “officer of the court and a key component of a system of justice, dedicated to a search for

truth” (emphasis added)); Lloyd v. Am. Exp. Lines, Inc., 580 F.2d 1179, 1186 (3d Cir. 1978)

2019] Insincere Evidence 1165

This may seem problematic for insincere rules. It is odd to think that a

system so focused on truth may be improved through the introduction of

falsehoods. But the situation is not so black and white.

First, even if trials were exclusively devoted to the search for truth, it

would not necessarily follow that insincere evidence would be counter to

this goal. The truth of what has happened is sought the same under either

a sincere rule or an insincere rule. In the driving example, the state seeks

to prove that a driver was going 56 mph (a truth in the world) regardless

of the sincerity or insincerity of the law. All that differs between the

sincere and insincere rule is the legal standard to which this behavior is

compared.

Second, there are reasons to question just how immutable truth-seeking

really is. There are philosophic reasons to doubt whether trials ever

uncover truth in an absolute sense of the term.

144

Litigants do not typically

have duties, or even incentives, to develop fully truthful records at trial.

145

Reasonable arguments can be made that fact-finding is basically a

functional concept: “legal truth,” as opposed to a substantive or platonic

truth.

146

This creates space for legal truths defined relative to insincere

rules.

Third, the rules of evidence and procedure already include many

examples of laws that frustrate truth-seeking. The most obvious is the

requirement of proof beyond a reasonable doubt in criminal cases. This

(“Any fact-finding process is ultimately a search for truth and justice, and legal precepts that

govern the reception of evidence must always be interpreted in light of this.”).

144

See generally Jerome Frank, Courts on Trial: Myth and Reality in American Justice 80–

102 (First Princeton Paperback ed. 1973) (critiquing the notion that the adversarial process

leads to the general discovery of truth); Sean P. Sullivan, Challenges for Comparative Fact-

Finding, 23 Int’l J. Evid. & Proof 100 (2019) (questioning when, if ever, a proposition of fact

could be proven true by any empirical method).

145

See Frank, supra note 144, at 81–89 (illustrating ways in which trial attorneys are

incentivized to conceal, rather than develop truthful testimony); Frankel, supra note 140, at

1032 (“My theme, to be elaborated at some length, is that our adversary system rates truth too

low among the values that institutions of justice are meant to serve.”). This is not to say that

proposals to address this have not been voiced. See id. at 1052–59 (proposing reform efforts

that would impose such duties). To date, however, there seems to be little movement toward

such an overhaul of the adversarial system.

146

See Summers, supra note 140, at 498 (“I define as ‘formal legal truth’ whatever is found

as fact by the legal fact-finder (judge or lay jurors or both), whether it accords with substantive

truth or not.”); cf. Edmund Morris Morgan, Some Problems of Proof Under the Anglo-

American System of Litigation 128 (1956) (“[T]he trial is a proceeding not for the discovery

of truth as such, but for the establishment of a basis of fact for the adjustment of a dispute

between litigants. Still it must never be forgotten that its prime objective is to have that basis

as close an approximation to the truth as is practicable.”).

1166 Virginia Law Review [Vol. 105:1115

standard is manifestly not facilitating truth-seeking.

147

It is aimed at an

important collateral objective: reducing the chances of wrongful

conviction when the stakes are high.

148

The reasonable doubt standard is

a prominent illustration, but it is hardly the only example.

Trials and evidence law serve many objectives,

149

and sometimes these

outweigh truth-seeking.

150

One example is the exclusion of relevant

evidence in order to incentivize desired behavior. Thus, subsequent

remedial measures are inadmissible to prove negligence in order to

incentivize the rapid remedy of potentially dangerous conditions.

151

Offers to pay medical expenses are inadmissible to show liability for an

injury in order to encourage the early payment of these expenses.

152

Exclusionary rules concerning statements made in settlement talks and

plea discussions prohibit the introduction of potentially damning

admissions in order to facilitate and encourage the settlement of

disputes.

153

Privileges are to the same effect. The attorney-client privilege

frustrates the discovery of truth in order to facilitate open communications

147

Frankel, supra note 140, at 1037 (“[I]n the last analysis truth is not the only goal. . . .

[T]he question is not at all ‘guilt or innocence,’ but only whether guilt has been shown beyond

a reasonable doubt.”); Weinstein, supra note 66, at 236 (describing requirements such as proof

beyond a reasonable doubt as “conscious distortions in the fact-finding process”).

