Health Care

• “What Do Health Insurance Deductibles Do to Health Care Spending Growth and Its Efficiency?” by Claudio Lucarelli, Molly Frean, Aliza Gordon, et al. SSRN Working Paper no. 3985356, December 2021.

The rationale for high deductible health care insurance accompanied by individual health savings accounts is the simple principle taught to all undergraduates in Economics 101: when consumers have their own money at stake, they are more careful with their spending. But health care seems to be different. Managed care and higher patient cost-sharing have all been shown to lower the level of spending as they are phased in, but after an initial implementation period, spending growth resumes its former rate.

Across the member countries of the Organisation for Economic Co-operation and Development, the rates of growth of health care expenditures since 1960 are remarkably similar despite large differences in industrial organization. From 1960 to 1998, the OECD median real per-capita health expenditure growth rate was 4.5% per year while the U.S. rate was 4.4%. From 2005 to 2019, the OECD real per-capita health expenditure growth rate averaged 2.2%; the U.S. rate was 2.1%, Switzerland 2.2%, Canada 2.1%, and the United Kingdom 2.1%. The conclusion I draw from the similarity in growth rates across developed countries, whose health care systems range from single-payer government systems, to largely private, and everything in between, is that policy differences matter less than all the political chatter about health care would suggest.

This paper asks whether high-deductible health care insurance plans with higher levels of cost-sharing generate a permanent reduction in the rate of growth of spending. The data come from claims from 2015 to 2018 from a large national insurance company and compare the spending of populations that have remained in high- or low-deductible health insurance plans for at least four years. Those enrolled in high-deductible plans do not experience a lower rate of spending growth for medical claims but do experience a lower spending growth rate for prescription drugs. The reduction in spending was for less cost-effective drugs (cost per quality-adjusted life year [QALY] above $50,000); there was no reduction in the growth rate for highly cost-effective drugs (cost per QALY below $50,000). The findings were robust to controls for the non-random nature of those who enroll in high- and low-deductible health plans (so-called selection effects).

Minimum Wage Effects on Low Wages

• “How Important Are Minimum Wage Increases in Increasing the Wages of Minimum Wage Workers?” by Jeffrey Clemens and Michael R. Strain. NBER Working Paper no. 29824, March 2022.

Minimum wage policy discussions suggest that the decision by states to increase the minimum is the important determinant of wage increase for low-wage workers. In this paper, the authors use Current Population Survey data from 2010 to 2019 in which the same respondents are asked about their wages 12 months apart. For those who were employed at both times, wage growth can be calculated.

Of those workers with wages within 50¢ of the minimum during the first interview, more than 70% of those employed 12 months later had a higher wage. For those who received an increase, their wages rose an average of $2.05 an hour. Thus, a large majority of low-wage workers cannot be described plausibly as “career minimum wage workers.”

The paper compares the wages of those in states that increased the minimum wage to the wages of those in states that did not. The effect of state minimum wage increases is real but small. Around 71% of minimum wage workers in states that did not increase their minimum wage at any point in the 2013–2018 period got a raise in any given year, compared to around 79% of minimum wage workers in states that did increase their minimum wage. In other words, minimum wage increases account for about 8% of the wage increases realized by minimum wage workers across the full sample period.

Consumer Credit Cards

• “Credit Cards and the Reverse Robin Hood Fallacy: Do Credit Card Rewards Really Steal from the Poor and Give to the Rich?” by Todd Zywicki, Ben Sperry, and Julian Morris. SSRN Working Paper no. 3984298, December 2021.

Many credit cards offer cash-back or other rewards for use. Some consumer advocates claim that, because merchants pay a higher interchange fee to process these cards, the rewards impose a hidden tax on consumers who pay cash because they pay the same prices as card-users. Cash-payers, in turn, are presumed to be lower-income than the card-users, which would make the reward schemes regressive. This paper assesses the evidence on whether this “reverse Robin Hood” theory is true.

Some stylized facts about cash and card transactions: First, consumer use of cash also imposes costs on merchants. The average retailer spends more than 9% of the value of its cash transactions counting, auditing, and depositing cash. Electronic transactions have lower transaction costs for merchants. Second, consumers who pay with cards spend two to four times more than those who use cash. The average transaction for those using rewards cards is 25% to 60% larger than those using non-rewards cards. And premium rewards-card transactions are 30% larger than regular rewards cards. So, merchants benefit from those who use rewards cards.

A necessary condition for the reverse Robin Hood hypothesis to be true is that consumers with different incomes buy the same goods at the same stores at the same prices and higher-income consumers pay with rewards cards while lower-income consumers pay with cash. If consumers with different incomes shop at different stores or buy different goods, merchants cannot pass on costs created by one type of consumer to other types. For example, if cardholders buy premium products, and cash users buy generic, and merchants charge higher margins for premium products, cardholders pay for their own rewards.

Even for those products bought by both cash and card customers within the same store, why don’t merchants offer cash discounts? Federal law guarantees merchants the right to do so. Most merchants do not, apparently because the costs of handling cash are real and substantial.

