The ACA and Opioid Deaths
“Health Insurance and Opioid Deaths: Evidence from the Affordable Care Act Young Adult Provision,” by Gal Wettstein. Forthcoming in Health Economics.
Accidental drug overdoses have become the leading cause of death for those under age 50, and the rate of death via opioids has increased dramatically in the last few years. In 2017, approximately 72,000 people died of a drug overdose in the United States, which is nearly twice as many as in 2013 and four times as many as at the turn of the 21st century.
The recent spike in drug mortality coincides with the advent of the Affordable Care Act, which greatly increased the ability of young people to obtain medical coverage. In the years following the ACA’s passage, insurance coverage for people between 18 and 25 increased from 70% to 87%. This has led some people to infer that the increase in insurance coverage contributed to the increase in opioid deaths. The rationale is that having a doctor and insurance coverage makes it easier for people to access and become addicted to opioids, despite attempts in recent years to restrict the availability of the drugs.
Despite the timing, it is not clear that the increase in opioid deaths has anything to do with the increase in health insurance coverage. In this paper, Gal Wettstein notes that, ex ante, the very opposite effect is possible: People with regular health care should have better health and thus have less reason to seek painkillers to begin with. Moreover, those who do become addicted will find it easier to access treatment via mental health counseling or medication, as well as follow-up care. Most importantly, addicts with insurance generally abuse prescription opioids, which are less risky than heroin or other narcotics bought illegally. Indeed, the Centers for Disease Control attributes fully 40% of all overdose deaths to fentanyl, an incredibly lethal drug that is often added to batches of heroin to accentuate the high it confers.
With such confounding intuitions, Wettstein turns to the data to attempt to determine if there is a connection between the ACA and opioid deaths. He uses the ACA’s health insurance provision for young adults as a quasi-experiment. The ACA allows children to remain on their parents’ health insurance until they turn 26, and this provision took effect upon the law’s passage on September 23rd, 2010, while many other ACA provisions did not take effect until 2014. The distinct implementation dates mean that we should see the effect of the ACA on drug deaths—if there is one—occur at different times for different age cohorts.
Two confounding events occurred between these two dates that make this natural experiment a bit less than ideal. The first is that marijuana became legal (or quasi-legal) in several states over this time period, and there is some evidence that marijuana dampens opioid usage by serving as another way for people to deal with pain. Also, use of naloxone, a medication that can rapidly alleviate the effects of an overdose, became more prevalent over this time. It is possible naloxone availability could reduce overdose deaths; it is also possible it could contribute to them through moral hazard: people may be more willing to take risks with opioids if they know naloxone is at the ready in case of overdose.
Another problem in this analysis, Wettstein notes, is that it can be difficult to discern precisely what, in fact, killed someone. Not all decedents get an autopsy and coroners may forgo one if they believe that they can easily discern a cause of death from circumstances and there is no next of kin insisting that an autopsy be done. People who spent years abusing drugs and died of a heart attack at a young age may have clearly had their lives cut short because of drug abuse, but the coroner may attribute their death to natural causes.
Wettstein looks at opioid-related deaths of people ages 19–29 by year and state from 2011 to 2016. The 29-year-olds in 2016 could obtain coverage in 2011. He compares this cohort’s overdose death rate to the overdose death rates via opioids for people age 32–36, which had no access to young-adult health insurance coverage.
He employs two distinct methods of analysis. The first is a difference-in-differences approach, which entails comparing the two groups. The second method is a simple dose-response model whereby he measures what happened after the implementation of the ACA while attempting to control for other factors.
Wettstein finds that the deaths from opioid abuse in the older cohort increased faster than the younger group post-ACA using the difference-in-differences approach. That leads him to tentatively conclude that health insurance access reduced deaths from opioids. However, there are caveats. One concern is that prescription opioids spill over between age groups within a state, confounding cohort comparisons. For instance, in their 2015 paper “How Increasing Medical Access to Opioids Contributes to the Opioid Epidemic: Evidence from Medicare Part D,” David Powell, Rosalie Liccardo Pacula, and Erin Taylor found that states with higher take-up rates for Medicare Part D were associated with greater drug abuse for non-retirees as well. In other words, it may be that the younger people having more insurance may actually increase access to drugs for older people as well. This “dilutes the experiment,” giving us one more thing that cannot be controlled for.
The regression results from the dose-response methodology show a decline in deaths from an increase in health insurance coverage. Wettstein does not discern any obvious stepwise linear tradeoff, but his data do suggest accumulated declines in drug abuse deaths. By looking at the entire panel of observations, he discerns a lagged effect to access to health insurance, with the reduction in death rates from health insurance access increasing in subsequent years. He estimates that a 1–percentage point increase in health insurance coverage ultismately reduces opioid deaths by 3.6 per 100,000, which is a 16.5% reduction.
Wettstein cautions against reading too much into his data, noting that 2011–2016 may turn out to be anomalous, with death rates much higher than in previous—and hopefully subsequent—eras. He concludes that health insurance seems to have reduced drug deaths from where they would be otherwise, and he suggests that it does so partly through the improved physical and mental health that regular access to health care begets. —Ike Brannon
Banking Regulation
“The Limits of Shadow Banks,” by Greg Buchak, Gregor Matvos, Tomasz Piskorski, and Amit Seru. October 2018. SSRN #3260434.
Traditional banks accept deposits that are federally insured, issue loans, are members of the Federal Reserve System, and are subject to safety and soundness banking regulations and examinations. Shadow banks, on the other hand, do not accept deposits; they raise money in the capital markets. They originate loans, not to hold in their portfolio, but to securitize and sell to other investors. Among those investors are the government-chartered Federal National Mortgage Association (Fannie Mae) and Federal Home Loan Mortgage Corporation (Freddie Mac), which also receive government subsidies.