148

See, e.g., In re Winship, 397 U.S. 358, 372 (1970) (Harlan, J., concurring) (“I view the

requirement of proof beyond a reasonable doubt in a criminal case as bottomed on a

fundamental value determination of our society that it is far worse to convict an innocent man

than to let a guilty man go free.”).

149

Summers, supra note 140, at 499–500 (“[T]rial court procedures and the rules of

evidence, though generally directed at substantive truth, are also designed to serve other ends

that actually come into play in a particular case.”); Weinstein, supra note 66, at 241 (“Even

were it theoretically possible to ascertain truth with a fair degree of certainty, it is doubtful

whether the judicial system and rules of evidence would be designed to do so. Trials in our

judicial system are intended to do more than merely determine what happened.”).

150

Monroe H. Freedman, Judge Frankel’s Search for Truth, 123 U. Pa. L. Rev. 1060, 1065

(1975) (“[I]n a society that respects the dignity of the individual, truth-seeking cannot be an

absolute value, but may be subordinated to other ends, although that subordination may

sometimes result in the distortion of the truth.” (footnote omitted)); Weinstein, supra note 66,

at 236 (“The law mandates many conscious distortions in the fact-finding process in order to

gain some supposed social advantage.”).

151

Fed. R. Evid. 407.

152

Fed. R. Evid. 409.

153

Fed. R. Evid. 408 (concerning compromise negotiations); Fed. R. Evid. 410 (concerning

plea discussions).

2019] Insincere Evidence 1167

by a person seeking legal counsel.

154

The doctor-patient privilege,

155

the

priest-penitent privilege,

156

the marital confidence privilege,

157

and the

spousal testimonial privilege

158

all incentivize some form of behavior or

communication that is deemed sufficiently beneficial to warrant distorting

the discovery of truth.

Other privileges and exclusionary rules operate in reverse:

manipulating the proof process to disincentivize socially undesirable

behavior. Rule 37 of the Federal Rules of Civil Procedure gives judges

flexibility to declare certain facts “true” or “false” as a way of penalizing

and thus deterring failures to comply with discovery orders.

159

In criminal

prosecutions, similar reasoning applies to the exclusion of evidence

obtained without a warrant, without a proper Miranda warning, or subject

to other procedural defects.

160

These rules frustrate the discovery of truth

in order to deter police misconduct.

154

See generally 1 McCormick on Evidence §§ 87–97 (Kenneth S. Broun ed., 7th ed. 2013)

(discussing the history, justifications, and scope of the attorney-client privilege in evidence

law).

155

See generally id. §§ 98–105 (discussing the history, justifications, and scope of the

physician-patient confidentiality).

156

See Trammel v. United States, 445 U.S. 40, 51 (1980) (“The priest-penitent privilege

recognizes the human need to disclose to a spiritual counselor, in total and absolute

confidence, what are believed to be flawed acts or thoughts and to receive priestly consolation

and guidance in return.”).

157

See generally Mikah K. Story, Twenty-First Century Pillow-Talk: Applicability of the

Marital Communications Privilege to Electronic Mail, 58 S.C. L. Rev. 275, 277–82 (2006)

(discussing the history and future of the marital communications privilege).

158

See generally R. Michael Cassidy, Reconsidering Spousal Privileges After Crawford, 33

Am. J. Crim. L. 339, 355–64 (2006) (discussing the origins, purposes, and evolution of the

spousal testimonial privilege).

159

See Fed. R. Civ. P. 37(b)(1)(A) (“If a party . . . fails to obey an order to provide or permit

discovery, . . . the court where the action is pending may issue further just orders. They may

include the following: (i) directing that the matters embraced in the order or other designated

facts be taken as established for purposes of the action, as the prevailing party claims; (ii)

prohibiting the disobedient party from supporting or opposing designated claims or defenses,

or from introducing designated matters in evidence . . . .”).

160

See, e.g., Miranda v. Arizona, 384 U.S. 436, 444 (1966) (“[T]he prosecution may not use

statements, whether exculpatory or inculpatory, stemming from custodial interrogation of the

defendant unless it demonstrates the use of procedural safeguards effective to secure the

privilege against self-incrimination.”); Escobedo v. Illinois, 378 U.S. 478, 491 (1964) (holding

that evidence is excluded when obtained by a violation of the constitutional right to

representation); Mapp v. Ohio, 367 U.S. 643, 659–60 (1961) (finding the need to exclude ill-

gotten evidence incorporated against the states).