A final requirement for the reverse Robin Hood hypothesis to be true is that, for items within a store that both cash and card customers purchase, the interchange fees charged to merchants for card use are passed through to all consumers through price increases. How much of the interchange fees are passed on to consumers through price increases?

Evidence from taxes on producers or changes in foreign exchange rates suggests that the pass-through rate for these charges is about 50% in the long run. More directly analogous evidence comes from the cap on debit card interchange fees mandated by the Durbin amendment to the 2010 Dodd–Frank financial reform legislation. (See Working Papers, Summer 2019.)

Did consumers benefit from the interchange fee reduction? The large banks affected by the debit-fee rule totally offset their $6.5 billion loss in debit card interchange fees by charging higher checking account fees. Monthly maintenance fees on checking accounts doubled, decreasing the share of consumers with free checking accounts from 60% to 20%. And an intensive study of gasoline stations found no reduction in prices for consumers. So, reductions in debit-interchange fees did not benefit consumers. Thus, the original incidence of such fees when they were first imposed may well have been largely on merchants.

But let’s assume the best case for the reverse Robin Hood hypothesis: cash and card consumers buy the same products in the same stores, and prices increase to reflect all card interchange fees; thus, cash customers pay a “tax.” Are card customers higher- income and cash customers lower-income? Two-thirds of adults earning less than $40,000 per year have credit cards. Some 98% of credit cardholders own a rewards card, including 82% of cardholders earning less than $20,000 per year. So even if cash customers pay a “tax,” the relationship between rewards card ownership and income is very weak. The distributional consequences are not obviously regressive.

The authors offer a more neutral summary of the evidence regarding the effects of rewards cards:

Those with better credit scores, regardless of income, benefit from rewards programs, which are partially “paid for” by interchange fees charged to merchants. Those interchange fees, in turn, may or may not be passed on to consumers who use cash, depending on whether those consumers buy the same goods and services from the same merchants as those using credit cards. But even then, the pass-through is a proximate result of decisions by merchants not to offer cash discounts, often because the administrative cost of doing so is greater than any benefit they would reap through larger margins on cash transactions.

Environment

• “Why Are Marginal CO₂ Emissions Increasing for U.S. Electricity? Estimates and Implications for Climate Policy,” by Stephen P. Holland, Matthew J. Kotchen, Erin T. Mansur, and Andrew J. Yates. Working paper, November 2020.

From 2010 to 2019, coal-fired electricity generation declined 48% and natural gas generation increased 58%. Coal combustion results in more carbon emissions per unit of electricity generated, thus average annual emissions declined 3.5%. This has led many to believe that electrification is the solution to all climate problems — including replacing natural gas use. In July 2019, Berkeley, Calif. became the first city in the nation to ban natural gas hookups in new construction or substantially renovated structures. Since then, other localities have also banned natural gas, including Brookline, Mass., San Jose, Calif., Seattle, Sacramento, and New York City.

Even though average carbon emissions have declined, marginal emissions have increased during the day as electricity use rises from its nighttime lows. How is that possible?

Historically, because of its lower marginal cost, coal-fired generation has been used to meet base load. Natural gas generation historically has had higher marginal cost but low fixed cost, and so it has been used to follow marginal increases or decreases in consumption during the day.

But the composition of natural gas electric generators has changed over time, from natural gas turbines to more efficient natural gas combined cycle units in which the waste heat from the turbines is used to make steam and generate additional electricity. Combined cycle’s low total cost has resulted in it displacing some baseload generation, meaning that now some coal-fired plants are available to follow demand during the day — and, in turn, produce higher carbon emissions when they are dispatched.

Since 2010, marginal carbon emissions increased 6% in the East and 15% in the West. The increase in marginal emissions for the United States overall was 7% over the last decade. This has occurred even though average emissions declined 28% over the same period, as combined cycle replaced baseload coal.

Public Attitudes Toward “Extreme” Vaccine Demand

• “Vaccine Hunters and Jostlers May Have Hurt the COVID-19 Vaccination Effort,” by Johanna Mollerstrom and Linda Thunström. SSRN Working Paper no. 4012923, March 30, 2022.

In the early months of 2021, as a limited supply of COVID vaccine became available and was earmarked for high-risk people, news stories began appearing about the lengths to which non-prioritized people would go to get a shot. Some stopped by pharmacies at the end of the day, hoping to get a dose of the perishable vaccine because it would otherwise be discarded; others falsely posed as priority recipients to get a jab. The stories became such a sensation that the police procedural Law & Order: Organized Crime featured a plotline about a mob-orchestrated theft and distribution of a truckload of the vaccine.

One might expect that stories of extreme demand would increase people’s preference to receive the medication. To measure that, Johanna Mollerstrom of George Mason University and Linda Thunström of the University of Wyoming conducted a survey experiment of the effect of exposure to stories of “hunting” (attempting to get about-to-expire vaccine) and “jostling” (improperly receiving the vaccine). They conducted the experiment twice, first with a group of 1,503 participants (though only 1,117 of them were ultimately relevant to the hunting/​jostling issue), and then a follow-up study of 800 more. The samples were representative of U.S. age, gender, and race/​ethnicity demographics, though not education (survey participants were more likely to have a four-year college degree).

