2019

HOOVER INSTITUTION
Summer Policy Boot Camp
Director’s Award Winners
1
2
HOOVER INSTITUTION SUMMER POLICY BOOT CAMP
DIRECTOR’S AWARD
Capping off an intensive week-long study and discussion of the core principles and
tools of public policy, students were invited to apply their knowledge by researching
and developing a policy proposal. Following the principles of Hoover scholarship, the
proposals emphasize a specific recommendation using facts, data, and well-constructed
arguments. The papers summarize the significance of the new policy and the expected
result, expressed in essays of 2,000 words.

After review of each submission, we recognized three participants with the Hoover
Institution Summer Policy Boot Camp Director’s Award based on their outstanding work. The
winning proposals demonstrate particular creativity in addressing complex policy issues.

We are proud to include these proposals written by the following winners of the 2019
Hoover Institution Summer Policy Boot Camp Director’s Award.

• Timothy Anderson
• Winston (Wenhao) Du
• Nick Sileo
• Kaila Webb

Sincerely,

Scott Atlas Joshua Rauh

David and Joan Traitel Senior Fellow Senior Fellow and Director of Research
Hoover Institution Hoover Institution
Ormond Family Professor of Finance
Stanford Graduate School of Business

3
4
5
 CONTENTS

Less Can Be More When Regulating Deepfakes........................... 7

By Timothy Anderson

Introducing a Modified Good Samaritan Reduced-Price

Mechanism to Combat Hunger and Food Waste....................... 13
By Wenhao Winston Du, Vanderbilt University

Bad Seeds: A Case against Federal Farm Subsidies................. 20

By Nicholas G. Sileo, Princeton University

A Local Approach to the Opioid Epidemic................................ 26

By Kaila Webb, Wellesley College

6
Less Can Be More When Regulating Deepfakes
By Timothy Anderson, Department of Electrical Engineering, Stanford
University

On November 2, 2017, a forum on Reddit.com was created for posting and dis-
cussing pornographic videos known as “deepfakes” which used AI to map faces
of female celebrities such as Gal Gadot and Maisie Williams onto pre-existing
video clips. The Reddit page quickly exploded to several thousand visitors—many
of whom created and posted new videos—and gained wide attention after being
featured in Vice Motherboard.1 While the Reddit page was shut down after only
three months, the deepfakes phenomenon laid bare to the broader public the su-
perhuman capabilities of readily-available AI, the ethical bankruptcy of modern
tech culture, and the chilling reality that anyone with basic programming skills and
access to a modern computer now has the power to violate innocent others or dis-
rupt national politics without leaving their desk.

Left: Original image from Forrest Gump.

Right: Deepfake video of Keanu Reeves mapped onto Tom Hanks.2

Computers Learn to Trick You

Deepfakes are built on “deep neural networks,” a broad family of AI algorithms

that are nearly ubiquitous in today’s data processing systems, from search engine
results to voice assistants and automated photo tagging. The neural networks be-
hind deepfakes differ from other AI in that, instead of classifying data such as
words in an audio clip, these networks, known as “generative networks,” learn to
generate new data that is indistinguishable from existing real-world examples.3
Generative network–based algorithms have become some of the most powerful
in AI and are now capable of creating believable fake data in a wide range of
areas.

An interesting application of these algorithms is translating images between do-

mains, such as daytime to nighttime photos or Monet’s style to Van Gogh’s. How-
ever, as virtual reality pioneer Young Harvill said about his invention, “a lot of the
killer app for some of the first technologies is porn,”4 and unfortunately, image

7
 translation proved no exception. The same technology that can change horses to
look like zebras can also change the face in a photograph; less than a year after
publication of the hallmark papers on image translation with generative networks,5
the first deepfakes were posted to Reddit.

Artificial Intelligence for Social Ill

If an anonymous Reddit user can create realistic compromising footage of celeb-

rities, what would stop a vengeful ex-lover from creating footage of you? What
about a plaintiff submitting AI-fabricated evidence to alter a court case? A rogue
actor creating a fake video of the president declaring war?

Deepfakes have been justifiably denounced as everything from misogynistic and

exploitative toward women6 to a national security disaster in the making,7 as well
as an existential threat to the idea of video evidence and news itself.8 The growth
in nefarious applications of AI is largely because AI algorithms are now so easy
to build. Today, thanks to free software maintained by Google, Facebook, and
Nvidia, anyone with a college freshman’s knowledge of mathematics and computer
programming can create their own AI algorithms. Such resources have been a boon
for research in AI and other data-centric fields, but the reduced barriers to entry
also mean that bad actors everywhere now have a new tool at their disposal.

Existing Laws Can Already Regulate Deepfakes

The public has begun to demand action from lawmakers on deepfakes, especially
in the wake of viral fake videos of Mark Zuckerberg and Nancy Pelosi,9 and Con-
gress has started to take action. In December 2018, Ben Sasse (R-NE) introduced
the Malicious Deep Fake Prohibition Act, which would criminalize the creation of
certain types of deepfakes.10 House Foreign Intelligence Committee chairman
Adam Schiff has also begun holding congressional hearings related to deepfakes
and other disruptive AI-based technologies.11

However, as John Villasenor points out, we likely already have legal frameworks
to regulate and prosecute many problematic types of deepfakes.12 While gener-
ative networks are often hyped as having “the gift of imagination,”13 this is not ac-
tually the case. The single most important trait of today’s AI is that it cannot invent:
AI can only imitate the real-world data that it ingests. While subtle, this aspect of
AI directly leads to how we can define, regulate, and—if necessary—prosecute
malicious deepfakes.

One domain of concern is false advertising. Given the availability of deepfake

audio generators for some popular internet personalities such as Joe Rogan, it
would be trivial for a corporation to synthesize and publish a fake endorsement
from one of these hugely influential individuals. However, false advertising is al-
ready forbidden by the Lanham Act: “Any person who . . . uses in commerce any

8
word, term, name, symbol, or device . . . false or misleading representation of fact
. . . shall be liable in a civil action.”14 The endorser’s image or voice was inherently
used in training the deepfake algorithm, and as such, the creator of a deepfake
endorsement could be held civilly liable for misuse under existing law.

Perhaps the most morally abhorrent application of deepfakes is hyperrealistic sim-

ulated child pornography. Again though, this application would require the use of
real images of children at some point in building the generative network algorithm,
and therefore would be prosecutable under the PROTECT Act of 2003, which pro-
hibits the use of images of minors in the creation of simulated child pornography.

Advancing Broader AI Efforts Can Also Fight Deepfakes

The uncomfortable reality is that the best course of action for policymakers may be
to do very little to address deepfakes directly. Existing laws cover many harmful
applications of deepfakes, and banning certain deepfakes would only hamper
research into benevolent technologies employing similar algorithms. The threat
from deepfakes is simply the worst-case end result of the democratization and
proliferation of AI. The cultural and technological environment that created deep-
fakes—not the resultant algorithms themselves—should be the focus of any policy.
To this end, there are four key areas in which we can concentrate our efforts to
combat deepfakes:

Incentivize industry to improve deepfake-detection algorithms. Incentivizing in-

dustry to prevent deepfakes is the one arena in which direct action may be the
best course. All malicious deepfakes falling under US law will be published on
privately owned internet platforms, so detection and removal of deepfakes must
necessarily be done by technology companies. We must require private technol-
ogy companies to implement deepfake safeguards on their platforms, and fine
those who fail to remove deepfakes violating existing laws. While some companies
such as Reddit and Twitter already have voluntarily banned malicious deepfakes,
others such as Facebook and internet pornography giant MindGeek have yet to
take significant action. Fining these companies would provide a monetary incentive
for preventing deepfakes.