1168 Virginia Law Review [Vol. 105:1115

The implications for insincere evidence are transparent. To the extent

that the behavioral benefits are sufficiently important, there is ample and

accepted precedent for frustrating truth to achieve that end.

C. Is Cost Reduction Enough?

A response to the previous argument might be that while truth-seeking

is sometimes frustrated to achieve important social ends, it isn’t clear that

cost reduction constitutes an important social end. Insincerity about the

actual normative content of law simply to economize on legal expenses is

unseemly.

Is it? Substantive laws often aim to encourage or deter certain

behaviors.

161

That effort is futile if the proof process gets in the way.

162

If

the proof needed to establish a violation of law were always prohibitively

costly, then the substantive law would never achieve its aim.

163

Even if

proof costs were only sometimes prohibitively costly, it would make

sense for lawmakers to try to reduce these frictions.

164

If the cost-reducing

effects of insincere evidence improves compliance with law, then—far

from unseemly—the use of insincerity serves the same laudatory ends as

the substantive law itself.

165

161

See Sanchirico, supra note 141, at 1220 (“When it comes to analyses of the ‘substantive

law,’ the idea that legal rules set incentives for everyday behavior—incentives to perform as

contracted, to disclose accurate financial information, to take reasonable precaution, to adopt

a safe product design, to eschew physical violence—occupies a central position.”).

162

See Summers, supra note 140, at 497–98 (“[W]ithout findings of fact that generally

accord with truth, the underlying policy goals or norms of the law could not be served. For

example, a rule designed to secure safety on the highways by setting a speed limit of 70 mph

and by punishing those who exceed it can not effectively serve its purpose if fact finders fail

to find the true facts as to the speed of actual offenders and so let speeders go free of penalty.”).

163

Weinstein, supra note 66, at 242 (“A system for determining issues of fact very accurately

in all tribunals might permit a few adjudications a year of almost impeccable precision. But

the resulting inability of the courts to have time to adjudicate the thousands of other pending

litigations would mean that justice would be frustrated; people could flout the substantive law

with relative impunity, knowing that the likelihood of being brought to trial was remote; and

plaintiffs would be forced to avoid litigation because of its extraordinary expense and delay.”).

164

See Chris William Sanchirico & George Triantis, Evidentiary Arbitrage: The Fabrication

of Evidence and the Verifiability of Contract Performance, 24 J.L. Econ. & Org. 72, 72 (2008)

(“The design of legal obligations, whether by a public body such as a legislature or by private

contract, should anticipate the enforcement process that induces compliance.”); see also id. at

73 (“Anticipating the judicial resolution of future disputes, contracting parties are likely to be

interested in the likelihood or cost of judicial truth-finding only to the extent that the court’s

ability to discern the truth efficiently improves contract incentives and the gains from trade.”).

165

Weinstein comes close to making this exact point. Weinstein, supra note 66, at 243

(“Unless we are to assume that the substantive law is perverse or irrelevant to the public

2019] Insincere Evidence 1169

In fact, this concern may be close to a strawman. Including cost

considerations among the social concerns of the legal system is not a new

proposition.

166

Alongside the discovery of truth, the Federal Rules of

Evidence list the elimination of “unjustifiable expense and delay” as one

of their core objectives.

167

The same phrase appears in the Federal Rules

of Criminal Procedure.

168

The Federal Rules of Civil Procedure provide

that they are to be construed “to secure the just, speedy, and inexpensive

determination of every action and proceeding.”

169

Practicality has never

compelled that cost considerations be ignored in promulgation of laws or

the structure of the legal process.

And if proof costs are included in the balancing of social interests

served by laws and the litigation system, then we are back to the same

cost-benefit analysis. So long as the benefits of insincere evidence

outweigh the costs, the frustration of truth is justifiable. An English court

made essentially this same observation nearly 200 years ago:

The discovery and vindication and establishment of truth are main

purposes certainly of the existence of Courts of Justice; still, [these

objectives,] however valuable and important, cannot be usefully

pursued without moderation . . . . Truth, like all other good things, may

be loved unwisely—may be pursued too keenly—may cost too much.

170

VI. CONCLUSION

In this Article, we have analyzed the ability of insincere evidence to

reduce proof costs. We have connected this effect to burdens of

persuasion and other legal concepts. We have also considered various

welfare, then its enforcement is properly the primary aim of litigation; and the substantive law

can be best enforced if litigation results in accurate determinations of facts made material by

the applicable rule of law.”). Here, however, it is not the accurate determination of facts, but

the cost of doing so, that determines the degree to which the litigation process will achieve the

substantive objectives of law.