The 2010 Dodd–Frank banking reform legislation, passed after the financial crisis in 2008 and the ensuing Great Recession, altered the regulation of traditional banks but not shadow banks. This paper argues that because shadow and traditional banks are partial substitutes for each other, the effects of regulatory reform on traditional banking has shifted some banking activity to the shadow sector, dampening any soundness benefits from Dodd–Frank.
The market for mortgages is segmented. Well-capitalized traditional banks issue so-called jumbo loans (above $484,350, or $726,525 in high-cost home areas) that cannot be sold to Fannie Mae and Freddie Mac. Thus, the traditional banks hold those loans on their balance sheets. Shadow banks originate loans to distribute to Fannie and Freddie or private investors through securitization. Poorly capitalized traditional banks with limited balance sheet capacity also originate loans to distribute to investors.
A central component of traditional banking reform has been increased capital requirements. The more equity in a traditional bank’s capital structure, the less likely depositors are to lose money if loans are not repaid in full. (See “Bank Capital Requirements,” Working Papers, Winter 2010–2011). Under Dodd–Frank, capital requirements were increased from 4% of assets (i.e., loans) in 2010 to 6% in 2015.
The central insight of this paper is that traditional banks rely on deposit insurance to attract deposits while shadow banks and poorly capitalized traditional banks rely on the government subsidies to Fannie and Freddie to facilitate their business model. As subsidies for traditional banks decline because of increased capital requirements, jumbo loan activity declines but “conforming” loan activity—loans below $484,350 or $726,525 in high-cost loan areas—increases to take advantage of the mortgage guarantees provided by Fannie and Freddie. —Peter Van Doren
Soda Taxes
“The Impact of Soda Taxes: Pass-Through, Tax Avoidance, and Nutritional Effects,” by Stephan Seiler, Anna Tuchman, and Song Yao. January 2019. SSRN #3302335.
Taxes on high-calorie beverages, i.e., “soda taxes,” have become a popular policy response to the obesity epidemic. Mexico implemented a nationwide soda tax in 2014. Estimates of its effects have used standard elasticity estimates that a 1% increase in soda price results in a 1–3% decrease in consumption. In a previous Working Papers column (“Soda Taxes,” Winter 2017–2018) I discussed research that lowered those estimates by considering the purchase of cheaper soda (switching brands) as a taxpayer response.
In the United States, beverage taxes have been enacted by localities rather than nationwide. A response to such a tax could be shopping outside the jurisdiction and avoiding the tax. On January 1, 2017, Philadelphia imposed a 1.5¢ per ounce tax on sweetened beverages. This was a large tax, amounting to $1.01 on a 2‑liter bottle that had a pre-tax price of $1.56—a 65% tax on the price. In comparison, the Mexican tax was 9% of the pretax average.
In this paper, the researchers found that the Philadelphia tax has had little, if any, effect on city residents’ consumption of caloric soda. The researchers found that beverage purchases within Philadelphia decreased by 42% after the tax, but that reduction was fully offset by an equivalent increase in purchases in stores outside of Philadelphia. —P.V.D.
Securities Regulation
“A New Market-Based Approach to Securities Law,” by Kevin S. Haeberle and M. Todd Henderson. August 2018. SSRN #3233122.
Three claims are used to justify modern securities regulation:
- Firms fail to disclose enough information.
- Firms disclose untruthful information.
- Insiders trade on information, reducing their incentive to release information and the incentive of outsiders to invest in information production.
Modern securities regulation attempts to solve all three of these problems through mandates and restrictions.
Government-mandated production of information results in the overproduction of information irrelevant to firms’ soundness (e.g., blood diamond disclosure, CEO pay ratios) and underproduction of relevant information. For instance, disclosure law has become a focal point and securities fraud litigation reinforces this legally defensible but mindless focal point. “Additional statements mean additional exposure to lawsuits based on the allegation that those statements are false or misleading,” note the authors of this paper. Class action investor fraud lawsuits result in overcompensation (the authors point out in a footnote that, since 1996, these suits have named 35,000 defendants and produced $95 billion in settlements) and are just a wealth transfer from one set of shareholders to another, with a healthy cut for the lawyers. (See “The End of Securities Fraud Class Action?” Summer 2006.) The marginal deterrence benefits from securities fraud are low because the payments are orders of magnitude greater than the actual level of fraud.
For the authors, the central economic problem with current securities regulation is that it mandates that firms provide information for free. The authors’ solution is to legalize payments for early access to public information. This money would generate incentives for firms to provide the information that investors want. Participation in class action securities fraud suits would be limited to those who paid for early access, acted on that information, and lost money because of fraudulent information. And corporate insider trading would be severely reduced because such behavior would now cost the firm money. Insider trading would undermine the firm’s profits from selling early access to information.
The most important objection to their proposal is that, in this new regime with advanced disclosure, there would be no uninformed investors from whom the knowledgeable could buy and sell securities. The uniformed, who did not pay for access, would avoid trading during the publicly announced time periods in which some investors get early access to the information, thereby protecting the uninformed from getting fleeced by the informed. Thus, the only people trading in these periods would be the informed. Given the belief that serious money is made only by the informed trading with the uninformed, there wouldn’t be anyone willing to pay for early access to information because they couldn’t make any money from that information. In this view, the wolves make money only by selling to the sheep.
The authors counter that informed people now trade with each other because they differ on the implications of information that they all possess. “Information can be valuable even when other people have it (if you have different predictions or can get to market first) or if it can be used to predict outcomes in related areas.” —P.V.D.