Implementing a policy of this sort would require the government to define and reg-
ularly update what is considered to be state-of-the-art deepfakes generation and
detection techniques. The Deepfakes Reports Act, introduced in July 2019 by US
Senator Gary Peters (D-MI),15 is a huge step towards developing such a standard.
This law would require the Department of Homeland Security to produce regular
reports on deepfakes, which in turn could be used to delineate what illicit content
should be detectable by private companies’ filters.

Expand fundamental AI research at government and academic institutions.

While we must significantly expand government-backed research specifically into

9
 deepfakes,16 this should be part of a much broader effort to improve American
AI capabilities. With Russia and China making large-scale government-backed
efforts into AI research, maintaining AI parity is now a national security concern.

However, the unfortunate reality of AI in the United States is that most of the
intellectual and computational resources are held by private industry. If we plan
to keep up in the global AI arms race, the United States must greatly expand re-
search into AI relevant to national security. Creating programs such as DARPA’s Me-
dia Forensics Office, which focuses on research related to fabricated images and
video, is a step in the right direction.17 Beyond expanding research at government
laboratories, we should also increase research grants from agencies such as the
National Science Foundation for projects focusing on national security–relevant AI.

Improve AI capabilities by improving education. Alongside our research capa-

bilities, we must expand our domestic intellectual capital in AI. While the United
States is home to many of the best AI research groups and technical universities, AI
is dominated by non-Americans working or studying at American universities and
corporations. While this is certainly no issue for the advancement of AI as a field,
it is a severe concern for America’s ability to integrate AI into our military and in-
telligence operations due to citizenship requirements for working in these areas. To
maintain AI parity with other great powers, the United States must grow the pool
of American students entering college with sufficient math and computer science
backgrounds to succeed in studying AI.

The inability of American undergraduates to meet the demand for AI engineering

jobs traces back to poor secondary education, particularly in STEM fields. How-
ever, improving domestic AI intellectual capital and readiness need not involve a
comprehensive overhaul of the education system. To address this challenge, it could
be sufficient to simply fund summer programs for a few thousand students to learn
programming and quantitative skills relevant to AI. Such programs could serve the
dual purpose of preparing top students to be competitive in technical fields at the
university level while providing a prestigious summer opportunity for low-income
high school students who would otherwise be unable to afford the often expensive
summer intensives offered at many top universities.

Promote ethics education alongside technical fields. Finally, the threat of deep-
fakes highlights the need to include law and ethics education alongside technical
instruction. All unethical technology begins with an unethical engineer: instead of
simply banning unethical AI, we should address the dearth of ethical concern shown
by many of today’s inventors and entrepreneurs. One approach would be to pro-
vide curriculum development grants to universities who integrate ethics education
into their introductory computer science curricula. By exposing students to the con-
cerns and perils surrounding technology early in their careers, we could help to
embed technology ethics in American tech culture.

10
Cynics may argue that expecting self-regulation of the tech industry is foolhardy—
the prevailing ethos of Silicon Valley is Facebook’s unofficial motto “Move fast
and break things”18—but recent events show that bringing AI ethics into the public
conscience is having an effect. For example, when DeepNude, a program that digi-
tally undressed women, was published earlier this year, GitHub, the largest website
for publishing computer code and now a Microsoft subsidiary, swiftly banned this
app and any related project.19 There still exists much work to be done in this area,
but tech culture in the United States is slowly shifting in a promising direction.

Summary and Conclusion

While deepfakes indeed pose a threat to our privacy and security, this threat is
far from insurmountable. When taking action on deepfakes, our legislators should
eschew rash solutions that will be detrimental to long-term AI advances and instead
pursue policies that address the underlying issues in technology culture and edu-
cation. By incentivizing private industry against deepfakes, promoting AI research,
and improving equity in and the ethics components of domestic STEM education, we
can effectively mitigate deepfakes, strengthen our domestic AI capabilities, and
continue to foster the environment of innovation which gave us such frightening and
fascinating technology in the first place.

1
Samantha Cole , “AI-Assisted Fake Porn Is Here and We’re All F——d,” Vice, December
11, 2017, https://www.vice.com/en_us/article/gydydm/gal-gadot-fake-ai-porn.
2
“Keanu Reeves as Forest [sic] Gump Deepfake—It’s Breathtaking!” YouTube,
https://www.youtube.com/watch?v=cVljNVV5VPw.
3
Ian J. Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sher-
jil Ozair, and Aaron Courville, “Generative Adversarial Networks,” 2014, arXiv:1406.2661.
4
Adam Fisher, Valley of Genius: The Uncensored History of Silicon Valley (As Told by the
Hackers, Founders, and Freaks Who Made It Boom (New York: Twelve, 2018).
5
Phillip Isola, Jun-Yan Zhu, Tinghui Zhou, and Alexei A. Efros, “Image-to-Image Translation
with Conditional Adversarial Networks,” 2016, arXiv:1611.07004; Jun-Yan Zhu, Taesung
Park, Phillip Isola, and Alexei A. Efros, “Unpaired Image-to-Image Translation using Cy-
cle-Consistent Adversarial Networks,” 2017, arXiv:1703.10593.
6
Drew Harwell, “Fake-porn videos are being weaponized to harass and humiliate women:
‘Everybody is a potential target,’” Washington Post, December 30, 2018, https://www.
washingtonpost.com/technology/2018/12/30/fake-porn-videos-are-being-weapon-
ized-harass-humiliate-women-everybody-is-potential-target/?noredirect=on.
7
James Vincent, “Watch Jordan Peele use AI to make Barack Obama deliv-
er a PSA about fake news,” The Verge, April 17, 2018, https://www.theverge.com/
tldr/2018/4/17/17247334/ai-fake-news-video-barack-obama-jordan-peele-buzzfeed.