166

See id. at 241 (stating that other objectives of the evidence system include “economizing

of resources, inspiring confidence, supporting independent social policies, permitting ease in

prediction and application, adding to the efficiency of the entire legal system, and tranquilizing

disputants”).

167

Fed. R. Evid. 102; see also Fed. R. Evid. 403 (including “undue delay” and “wasting

time” among the balancing factors on which a judge may exclude an otherwise relevant item

of evidence).

168

Fed. R. Crim. P. 2 (“These rules are to be interpreted to provide for the just determination

of every criminal proceeding . . . and to eliminate unjustifiable expense and delay.”).

169

Fed. R. Civ. P. 1 (emphasis added).

170

Pearse v. Pearse (1846) 63 Eng. Rep. 950, 957; 1 De G. & Sm. 11, 28–29.

1170 Virginia Law Review [Vol. 105:1115

qualifications to the effectiveness and desirability of insincere evidence.

Rather than summarizing that work, we conclude by sketching some open

questions and pathways for future research.

For one thing, insincere lawmaking complicates the notion of

legislative intent.

171

To reduce proof costs through insincerity, legislators

must drive a wedge between what they consider to be socially optimal

behavior for society (their sincere intent) and what they publicly say to

better effect outcomes (insincerity in the text of a statute and, possibly, in

the legislative history that accompanies it). The idea that legislators would

bluster during floor debates is not shocking, but the idea that they might

exaggerate their sincere preferences in the substance of law is less

familiar. The more one thinks insincere lawmaking occurs in practice, the

less compelling it becomes to equate legislative intent with traditional

sources like statutory text and legislative history.

The same tensions are magnified when agencies need to justify

exercises of their rulemaking authority. Under the Administrative

Procedure Act, as part of the rulemaking process, an agency must produce

“a concise general statement of [the] basis and purpose [of the adopted

rule].”

172

While this statement need not be exhaustive, the expectation is

that it will indicate the policy questions at issue as well as the agency’s

reason for choosing its rule.

173

Judges often focus on this explanatory

statement in deciding whether agency action should be set aside as

arbitrary and capricious.

174

Though deferential, that review is not

toothless.

175

Again, the use of insincerity complicates matters. Suppose,

for example, that a regulatory agency faced the speed limit problem we

have discussed. The need to explain itself would seem like an obstacle to

171

Of course, legislative intent is already complicated. See, e.g., Kenneth A. Shepsle,

Congress Is a “They,” Not an “It”: Legislative Intent as Oxymoron, 12 Int’l Rev. L. & Econ.

239 (1992).

172

Administrative Procedure Act, 5 U.S.C. § 553(c) (2012).

173

See, e.g., Indep. U.S. Tanker Owners Comm. v. Dole, 809 F.2d 847, 852 (D.C. Cir. 1987)

(“[S]uch a statement should indicate the major issues of policy that were raised in the

proceedings and explain why the agency decided to respond to these issues as it did,

particularly in light of the statutory objectives that the rule must serve.”).

174

Administrative Procedure Act, 5 U.S.C. § 706(2)(A) (2012) (requiring a reviewing court

to “hold unlawful and set aside agency action . . . found to be . . . arbitrary, capricious, an

abuse of discretion, or otherwise not in accordance with law”).

175

See Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43 (1983)

(endorsing fairly exacting hard look review of agency rulemaking); Citizens to Pres. Overton

Park, Inc. v. Volpe, 401 U.S. 402, 415 (1971) (explaining that a presumption of regularity “is

not to shield [agency] action from a thorough, probing, in-depth review”).

2019] Insincere Evidence 1171

using an insincere rule. Can the agency say to a court that it intentionally

adopted a too-strict rule to improve enforcement? Maybe not. On the

other hand, if the point of delegation to the agency was to achieve a given

enforcement outcome, and if the agency better achieves that outcome with

an insincere rule, then invalidation of the rule by the reviewing court

would undermine the intent of the legislature in empowering the agency

in the first place.

Before concluding, we should say something about the rule of law.

That divine, contested concept lurks in the background of every page.

176

One might wonder whether insincerity undermines the rule of law,

especially its accessibility. Law, the argument goes, must be publicized

and intelligible so that people can understand it.

