Cigarette Taxes and Food Stamps

"Behavioral Responses to Taxation: Cigarette Taxes and Food Stamp Take-Up," by Kyle Rozema and Nicholas Ziebarth. March 2015. SSRN #2586410.

In recent issues of Regulation, authors have presented evidence that the consumption of cigarettes by those who continue to smoke—both adults and teenagers—is now unaffected by cigarette tax increases ("Cigarette Taxes and Smoking," Winter 2014–2015 and "Working Papers," Winter 2015–2016). That is, current smokers have such strong preferences for smoking that additional cigarette tax increases do not reduce smoking further and cannot be justified by health benefits.

For low-income smokers, the price effects of the tax increases are large. From 2000 to 2010, cigarette prices doubled from $3 to $6 a pack. A full-time minimum-wage worker's expenditure on a pack a day has increased from 4% to over 10% of a day's income.

How have they managed? It appears that many have turned to food stamps. During the same time period, the share of the population receiving food stamps has more than doubled, from 6% to 15%. And low-income smoking households are 50% more likely to enroll in food stamps relative to low income non-smoking households.

Are these stylized facts causally related? The authors examine whether exogenous variation in the imposition of state cigarette tax increases changes food stamp enrollment. They conclude that each dollar of tax increase raises the probability of an eligible (low-income) non-enrolled smoking household signing up for food stamps by between 2 and 3 percentage points from a baseline probability of 25%.

President Obama proposed increasing federal cigarette taxes by about $1 in 2013. The estimated revenue increase was $8 billion. The authors use their work to estimate that, because of the higher tax, food stamp enrollment would increase by about 400,000 people, costing the program $500 million.

E-cigarettes and Adolescent Smoking

"How do Electronic Cigarettes Affect Adolescent Smoking," by Abigail S. Friedman. April 2015. Available at http://scholar.harvard.edu/files/afriedman/files/how_do_electronic_ciga….

In May the U.S. Food and Drug Administration issued regulations on e-cigarettes under authority granted by the Tobacco Control Act of 2009. In response, many policy analysts, including my Cato colleague, Walter Olson, have argued that requiring e-cigarettes to navigate an FDA-approval process is a bad idea because e-cigarettes are less harmful than traditional cigarettes. ("The FDA's Slow-Motion Ban of e-Cigarettes," Ricochet, May 9, 2016.) To be sure, the FDA, in its final rule, acknowledged that many believe that "the aerosol is completely harmless or significantly less harmful than tobacco smoke from combusted tobacco products." And the "FDA recognizes that the aerosol that is exhaled by users of some e-cigarettes and similar electronic apparatus may not pose as much harm as smoke emitted from combusted tobacco products." But, sadly, "the Tobacco Control Act does not require that FDA make a finding that a product is harmful in order to deem it subject to chapter IX of the FD&C Act; FDA is authorized to deem any product that meets the definition of a "tobacco product" pursuant to section 901 of the FD&C Act."

At least according to the FDA, Congress did not instruct it to reduce the risks associated with traditional tobacco use. Instead, Congress told the FDA to regulate tobacco or anything, such as nicotine, derived from tobacco, and that is what the FDA is going to do.

If the courts or Congress instructs the FDA to consider risk reduction in its treatment of e-cigarettes, then the findings of Abigail Friedman's paper would seem relevant. She compares conventional smoking rates in states that have enacted e-cigarette sales bans for minors to smoking rates in states that have not banned sales to minors. She includes controls for state and year fixed effects. She concludes that cigarette use among 12–17 year-olds increases by 0.7 to 1 percentage points (relative to other states) after a ban on e-cigarette sales to minors is enacted.

Does easy access to e-cigarettes induce kids to "smoke" who would not smoke if only conventional cigarettes were available? Friedman estimates propensity-to-smoke regressions using demographics and other relevant variables. Those predicted to be unlikely to engage in conventional smoking show no change in recent use of e-cigarettes, while those predicted to be likely smokers exhibit falling conventional cigarette use over time and increased e-cigarette use.

Corporate Inversions

"Are Corporate Inversions Good for Shareholders?" by Anton Babkin, Brent Glover, and Oliver Levine. December 2015. SSRN #2700987.

Among industrialized countries, the United States has the highest corporate tax rate, at 35 percent. To take advantage of lower rates in other countries, some U.S. firms elect to sell themselves to smaller foreign firms, a process called "inversion." These sales have drawn considerable ire from U.S. politicians and activists, and the Obama administration has vowed to "do something" about it.

In April, the U.S. Treasury Department issued regulations to throw sand in the gears of inversions. The commentary on these regulations—both pro and con—was the usual food fight about growth, taxes, and investment in the United States. Underlying all of that talk was the assumption that inversion reduces taxes for shareholders. But that assumption is incorrect in many instances.

The tax consequences of inversions are complicated because they are taxable events. That is, individual shareholders are taxed on the increased value of their shares. This can result in different tax outcomes from inversions for shareholders who have held the stock for a long time prior to the inversion and short-term shareholders (including corporate officers exercising company stock options).

For an inversion to be advantageous, the present value of the corporate tax reduction must be larger than the capital gains tax payments. When the reduction in the effective corporate tax rate is modest, those shareholders who purchased the stock near the current market price and thus have little individual capital gain to be taxed are net winners while longer-term shareholders who purchased the stock at a price much lower than the current market price, and thus have large capital gains, are net losers.

The authors study 73 inversions that occurred from 1983 to 2014 for which equity price data are available. Using historical price and turnover data for the 73 inversions, the authors estimate returns for the average, median, 10th percentile and worst-off shareholder. For those investors who had owned stock for five years or more and whose firms had 35 percent foreign earnings, the average return (in the worst 10 percent of the distribution of inversions) was –2.21%.

So if many long-term shareholders lose from inversions, why do they occur? The authors focus on the difference in returns to CEOs and long-term shareholders. The return earned by CEOs of inverted companies is different than the return of average shareholders because the CEOs have options rather than stock and the difference between the option strike price and market price is not a capital-gain taxable event. In their data, the average CEO had capital gains of $1.3 million and paid a capital gains tax of $538,000, but they also had increased value of options of $530,000.

In many circumstances, the CEO has incentives to invest that are not well-aligned with long-term shareholders. The authors show that the higher the option compensation of a CEO, the greater the likelihood of an inversion, controlling for industry and year fixed effects, normal firm characteristics, as well as the foreign share of taxes.

Public Housing and Crime

"Externalities of Public Housing: The Effect of Public Housing Demolitions on Local Crime," by Danielle H. Sandler. March 2016. SSRN #2749322.

High-rise public housing has long been associated with crime in popular culture. This paper examines crime rates at the block level before and after the demolition of public housing buildings in Chicago from 1995 through 2010. The "treatment" radius is within 0.25 miles of demolition. The control sample radius is within 3 miles of demolition. Crime is reduced 0.047 crimes per month per block for every 100 units demolished or 2.4 percent for every building demolished. No increase in crime occurred within the control area around a demolition, so crime was not merely displaced. For the average demolition, total crime decreased by over 8% and murders and assaults decreased by over 30%.