11
 8
Riana Pfefferkorn, ““Deepfakes’: A New Challenge for Trial Courts,” NWSidebar, June
24, 2019, https://nwsidebar.wsba.org/2019/03/13/deepfakes-a-new-challenge-for-tri-
al-courts; Oscar Schwartz, “You thought fake news was bad? Deep fakes are where truth
goes to die,” The Guardian, November 12, 2018, https://www.theguardian.com/technolo-
gy/2018/nov/12/deep-fakes-fake-news-truth.
9
Robert Chesney, Danielle Citron, and Quinta Jurecic, “About That Pelosi Video: What to
Do About ‘Cheapfakes’ in 2020,” Lawfare, June 2, 2019, https://www.lawfareblog.com/
about-pelosi-video-what-do-about-cheapfakes-2020.
10
Malicious Deep Fake Prohibition Act of 2018, S.3805, introduced December 21, 2018,
115th Congress (2017–2018), https://www.congress.gov/bill/115th-congress/sen-
ate-bill/3805.
11
Elizabeth Culliford, “House Intelligence chief presses social media companies on deep-
fake policies,” Reuters, July 15, 2019, https://www.reuters.com/article/us-usa-elec-
tion-deepfakes/house-intelligence-chief-presses-social-media-companies-on-deep-
fake-policies-idUSKCN1UA2GC; Eric Johnson, ““We’re not ready’ for foreign election
interference in 2020, says Rep. Adam Schiff,” Vox, July 22, 2019, https://www.vox.com/re-
code/2019/7/22/20702196/adam-schiff-deepfakes-nancy-pelosi-google-twitter-face-
book-2020-youtube-kara-swisher-decode-podcast.
12
John Villasenor, “Artificial intelligence, deepfakes, and the uncertain future of
truth,” Brookings Institution, February 14, 2019, https://www.brookings.edu/blog/
techtank/2019/02/14/artificial-intelligence-deepfakes-and-the-uncertain-future-of-truth.
13
Martin Giles, “The GANfather: The man who’s given machines the gift of imagination,” MIT
Technology Review, February 27, 2018, https://www.technologyreview.com/s/610253/
the-ganfather-the-man-whos-given-machines-the-gift-of-imagination.
14
15 U.S.C. § 1125.
15
“Peters, Colleagues Introduce Bipartisan Bill to Respond to Growing Threat of Deep-
fakes,” Homeland Security & Governmental Affairs, Minority Media, July 3, 2019, https://
www.hsgac.senate.gov/media/minority-media/peters-colleagues-introduce-bipartisan-
bill-to-respond-to-growing-threat-of-deepfakes.
16
Villasenor, “Artificial Intelligence.”
17
Matt Turek, “Media Forensics (MediFor), Defense Advanced Research Projects Agency
(DARPA), https://www.darpa.mil/program/media-forensics.
18
Fisher, Valley of Genius.
19
Arwa Mahdawi, “An app using AI to ‘undress’ women offers a terrifying glimpse into the
future, The Guardian, June 29, 2019, https://www.theguardian.com/commentisfree/2019/
jun/29/deepnude-app-week-in-patriarchy-women; Joseph Cox, “GitHub Removed Open
Source Versions of DeepNude,” Vice Motherboard, July 9, 2019, https://www.vice.com/en_
us/article/8xzjpk/github-removed-open-source-versions-of-deepnude-app-deepfakes.

12
Introducing a Modified Good Samaritan Reduced-Price
Mechanism to Combat Hunger and Food Waste
By Wenhao Winston Du, Vanderbilt University

Introduction

The federal government has long recognized the unfortunate truth that hunger and
food waste are pressing issues in the United States. Besides providing and funding
a litany of direct federal assistance programs (e.g., SNAP) to address the prob-
lems, the government has also recognized the importance of promoting the roles of
individuals, corporations, and nonprofits in this common endeavor.

To that effect, under section 2135 of the Tax Reform Act of 1976, Congress put
forth special, enhanced tax deductions for corporate business donations of food
inventory to nonprofit entities.1 Two decades later, President Clinton signed into law
the Bill Emerson Good Samaritan Act of 1996. The law provided strong liability
protection, under a clearly outlined framework, for businesses and individuals do-
nating food inventory to nonprofits serving the needy.2

Together, these pieces of legislation incentivized companies in the food and gro-
cery sector to receive certified tax deductions for donating, instead of writing
down, surplus inventory. The tax-deduction rules were temporarily expanded by
the Katrina Emergency Tax Relief Act (KETRA) in 2006.3 This expansion, which
extended the donation deductions to all taxpayer groups (including farmers and
small businesses), was made permanent as part of the Protecting Americans from
Tax Hikes (PATH) Act of 2015.4 According to Feeding America, a nationwide food
bank network consisting of more than two hundred members, this legislation dras-
tically increased restaurant food donations (jumping 137 percent in 2006) to its
member institutions.5

Under the law (current as of September 2019), for every $1 of fair-market value
foods or produce donated, a company can receive a deduction on its tax burden
of 25 cents. This allowance, which can be rolled over the succeeding five years,
is capped at 15 percent of taxable income in the case of C corporations, or 15
percent of aggregate net income for all other taxpayer classifications.

The Problem

Despite its successes, the federal government can do more to effectively ensure
that unwanted but still edible food ends up on people’s plates instead of in the
landfill. There are still gaps in current law open for improvement.

1. First, donors can only (for good reasons) donate to 501(c)(3) nonprofit entities
exempt under section 501(a).6 The indirect consequence is that food dona-

13
 tions can only benefit communities that are close enough to have a designat-
ed food bank, leaving out less-populated rural communities. In 2018, 12.7
percent of rural Americans were found to be food insecure.7 This has forced
many food banks to deploy limited resources towards mobile food pantries in
order to serve the needy.8
2. Moreover, since such deductions are valid without respect to the type of food
donated, donors are incentivized to donate foodstuffs that are easy to do-
nate. This means processed foods with long shelf lives are received in higher
volume than others. For example, many food banks across the United States
are inundated with cereal grains and junk foods while struggling to pro-
cure healthier fresh produce.9 Feeding America, which conducted a survey of
about 60,000 food pantry and food bank users across the country, reported
in 2014 that fresh fruit and vegetables (55.0 percent), meat (47.1 percent),
and dairy products (40.0 percent) were identified by survey respondents as
the most unavailable desired products.10
3. Yet while demand for these items is high, there is ample evidence that a large
amount of this desired produce goes to waste in supermarkets. From 2011 to
2012, the USDA Economic Research Service conducted a nationwide study on
a sample of supermarkets (consisting of more than 2,900 stores in 45 states).
It found that about 12.6 percent of fruit, 11.6 percent of vegetables, and
22.6 percent of fresh meat were lost. Industry groups cited by the report esti-
mated operational breakdowns to account for 64 percent of this loss in 2012.
This represents almost twice the amount these same industry groups pinned
on theft (35 percent).11 In other words, there is still a supply-and-demand
mismatch when it comes to donation viability.
4. Data collection is another problem. In order to effectively understand this
policy challenge of food waste, policy analysts require a fine-grained under-
standing of what exactly is wasted, and how much. The best measurement,
the Economic Research Service (ERS) Loss-Adjusted Food Availability
data report, is conducted once per decade and is based only on supermar-
ket shrinkage.12 This gives policymakers a very fuzzy view of the food waste
problem. Other numbers, such as a 40-percent-waste statistic included in a
highly cited 2009 paper, are derived from a back-of-the-envelope calcula-
tion of calories lost rather than actual edible calories.13

The Proposed Policy

As a first step toward resolving these issues, Congress should consider the passage
of a Good Samaritan reduced-price mechanism to spur point-of-sale establish-
ments (e.g., restaurants, grocery stores, and wholesalers) to discount foodstuffs sold
to low-income individuals and 501(c)(3) non-profit entities. Specifically, it should do
the following:

1. Allow food establishments and grocery stores to receive the existing en-
hanced deductions for the discounted sale of “apparently wholesome” foods