177

To the extent that

insincere rules are adequately publicized, we do not see an obvious

tension. But perhaps they cut against accessibility. Perhaps when

deployed by judges, insincere rules could cut against norms of candor.

178

We do not work through these questions here.

We will, however, address a point about the perception of the rule of

law. That perception, one might argue, would suffer if people discover

the state to be deliberately misrepresenting its desired behavior. We are

not sure this is right. Perception of the rule of law probably depends on

both the law in books and the law in action. If people perceive the law in

books to be too harsh, perceptions may suffer. But if people perceive the

law in action to be about right, perceptions may improve. Insincere rules

have both effects. How the scales of perception balance is yet unknown.

176

See Brian Z. Tamanaha, On the Rule of Law: History, Politics, Theory 3–4 (2004)

(noting widespread support for the “rule of law,” but also substantial ambiguity and even

disagreement about what exactly this concept entails); Jeremy Waldron, The Rule of Law,

Stanford Encyclopedia of Philosophy (June 22, 2016), https://plato.stanford.edu/entries/rule-

of-law/ [https://perma.cc/GNX2-P2LH] (surveying disagreements about the desirability and

definition of rule of law under the headings of “The Contestedness of the Rule of Law,”

“Opposition to the Rule of Law,” and “Controversies about Application”).

177

See, e.g., Paul Gowder, The Rule of Law and Equity, 32 Law & Phil. 565, 582–85 (2013)

(discussing publicity of law as one uncontroversial aspect of rule of law concepts); Waldron,

supra note 176, § 2 (“[L]aw should be epistemically accessible: it should be a body of norms

promulgated as public knowledge so that people can study it, internalize it, figure out what it

requires of them, and use it as a framework for their plans and expectations and for settling

their disputes with others.”).

178

Cf. Micah Schwartzman, Judicial Sincerity, 94 Va. L. Rev. 987 (2008) (arguing for the

importance of sincerity and candor in judicial opinions).

1172 Virginia Law Review [Vol. 105:1115

TECHNICAL APPENDIX

The main text of this Article provides an intuitive explanation of our

theory of insincere evidence. Here, we extend that theory with a modest

formalization. The only unusual aspect of this model is its reliance on

likelihood concepts instead of posterior probabilities. One of us (Sullivan)

has elaborated elsewhere on the superiority of likelihood analysis as a

model of legal persuasion.

179

Among other things, our reliance on

likelihoods allows this model to avoid the always awkward problem of

specifying the prior distribution of violations in the population.

180

Suppose the state seeks to regulate some privately beneficial but

socially costly activity. Let

$ % &

be the level of activity a given

individual chooses to undertake. All else equal, this person prefers higher

levels of

$

. Let

' % &

be the legal limit that state lawmakers have adopted

for this activity. Activity levels

$ ( '

are illegal and subject to

punishment by the state.

Suppose that state agents can perfectly observe a person’s choice of

$

,

but cannot credibly disclose this information to the tribunal. Perhaps the

agents cannot credibly disclose their knowledge because the fact-finder

correctly predicts that they would exaggerate the extent of the observed

violation if given the chance. Instead, the state is limited to proving a

violation of law by objective evidence—a noisy signal of the actual

activity level. For sake of clarity, assume that the combined weight of the

objective evidence is distributed according to a normal distribution,

centered on the true activity level, and dispersed according to an

exogenous noise parameter,

) ( *

, such that the evidence draw in a given

case is

+,-.$/ )01

.

181

Agents of the state do not draw from the evidence

distribution until after the costs of trial have already been borne.

182

179

Sullivan, supra note 60.

180

Cf. Davis, supra note 66 (noting the complexities that population distributions (i.e., prior

probabilities) have in a probability-based model of the trial process).

181

The following generalizes in an obvious way to other evidence distributions.

182

This setup is an easy way to think about the situation, but it is not the only way to

understand the model. Alternatively, the regulator could be understood as having no direct

ability to observe

$

, but instead having access to an unbiased signal of

$

. Subject to some

complications around the possible communication of asymmetric information by the

individual, results in this alternative specification of the model seem much the same as in the

above specification. Another approach would be to interpret the state as having access to all

available evidence of the regulated party’s activity level, but subject to uncertainty about how

the fact-finder will weigh this evidence at trial. Again, this approach would seem to result in

largely the same model as the above specification.

2019] Insincere Evidence 1173

Violations of law are established by producing enough evidence to

convince the fact finder that

$ ( '

by the applicable burden of persuasion.

While it might seem natural to model this persuasion process in terms of

the posterior distribution

2.$3+1

, this would require us to assume a prior

distribution

2.$1

, which we do not do. Instead, we model the proof

process via likelihood reasoning. Intuitively, the likelihood approach

involves an iterative comparison of pairwise likelihood-ratio tests in order

to determine whether

$ ( '

has been established. The end result is that

proof of

$ ( '

under any given burden of persuasion requires an evidence

draw,

+

, such that the following condition is met:

456

789: ;.+3$</ ) 01

456

78=: ;.+3$</ )01( >/

where

;.+3$</ ) 01

is the probability density of the normal distribution,

evaluated at the observed evidence draw, under the assumed centrality

parameter,

$<

. The

>

term in this likelihood ratio is a numeric

representation of the applicable burden of persuasion. Sullivan has argued

that

> ? @

corresponds to preponderance of the evidence; that something

like

@ A > A @*

represents clear-and-convincing evidence; and that

something like

> ( @*

represents proof beyond a reasonable doubt.

183

In the case of evidence draws distributed according to the normal

distribution, it can be shown that the above general condition for proof of

$ ( '

by any arbitrary burden of persuasion requires an evidence draw,

+

, such that

+ ( ' B

C

D)0EFG >/

with all variables as defined above.

184

At the point where

$

is known but

+

is not known, then for a given choice of legal limit,

'

, and for given

183

Sullivan, supra note 60, at 26–37.

184

Here is an informal proof. Since

> H @

in every interesting application, a necessary

condition for the state to win at trial is

+ ( 'I

For the preponderance of the evidence standard,

this condition is both necessary and sufficient. For higher burdens of persuasion, assuming

+ ( '

, the maximum likelihood estimator of

$

is simply

+

, and the constrained maximum

likelihood estimator on the assumption that

$

′

J '

is simply

'

. As such, sufficient evidence

of violation has been produced whenever

;

.

+

3

+/ )0

1

( ;

.

+

3

'/ )0

1

K >

. Simplifying and

solving this expression for

+

leads to the stated condition.

1174 Virginia Law Review [Vol. 105:1115

levels of the burden of persuasion,

>

, and noise parameter,

)

, the

probability of successfully proving a violation of law for activity level

$

is

2

.

$/ '/ >

1

? @ L M

N

' B

C

D)0EFG >

O

$/ )0

P

/

with

M

.

Q

3

$/ )0

1 the normal cumulative distribution function.

With this definition of the probability of

2

.

$/ '/ >

1, we can

operationalize the remainder of the model. While we would ideally focus

on the state’s optimal choice of legal limit and enforcement strategy, the

assumptions needed to do so are more than we wish to defend at this

point.

185

To obtain tractable results without limiting our approach, we

characterize how the choice of

'

affects the necessary condition for non-

strategic enforcement of laws.

A necessary condition for the rational enforcement of any law is that

the expected marginal cost of attempting to prove a violation of law

should not outweigh the expected marginal benefit. Formalizing this

necessary condition requires only modest assumptions. Specifically, we

assume the existence of a known functional form for the cost of trying to

prove a violation of law:

R

.

$/ '

1

( *

. We also assume a known functional

form for the benefit of successfully proving a violation of law by any

given choice of activity level. This is a reduced-form expression for

whatever benefit society gets from the (unmodeled) marginal deterrence

arising from successful punishment of a given level of violation:

S.$/ '1

.

Combining these terms with the probability of successful proof at trial,

the necessary condition for the state to attempt to prove a violation of law

is

S

.

$/ '

1

K 2

.

$/ '/ >

1

( R

.

$/ '

1

I

185

For example, these assumptions would include precise specification of the welfare

implications of different choices of

'

; specification of individuals’ exact preferences over all

possible combinations of activity level, penalty, and probability of incurring a penalty,

including any variation in this preference, in risk tolerance, or in strategic type; and

specification of the game-theoretic equilibrium strategies individuals adopt in deciding when

to violate laws, and that both individuals and the state adopt in deciding when to litigate

disputed violations of law.

2019] Insincere Evidence 1175

As we explain in the Article, this necessary condition is more

conveniently illustrated with the probability term on the cost side of the

inequality:

S

.

$/ '

1

(R.$/ '1

2.$/ '/ >1I

Insincere Evidence

Abstract and Figures