14
 (as defined under the 1996 Bill Emerson Good Samaritan Act) at prices no
more than 25 percent of fair-market value (as outlined by the 2015 Protect-
ing Americans from Tax Hikes Act) to persons in these two groups:
a. Individuals and households approved to receive federal SNAP
food stamp benefits.
b. 501(c)(3) nonprofit entities that will use the donations for the
purpose of serving individuals and families in need.
2. In order to qualify for this treatment of sales, the retailer must ensure the
following:
a. That it (the retailer) allows for the sale of all food items in ac-
cordance with federal and state law via the federal Electronic
Benefits Transfer (EBT) system.
b. That the annual fair-market value total of such reduced-price
sales does not exceed the annual fair-market value the same re-
tailer made in accepted food donations to a 501(c)(3) nonprofit
organizations.
c. That it cooperates, to a reasonable degree, with federal agen-
cies on data collection and surveys of inventory waste.
3. Qualifying retailers shall be entitled to receive, unless in the case of gross
negligence, full liability protections under all relevant conditions set forth by
the Bill Emerson Good Samaritan Food Donation Act.14
4. An additional $10 million should be allocated annually, beyond the existing
$25 million designated in the 2018 Farm Bill,15 to the USDA ERS to fund an-
nual studies encompassing research of the food distribution chain and food
bank inventories in the United States. In addition, the USDA shall track the
sale and donations made by participating Good Samaritan reduced-price
retailers, and the ERS shall summarize these findings in annual reports to Con-
gress and the general public.
5. An allocation of $25 million should be made annually, for the next five fiscal
years, to inform qualified retailers of the Good Samaritan reduced-price
sale mechanism.

Analysis

The proposed reduced-price mechanism allows qualified businesses to take al-

ready pre-existing tax deductions in a different manner. It lowers barriers to distri-
bution in two ways: First, with up to 50 percent of the fair-market value of surplus
foodstuffs up for grabs, retailers such as grocery stores are incentivized to try to
sell them and are rewarded for innovating and expanding their logistics capabil-
ity. Second, a large percentage of the needy can purchase from a grocery store
instead of finding a food pantry.

As a consumer choice–based approach, this discount program will offer recipients

access to better nutritional balance. On the demand side, it allows needy end con-
sumers, as well as the intermediary nonprofits, to exert their preferences on what

15
 they receive. On the supply side, it alleviates friction due to economies of scale.
Small grocers that are currently not able to carry certain fresh produce (due to the
risks in filling a large minimum supply order), may find the accounting to be a bit
more in their favor. This spillover effect benefits all customers.

It is not expected that retailers would significantly decrease current donations of

their surplus food under the passage of this policy. This is because the proposed
policy only allows for the sale of fully “wholesome” food, while baseline require-
ments for donations to nonprofits remain laxer. Moreover, as a last-line safeguard
against “switching” behavior, section (2) directly ties total sales under the re-
duced-price mechanism to the value of straightforward donations.

The budget cost of this policy will likely be minimal and may actually be net
positive for federal revenues. First, because retailers will need to accept EBT, this
means federal SNAP dollars are stretched to provide more value to recipients.
Secondly, because the reduced-price sales are treated as revenue, participating
retailers may wind up with a higher taxable net income. This is especially relevant
as grocery stores and restaurants are notoriously low-margin businesses, with av-
erage industry margins estimated to be 1.5 percent in 2016.16 Finally, according
to the Congressional Budget Office, the 2016 PATH Act provision that expanded
the cap on charitable food deductions from 10 to 15 percent of income amounted
to only $2.2 billion over ten years.17 This policy should cost even less, as it merely
attempts to increase the ability of retailers to receive the deduction rather than
increasing the cap any further.

Differences with Similar Proposals

An earlier iteration of the proposed Good Samaritan reduced-price mechanism

was introduced by Representative Chellie Pingree of Maine in the Food Recovery
Act of 2017.18 Pingree’s legislation, which failed to gain traction, differs with this
paper’s proposed policy in two major ways:

1. It does not make applicable the enhanced deductions (as provided by the
2015 PATH legislation) to reduced-price sales.
2. It defines the Good Samaritan reduced price as “an amount not greater
than the cost of handling, administering, and distributing” the food as op-
posed to this paper’s proposal of using 25 percent of the food product’s
fair-market value.

Without the additional tax deduction qualification, this bill will alleviate only the
burdens of nonprofits. It will not be as effective at drawing in private retailers to
participate. In addition, by tying the price to operational cost rather than the actu-
al value of the product, the Pingree bill inadvertently disallows a key incentive for
innovation and cost reduction.

16
Expected Result and Evaluation

This policy sets out to do three things:

1. Decrease food waste by restaurants, grocery stores, and other retail-

ers by increasing the proportion of unsold inventory that is diverted to
needy end consumers.
2. Increase the nutritional diversity of foods offered by nonprofits and giv-
en or purchased by low-income and needy Americans.
3. Making point-of-sale businesses aware of the Good Samaritan re-
duced-price mechanism in order to accomplish the above two points.

Through Section (5)’s study of food waste in the United States, policymakers will
have the measurements to evaluate the impact of the proposal’s first objective.
Next, by tracking food bank inventories as well as reduced-price purchases made
through EBT, policymakers can evaluate the nutritional impact post-policy enact-
ment. Finally, to ensure this policy is deployed effectively, marketing to qualified
businesses is required. The popularity of this option can be measured in terms of
the number of businesses enrolled as well as the amount in sales of reduced-price
food sold.

Conclusion

Food waste in the United States is a major problem. Every year, up to $218 bil-
lion a year, or 1.3 percent of GDP, is spent on growing, processing, transporting,
and disposing food never eaten. 19 This waste occupies 21 percent of US landfill
content and releases greenhouse gases equivalent to that of 37 million cars.20
Yet at the same time, hunger persists as well, with 11.1 percent of US households
deemed food insecure in 2018.21 Through the expansion of donor incentives and
food waste research funding, this policy will lay the groundwork for tackling both
issues together.

1
Tax Reform Act of 1976, Pub. L. No. 94–955, § 2135, 90 Stat. 1520 (1976), 410, https://
www.govinfo.gov/content/pkg/STATUTE-90/pdf/STATUTE-90-Pg1520.pdf.
2
Bill Emerson Good Samaritan Food Donation Act, Pub. L. No. 104–210, 110 Stat. 3011
(1996), https://www.congress.gov/104/plaws/publ210/PLAW-104publ210.pdf.
3
Katrina Emergency Tax Relief Act of 2005, Pub. L. No. 109–73, 119 Stat. 2016 (2005),
https://www.congress.gov/109/plaws/publ73/PLAW-109publ73.pdf.
4
This was part of negotiated 2016 federal omnibus spending bill Consolidated Appropria-
tions Act of 2016, Pub. L. No. 114–113, § Division Q: Protecting Americans From Tax Hikes
Act of 2015, 129 Stat. 3040 (2015), https://www.congress.gov/114/plaws/publ113/
PLAW-114publ113.pdf.

17
 5
“Feeding America Urges Swift Vote On Expired Tax Provisions,” Feeding America, June
8, 2012, https://web.archive.org/web/20170913220301/http://www.feedingamerica.
org/about-us/press-room/feeding-america-urges-swift-vote-on-expired-tax-provisions.
html.
6
Internal Revenue Code of 1986, 26 USC § 170(e)(3)(C) (1986), http://uscode.house.gov/
quicksearch/get.plx?title=26&section=170.
7
Alisha Coleman-Jensen, Matthew P. Rabbitt, Christian A. Gregory, and Anita Singh, House-
hold Food Security in the United States in 2018, ERR-270, US Department of Agriculture, Eco-
nomic Research Service, 2019, https://www.ers.usda.gov/webdocs/publications/94849/
err-270.pdf?v=963.1.
8
Elaine S. Povich, “Mobile Food Banks Roll to Isolated, Rural Poor,” Pew Charitable Trusts,
Stateline, July 16, 2018, https://www.pewtrusts.org/en/research-and-analysis/blogs/
stateline/2018/07/16/mobile-food-banks-roll-to-isolated-rural-poor.
9
Small food banks struggle to source fresh produce, meat, and dairy, and find plac-
es to store it. Caitlin Dewey, “Charities are realizing that poor people also deserve
healthy food,” Washington Post, May 12, 2017, https://www.washingtonpost.com/
news/wonk/wp/2017/05/12/charities-are-finally-realizing-that-poor-people-also-de-
serve-healthy-food/.
10
Nancy S. Weinfield, Gregory Mills, Christine Borger, Maeve Gearing, Theodore Maca-
luso, Jill Montaquila, and Sheila Zedlewski, “Hunger in America 2014: National Report,”
prepared by Westat and the Urban Institute for Feeding America, 30, http://help.feedin-
gamerica.org/HungerInAmerica/hunger-in-america-2014-full-report.pdf.
11
Jean C. Buzby, Jeanine T. Bentley, Beth Padera, Jennifer Campuzano, and Cara Ammon,
Updated Supermarket Shrink Estimates for Fresh Foods and Their Implications for ERS Loss-Ad-
justed Food Availability Data, EIB-155, US Department of Agriculture, Economic Research
Service, June 2016, https://www.ers.usda.gov/webdocs/publications/44100/eib-155.
pdf?v=0.
12
Buzby et al, Updated Supermarket Shrink Estimates. See also Jean C. Buzby, Jeanine T.
Bentley, Beth Padera, Jennifer Campuzano, and Cara Ammon, “Estimated Fresh Produce
Shrink and Food Loss in U.S. Supermarkets,” Agriculture 5, no. 3 (August 4, 2015): 626–46,
https://doi.org/10.3390/agriculture5030626.
13
Kevin D. Hall, Juen Guo, Michael Dore, Carson C. Chow, “The Progressive Increase of
Food Waste in America and Its Environmental Impact,” PLOS ONE 4, no. 11 (2009):e7940,
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0007940.
14
42 USC 1791.
15
“Agriculture Improvement Act of 2018,” Pub. L. No. 115–334, 132 Stat. 4490 (2018),
https://www.congress.gov/115/plaws/publ334/PLAW-115publ334.pdf.
16
Madeline Hurley, IBISWorld Industry Report 44511: Supermarkets & Grocery Stores in
the US, IBISWorld, May 2016, https://www.scribd.com/document/325830175/44511-Su-
permarkets-Grocery-Stores-in-the-US-Industry-Report.

18
17
Note: the CBO scored these specific provisions when they were part of a standalone bill.
See Nathaniel Frentz, “Congressional Budget Office Estimate for H.R. 644,” Congressional
Budget Office, February 4, 2015, https://www.cbo.gov/sites/default/files/114th-con-
gress-2015-2016/costestimate/hr644_0.pdf; America Gives More Act of 2015, H.R. 644,
114th Cong. (2015), introduced by Rep. Tom Reed, https://www.gop.gov/bill/h-r-644-
fighting-hunger-incentive-act-of-2015.
18
Food Recovery Act of 2017, H.R. 3444, 115th Cong. (2017), introduced by Rep. Chellie
Pingree, https://www.congress.gov/bill/115th-congress/house-bill/3444.
19
“A Roadmap to Reduce US Food Waste by 20 Percent”, ReFed, March 2016. https://
www.refed.com/downloads/ReFED_Report_2016.pdf.
20
Dana Gunders and Jonathan Bloom. “Wasted: How America is losing up to 40 percent of
its food from farm to fork to landfill.” Natural Resources Defense Council, 2017. 4. https://
www.nrdc.org/sites/default/files/wasted-2017-report.pdf.

21
Alisha Coleman-Jensen, Christian A. Gregory, and Matthew P. Rabbitt, “Food Security
Status of U.S. Households in 2018,” US Department of Agriculture, September 2019.

19
 Bad Seeds: A Case against Federal Farm Subsidies
By Nicholas G. Sileo, Princeton University
At the founding of the United States, Thomas Jefferson is said to have imagined
a nation of gentleman farmers tending to the land of the nation through their
work and civic involvement. Writing to George Washington, he reflected, “Ag-
riculture . . . is our wisest pursuit, because it will in the end contribute most to
real wealth, good morals & happiness.”1 But the system of corporate welfare
that has grown a part of American agriculture is much closer to “government
. . . wasting the labours of the people, under the pretence of taking care of them”2
than any sort of Jeffersonian democracy.

Farm subsidies have ballooned, becoming a major government expense. In 2018

alone, farmers received $18 billion in direct farm subsidies.3 To justify this sticker
price, the American public would need a cornucopia of public benefits, but the
fruits of this labor appear lacking. Current farm subsidy programs increase the
price consumers pay for items on the shelf; distort market incentives for farmers,
leading to inefficient production; concentrate wealth through the means of gov-
ernment redistribution; pose a worsened risk to environmental impact; aggregate
international trade relations; and are subject to bureaucratic waste, fraud, and
abuse.4

There is a bushel of reasons to rethink farm subsidies, but the political incentives
that have led to the currently bloated system are the very same that will make
reform difficult—subsidy benefits are concentrated whereas costs are dispersed.
Billions of dollars distributed evenly to farmers, composing only 1.3 percent of
the American workforce,5 would be motivation enough for the group to focus on
lobbying over tending fields. Yet, much like the corn syrup being produced on the
taxpayers’ dime, subsidies are highly concentrated. The top 20 percent of agricul-
tural subsidy recipients receive over 90 percent of all subsidies; the top 10 percent
still maintain over three-fourths of all subsidies; and the top 1 percent alone rake
in over a quarter of subsidies by themselves.6 With more than six thousand entities
receiving upwards of $1 million in farm subsidies, there is plenty of reason for
interest groups to keep farm bills alive.7 Meanwhile, the costs for these programs
are widely dispersed across taxpayers, making it a relatively minor political issue
for most voters. As American Farm Bureau chief economist Bob Young points out,
detailing costs of the programs (using the extremely conservative estimate of $7
billion per year in farm subsidies), “That works out to $23 per capita on an annual
basis, six cents per day, or two cents per meal.”8 While not enough of a burden to
fire up the electorate, hiding the costs of subsidies across the populace does not
change the fact the it costs quite a bit more than peanuts.

Given their current state, agricultural subsidies need much more than some light
pruning. Crop insurance, Agriculture Risk Coverage (ARC), and Price Loss Coverage
(PLC) all should be cut to allow a free agricultural market to once again grow back

20
into place. Smaller programs should also be evaluated to determine what tangi-
ble public benefit they offer, and likely trimmed or cut as well. Putting an end to
unneeded distortion in the agricultural market will help honest farmers, consumers,
and taxpayers alike.

Data Analysis

Proponents of agricultural subsidies argue that the programs are designed to sta-
bilize commodity markets, ensure security in food production, and aid impover-
ished farmers and communities, among a variety of other benefits. However, multi-
ple studies have failed to find any tangible benefits from these programs.9 In fact,
a study by the Kansas City Federal Reserve Bank found that counties receiving the
highest per-capita payment had population and employment growth both below
average.10 While these subsidy programs are massive, the majority of American
agriculture functions perfectly well already, with over two-thirds of American farm-
ers producing without government support.11

What these subsidies have been proven to do is distort market incentives. Accord-
ing to the US Department of Agriculture (USDA), sales for the most subsidized crops
grew only 14 percent from 1980 to 2005, whereas nonsupported foodstuff like
fruits, vegetables, and nuts increased by over 186 percent during that same peri-
od.12 Additionally, since most agricultural subsides are production based, subsidies
create an undue incentive for negative environmental impact by promoting over-
use of fertilizers and pesticides.13 While the subsidies incentivize overutilization of
some lands, in 2017 alone $1.8 billion was paid out so that farmers would not use
their land at all.14

These are not funds going to help impoverished farmers that barely get by. In
2016 the median income of all farm households was 29 percent higher than the
national average, and mean income was 42 percent higher.15 Among farm house-
holds, subsidies further advantage those who are already flourishing in the private
sector. USDA data highlights that, of farms with revenues below $100,000, only
one-third receive subsidies, whereas three-fourths of those above that figure do.16
With over $626 million being paid out in farm subsidies to urban areas between
2015 to 201717 and over 10 percent of those on the Forbes 400 list of wealthiest
Americans receiving subsidies18 it’s a stretch to claim these programs are helping
struggling farmers.

By all accounts, the agricultural industry is doing very well. Between 2004 to 2006,
the collective net worth of American farmers grew by an average $90 billion each
year, making the debt-to asset ratio for farmers the lowest it’s been in nearly half
a century.19 Farms are not just doing well by their own standards but also as com-
pared with other sectors in the economy. In 2017 the bankruptcy rate was only
2.4 per 10,000 farms versus the average bankruptcy rate of 8 per 10,000 for
all businesses.20

21
 It’s worth noting too that these subsidies bring a host of other problems with them.
There are serious conflicts of interest in restricting these programs, with thirty-two
current members of Congress receiving farm subsidies.21 The bureaucracy that ad-
ministers the programs is rampant with waste, including embarrassing mistakes like
admitting to $3.7 billion in “improper payments” in 2004 and $1.1 billion paid to
deceased farmers in 2011.22 This is all while protectionist agricultural tariffs and
quotas cost Americans about $10 billion more at the store each year,23 and while
studies by the Congressional Budget Office find that removing these farm subsidies
and trade barriers would bolster the economy.24

A departure from the status quo could cause concern, especially for something
as vital as the food supply, but the United States would not be the first nation to
end these subsidies. In the 1980s, New Zealand cut its agricultural subsidies to the
bone so that today Kiwi farmers only receive 3 percent of their income from the
government as compared to 16 percent in the United States.25 Following the cuts,
not only was the agricultural industry sustained but New Zealand saw productivity,
earnings, and output rise.26 By Kiwi farmers’ own account, according to the Fed-
erated Farmers of New Zealand, that nation’s example of nixing these programs
“thoroughly debunked the myth that the farming sector cannot prosper without
government subsidies.”27

Policy Recommendation

The simple answer to the complex web of problems federal farm subsidies create
is to cut the programs. The three largest farm subsidy programs (crop insurance,
ARC, PLC) are the natural place to start given that they make up of a vast segment
of total expenditures.28 Those three programs are designed to mitigate risk for
farmers, but instead they distort incentives of crop choice and land management,
with over 70 percent going to only to three crops of corn, soybeans, and wheat,
while running up enormous expenses for taxpayers. 29

Various conservation programs, such as those that pay farmers not to produce,
should be cut given they spur nonoptimal usage of land. Instead, long-term pro-
duction incentives will be able to drive proper land use. Other programs such as
disaster aid and research and development spending should be further evaluated
for their effective benefits. While these programs do not have the same mar-
ket-distorting effects, if their public benefit is minor then they should likewise be cut.

Protectionist foodstuff tariffs should be reduced to encourage competition within

the market and lower consumer prices. This would further allow for specialization
of agricultural production and provide the stability to the food supply that many
of these programs claim to seek.

22
Methods for Evaluation

Restoring proper markets to the agricultural industry is likely to have a variety

of benefits for both farmers and consumers. While some of these benefits will be
difficult to immediately quantify (such as proper R&D incentives, lessened strain on
international trade agreements, and dietary changes as a result of removing price
distortions) others will be more readily observable using government and market
data. Four methods that can be used to evaluate the effectiveness of cutting farm
subsidies are: the lessening of the tax burden, total agricultural production, con-
sumer food prices, and farmer welfare.

Every dollar not spent on agricultural subsidies is a dollar less taken from the tax-
payer. Therefore, the direct dollar amount cut from these subsidy programs can be
counted as a positive benefit towards the policy. It is worth noting that lessening the
tax burden frees up citizens to either further consume or save/invest in the econo-
my which will have further benefits.

Agricultural production can be tracked in terms of GDP. While it is likely to observe

some temporary fluctuations in production and between different products as sub-
sidies and price controls are lifted, long-term aggregate production trends can be
used to track the effectiveness of the policy. Additionally, the prices of foodstuffs
can be tracked to evaluate consumer surplus. A GDP deflator approach for food-
at-home purchases should be used to allow for consumer substitution as opposed
to a more rigid CPI basket-of-goods approach that may over- or understate the
impact of the subsidies.

Finally, the cut in subsidies is not designed to hinder farmers but to work to their
long-term benefit. Average pretransfer income of farmers should be tracked to
see how the removal of the subsidies effects their welfare. Given the concentration
of the subsidies to major farming corporations and the distorting effect they have
on the market, it is likely, with increased efficiency, that average farmers will see
an increase in their income as well.

Recommendation

Modern agricultural subsidies in the United States were first introduced in the
1930s along with other New Deal programs.30 Like many New Deal programs, this
policy spurred poorly aligned incentives and has grown to become a massive but
entrenched drag on the American taxpayer and private enterprise. The removal
of these subsidy programs would make America one of the world’s few developed
nations that allow for the free market to flourish within agricultural production. This
should allow for lower prices on shelves, more dynamic and fast-growing food
production, and more prosperous farmers, all while reducing federal expenditures
by billions annually. America’s current cash crop is no longer tobacco or cotton but
taxpayer subsidies. If these programs are cut, once again the nation’s agricultural
sector can get back to doing what it does best—growing.

23
 1
Thomas Jefferson, Letter to George Washington, Thomas Jefferson Foundation, August 14,
1787, http://tjrs.monticello.org/letter/98.
2
Thomas Jefferson, Letter to Thomas Cooper, Thomas Jefferson Foundation, November 29,
1802, https://www.monticello.org/site/research-and-collections/wasting-labours-peo-
ple-quotation.
3
US Department of Agriculture, “The United States Farm Subsidy Information,” Environ-
mental Working Group, n.d., https://farm.ewg.org/region.php?fips=00000&progcode=-
total&yr=2018.
4
Chris Edwards, “Agricultural Subsidies,” Downsizing the Federal Government, April 16, 2018.
5
Kimberly Amadeo, “Farm Subsidies with Pros, Cons, and Impact,” The Balance, July 4,
2019.
6
United States Department of Agriculture, “Commodity subsidies in the United States to-
taled $217.5 billion from 1995-2019,” Environmental Working Group.
7
Adam Andrzejewski, “Mapping The U.S. Farm Subsidy $1M Club,” Forbes, August 18,
2018.
8
Quoted in Daniel Griswold, “Should the United States Cut Its Farm Subsidies?” Cato Insti-
tute, April 27, 2007.
9
Daniel A. Sumner, “Agricultural Subsidy Programs,” The Library of Economics and Liberty,
n.d., https://www.econlib.org/library/Enc/AgriculturalSubsidyPrograms.html.
10
Griswold “Cut Farm Subsidies?”
11
Griswold “Cut Farm Subsidies?”
12
Griswold “Cut Farm Subsidies?”
13
Griswold “Cut Farm Subsidies?”
14
Andrzejewski, “Farm Subsidy $1M Club.”
15
Edwards, “Agricultural Subsidies.”
16
Edwards, “Agricultural Subsidies.”
17
Andrzejewski, “Farm Subsidy $1M Club.”
18
Amadeo, “Farm Subsidies.”
19
Griswold, “Cut Farm Subsidies?”
20
Edwards, “Agricultural Subsidies.”
21
Edwards, “Agricultural Subsidies.”
22
Andrzejewski, “Farm Subsidy $1M Club.”
23
Griswold, “Cut Farm Subsidies?”

24
24
Edwards, “Agricultural Subsidies.”
25
Griswold, “Cut Farm Subsidies?”
26
Edwards, “Agricultural Subsidies.”
27
Edwards, “Agricultural Subsidies.”
28
Edwards, “Agricultural Subsidies.”
29
Edwards, “Agricultural Subsidies.”

30 Sumner, “Agricultural Subsidy Programs.”

25
 A Local Approach to the Opioid Epidemic
By Kaila Webb, Wellesley College

Context

In 2017, the Trump administration’s Commission on Combating Drug Addiction and

the Opioid Crisis published its final report. It included two notable conclusions: fed-
eral funding must be streamlined and organized to make it accessible to states,1
and addiction is heavily influenced by environmental factors.2 This has resulted in
an opioid crisis that mimics cancer—in that it’s really a myriad of different soci-
etal failures that end with a common symptom and diagnosis. My research has
shown that states with similar opioid overdose rates achieve these dismal figures
in different ways. Massachusetts’s overdoses occur largely in a window of wealthy
communities with median incomes of $100,000 to $125,000.3 This starkly contrasts
with what many assume to be the typical epidemic community profile,4 even though
Massachusetts has the seventh highest number of opioid overdoses per state.5 So
how do wealthy communities in New England and poorer communities in the Mid-
west come to have similar overdose statistics? It’s clear that we researchers do
not currently have enough data to reliably determine the root of the issue in each
community, and thus cannot craft conventional mandate-based policy to address
this crisis.

The majority of policy makers do not recognize that each community is arriving
at the epidemic stage due to different environmental causes. Even the aforemen-
tioned final report focused much of its recommendation on promoting Prescription
Drug Monitoring Programs (PDMPs), which have not been shown to statistically
improve overdose or opioid consumption rates.6 While they do improve Schedule
II drug use rates (such as fentanyl and oxycodone),7 they do not affect Schedule I
drugs like heroin. Some studies have suggested that PDMPs are simply forcing users
away from medical opiates to street drugs with the same fatal results.8

These state-run programs receive a substantial portion of the $28 million in funds
from the Center for Disease Control’s Overdose Protection in States branch.9 The
branch is formed of three programs: Prescription Drug Overdose: Prevention for
States (PfS); the Data-Driven Prevention Initiative (DDPI); and the Enhanced State
Opioid Overdose Surveillance (ESOOS). The first, PfS, specifically seeks to maxi-
mize PDMPs and funds such programs in twenty-nine states. The ESOOS provides
financial support for states to produce frequently updated data on the epidemic,
and support in sharing this data across state lines. The Center for Disease Control
has also recently announced that over the next three years it will be providing
over $900 million to state and local governments to track opioid overdose data.10
While this provides more local government support than ESOOS does, it does not
provide a policy solution to solving the opioid epidemic, it merely monitors it.

26
The Policy

I propose that all funds currently used by PfS to promote PDMPs be redirected into
block grants for innovative local governments, under the condition that they report
community statistics to their states and show improvement over time. State-level
politicians and university professors were usually surprised by that my research
found wealthier communities to have a higher overdose rate than neighboring
poorer communities. But, a quick dive into local community websites showed that
higher academia was among the last to identify communities struggling with the
opioid epidemic. Locals noticed it first, for obvious reasons.

For example, the small Massachusetts town of Billerica was identified as a hot spot
for high per-capita opioid overdoses in the state. Billerica town officials didn’t
need to be told they were a hot spot—they knew it from living there. In 2017
(a year before I began my research), they held raffle events to educate seniors
on how to keep their prescription medications safe. The first item on the agenda?
“Hear from an Addict on how he ripped off his grandparents.”11 Billerica already
knew they had a higher-than-usual population of elderly resident, without check-
ing census statistics or setting up a meeting with local politicians. They pulled in
an “enterprise bank manager” to speak with those in attendance on how to close
and check their bank accounts. Most important, like any local event, there was a
stronger degree of trust in learning this advice from community leaders rather than
from researchers, state officials, or pamphlets. Unlike larger federal programs, this
influential event cost a total of $50, simply as a raffle prize to entice attendees.
With far less in resources, a local event was able to provide community-specific
education before quarterly state statistics identified the issue.

Locality also matters when determining how to stop drugs from entering a commu-
nity. Areas near the Mexican border face different supply-chain challenges than a
wealthy New England town does. In one place it might be more effective to combat
gang presence, which facilitates heroin access. In another it might be worthwhile to
focus more heavily on doctor shopping and pain-management protocols. In most
states, communities within one hundred miles of one another will have these dif-
ferences in supply chains, to varying degrees. While it is useful to facilitate data
exchange at the state level to help communities understand these supply chains,
giving local governments the power to enact and craft addiction prevention and
education policies puts money where the most knowledge is.

With regard to solutions targeting opioid supply and opioid abuse, locals have
the most useful ideas for how to help their own communities. By providing grants
through their state governments, citizens and local nonprofits can focus their efforts
on improving a single community hit hardest by the opioid crisis. If conventional,
large-scale (state and federal) programs were effective, then we would have seen
drastic improvement over the last two decades. Since we haven’t, we now need to
allow communities to recognize their own needs and provide their own solutions.

27
 Oversight

Strong state oversight should be used to protect these federal funds from abuse.
Rather than offering a blank check, these block grants would be re-approved
annually by state governments under the condition of total data transparency and
success within a year and a half. Such programs will need to electronically deliver
biannual reports containing the number of attendees to local programs, dates of
program activities, and invoices for any program costs. These reports will be made
immediately viewable by the public via an online portal. Programs may only be
started in communities that have per-capita opioid overdoses above the state av-
erage, and they must show a reduction in that overdose rate within the first year
and a half to maintain funding. After their initial success period, further reductions
must be shown every two years afterward, and transparency reports must continue.
States should individually determine what improvement to the per-capita overdose
rate can be considered a success.

Any states receiving these grants must then summarize these transparency reports
by success rates, program descriptions, and community statistics. Furthermore, any
state accepting these grants must maintain overdose statistics by city, and they
must be made available to the public digitally. Without them states cannot ac-
curately determine whether a community is above the state average for opioid
overdoses per capita, and thus whether or not they are eligible for funding. These
numbers must be made publicly accessible so that local community leaders can
independently determine a proper course of action. Citizens and nonprofits who
wish to help should not have to work through the bureaucratic process simply to
ascertain which areas are suffering.

The ESOOS program already provides funding to thirty-three states so that they
can provide prompt, up-to-date data on the opioid crisis.12 This application is
competitive, and my proposal should be tied to this existent policy. If states are
already receiving funding to develop their fact-finding systems on this issue, then it
makes most sense to allow those states to apply for block grants that function off
of that data.

Possible Results

My research also found that, at least in Massachusetts, wealthy communities with
overdose hot spots were surrounded by lower-income communities with slightly low-
er per-capita overdose rates.13 This suggests that certain communities may serve
as hubs of opioid abuse, leaking into surrounding areas. If states provide funds to
spot-treat these communities with innovative approaches, we may see surrounding
communities benefit without additional concerted efforts.

Over time and if successful, this program could be expanded by increasing the
number of grants available to ESOOS-participating states. Successful programs

28
could become models in their states, and suggested to counties that have similar
community profiles. A small town in western Massachusetts has far more in common
with a small town in the Midwest than it does with the Boston metropolitan area. By
encouraging creative solutions in communities with willing leaders and documenting
how they built their successes, they could be artificially replicated in less organical-
ly supported communities. Communities would no longer have to wonder whether
mental health support, methadone clinics, or drug busts would be most effective.
The federal government would have a register of situations in which each solution
was or was not successful.

If enacted, this policy would take the guess work out of the opioid crisis. It would
effectively outsource research to civilians willing to help, without requiring that oth-
er communities wait for their state government to figure out an effective solution to
the problem. It also would allow success to be defined as reduction in harm (fewer
overdose deaths) rather than a reduction in medical prescriptions (less Schedule II
drug use). Doctors currently describe their narcotic prescription process with words
like “blame” or “guilt,” due in no small part to its regulation in an effort to reduce
opioid addiction.14 By seating future addiction prevention programs in voluntary,
community-based approaches, we absolve doctors to some degree of this guilt. If
a group chooses to install Narcan kits in common overdose locations,15 it prevents
overdoses before addicts enter the hospital. If communities successfully increase
mental health support for those at risk of addiction, it prevents doctor shopping
from ever occurring.

Current federal and state regulations that attempt to combat the opioid crisis take
a top-down approach. Criminalizing drugs made people turn to illegal supply
chains. Requiring a prescription for Naloxone makes the drug difficult to acquire
during a life-threatening event. Regulation is a choke hold on innovation, and all
standard, homogeneous approaches to the opioid crisis have failed. Given the
chance, local governments can and do care for their citizens. Giving them more
autonomy on such a critical issue at worst does nothing. At best, it saves lives.

1
“The Commission urges Congress . . . to block grant federal funding for opioid-related
and SUD-related activities to the states, where the battle is happening every day. . . . The
Commission believes that ONDCP [Office of National Drug Control Policy] must establish a
coordinated system for tracking federally-funded initiatives.” Chris Christie, Charlie Baker,
Roy Cooper, Patrick J. Kennedy, Bertha Madras, and Pam Bondi, report of the President’s
Commission on Combating Drug Addiction and the Opioid Crisis, 2017, https://www.white-
house.gov/sites/whitehouse.gov/files/images/Final_Report_Draft_11-1-2017.pdf.

2
“Powerful environmental factors can shape the course of heroin addiction… these results
illustrate that powerful environmental factors may influence the course of heroin addiction.”
Chris Christie et al., Combating Drug Addiction.

29
 3
Kaila Webb, “The Opioid Crisis’ Wealth Window: There’s a Network of Overdoses on the
East Coast,” Pioneer Institute, August 17, 2018. https://pioneerinstitute.org/blog/the-opi-
oid-crisis-wealth-window-theres-a-network-of-overdoses-on-the-east-coast.

4
Katherine M. Keyes, Magdalena Cerdá, Joanne E. Brady, Jennifer R. Havens, and Sandro
Galea, “Understanding the Rural–Urban Differences in Nonmedical Prescription Opioid Use
and Abuse in the United States,” American Journal of Public Health 104, no. 2 (February
2014):e52–59.

5
National Institute on Drug Abuse, “Opioid Summaries by State,” revised May 2019,
https://www.drugabuse.gov/drugs-abuse/opioids/opioid-summaries-by-state.

6
Gil Kerlikowske, Christopher M. Jones, Regina M. Labelle, and Timothy P. Condon, “Pre-
scription Drug Monitoring Programs—Lack of Effectiveness or a Call to Action?,” Pain Med-
icine 12, no. 5 (May 2011):687–89.

7
Kerlikowske et al, “Prescription Drug Monitoring Programs.”

8
Richard C. Dart, Hilary L. Surratt, Theodore J. Cicero, Mark W. Parrino, S. Geoff Severt-
son, Becki Bucher-Bartelson, and Jody L. Green, “Trends in Opioid Analgesic Abuse and
Mortality in the United States,” New England Journal of Medicine 372, no. 3 (January
15, 2015):241–48; Theodore J. Cicero, Matthew S. Ellis, and Hilary L. Surratt, “Effect of
Abuse-Deterrent Formulation of OxyContin,” New England Journal of Medicine 367, no. 2
(July 12, 2012):187–89.

9
Mark R. Jones, Matthew B. Novitch, Syena Sarrafpour, Ken P. Ehrhardt, Benjamin B. Scott,
Vwaire Orhurhu, Omar Viswanath, Alan D. Kaye, Jatinder Gill, and Thomas T. Simopoulos,
“Government Legislation in Response to the Opioid Epidemic,” Current Pain and Headache
Reports 23, no. 6 (May 1, 2019):40.

10
“Trump Administration Announcing Nearly $2B in Opioid Grants.” Associated Press, Sep-
tember 9, 2019, https://apnews.com/aff5fa628c6146dab958f21ad868ca7a.

11
“Opioids & How NOT to Be a Victim,” Town of Billerica event listing, May 20, 2017,
https://www.town.billerica.ma.us/Calendar.aspx?EID=515.

12
“Enhanced State Opioid Overdose Surveillance,” Centers for Disease Control and Pre-
vention, page last reviewed July 16, 2019, https://www.cdc.gov/drugoverdose/foa/state-
opioid-mm.html.

13
Kaila Webb, “A Wealth of Data: A Map of the Massachusetts Opioid Epidemic,” Pioneer
Institute, August 17, 2018. https://pioneerinstitute.org/blog/a-wealth-of-data-a-map-of-
the-massachusetts-opioid-epidemic.

14
Angela Y. Esquibel, and Jeffrey Borkan, “Doctors and Patients in Pain: Conflict and Col-
laboration in Opioid Prescription in Primary Care,” PAIN 155, no. 12 (December 2014):
2575–82.

15
Glenn Forbes, “City of Green To Install NARCAN Kits In Hotels,” Ideastream, May 28,
2019, https://www.ideastream.org/news/city-of-green-to-install-narcan-kits-in-hotels.

30
NOTES

31
 #HISPBC

Hoover Institution, Stanford University

434 Galvez Mall
Stanford, CA 94305-6003
650-723-1754

32