Surprisingly, Mollerstrom and Thunström found (95% confidence) that exposure to stories of extreme demand made survey participants less likely to want to receive a jab or recommend it to friends and family, either immediately or within two months. This result from the initial survey prompted the researchers to assemble the follow-up, which asked participants about their emotional responses to the stories of vaccine hunting and jostling. Respondents exposed to such stories indicated increased feelings of sadness, anger, and disgust.

Mollerstrom and Thunström interpret the results as indicating that stories of extreme demand may “negatively affect[] the general public’s enthusiasm for getting vaccinated and/​or recommending others get the vaccine.” This seems unnecessarily pessimistic. Among their other survey results, they found that respondents who were exposed to the hunter/​jostler stories were as interested in receiving further information about the vaccine as respondents who did not read of hunters and jostlers. The overall sample indicated a higher willingness to be vaccinated than the broader public expressed in other surveys. This suggests that the extreme demand stories only made people more likely to “wait their turn” to receive the vaccine, not that their “enthusiasm” for vaccination was dampened.

Pandemics, Liberal Democracy, and Quality of Life

• “Can Governments Deal with Pandemics?” by Vincent Geloso and Ilia Murtazashvili. SSRN Working Paper no. 3671634, December 3, 2021.

In the first year of the COVID-19 pandemic, the United States was criticized (often by Americans) for its high infection and mortality rates as compared to other developed countries. In the second year, after vaccines against the coronavirus became available, “red states” — those that went to Donald Trump in the 2020 presidential election — were chastised (often by Joe Biden supporters) for their lower vaccination rates and higher current infection and mortality rates than blue states. Critics charged the disease inflicted its heaviest tolls on places where people allegedly have outsized concern for (to borrow the mocking phrase of The Bulwark executive editor Jonathan V. Last) “muh freedom.”

Now, two-and-a-half years into the pandemic, the correlation between politics and the disease’s progress has attenuated somewhat. The United States still leads the Group of Seven developed countries in mortality rate from the disease, but its performance isn’t much worse than fellow G7ers Italy, the United Kingdom, and France, and those numbers periodically shift as different variants of the virus strike different areas. The same goes for U.S. states: the disease had some of its highest mortality rates in the “red” states of Alabama, Louisiana, Mississippi, and Oklahoma, but it also hit hard in “purple” Arizona and “blue” Michigan, New Jersey, and New York.

Still, the “muh freedom” critics have a point. Though liberal democracies like Australia, New Zealand, and Japan are on the low end of the international COVID mortality rate spectrum (and, worth noting, are islands), they are joined there by illiberal China, Cuba, Belarus, Vietnam, and Pakistan. In a pandemic where testing, quarantining, vaccination, and disclosure of health status can slow the spread of the disease, more coercive governments seem to have advantages over more liberal governments in combating this negative externality.

In this working paper, Vincent Geloso of King’s University College (Canada) and Ilia Murtazashvili of the University of Pittsburgh find empirical support for the coercive government–COVID policy correlation. Within the United States, they cite a 2021 Southern Economic Journal article by Joshua C. Hall and Bryan C. McCannon (both of West Virginia University) finding that states with heavier economic regulation were quicker to adopt stay-at-home orders to fight the pandemic. Internationally, Geloso and Murtazashvili use an ordinary least squares regression to find a negative correlation between the adoption of “stringent” pandemic policies and countries’ Economic Freedom of the World scores (99% confidence) and Polity Index scores (90% confidence). So, more freedom, less pandemic-suppressing policies.

However, Geloso and Murtazashvili argue, there is more to the story. It’s wrong to think about the benefits of coercive countries’ COVID policies without appreciating the broader harm of coercive government: “A government that can use coercion for good can also use it for less enlightened reasons.” More- and less-free countries adopt “institutional bundles” of policies that either nurture or suppress social and economic freedom. Such freedom may limit pandemic response, but it nurtures economic growth that provides more nutritious food, more fruitful research and development, better and safer infrastructure, and plenty of other goods that boost human health and happiness. In health care alone, communicable diseases cause only a small percentage of human deaths; plenty of other causes are suppressed by innovation nurtured by economic freedom. Write the authors, “Economically free democracies tend to enjoy faster economic growth, which in turn leads to better health outcomes with respect to noncommunicable diseases.” (I suppose it should be noted that the United States’ economic abundance also contributes to some of its leading causes of death.)

It’s even possible that, in the long term, these bundles of freedom will result in a better outcome for the COVID pandemic. Citing a 2021 Review of Austrian Economics article by Mark Pennington (King’s College, London), Geloso and Murtazashvili write that “the coronavirus pandemic is an example of what Hayek called a complete policy problem, with uncertainty arising from the epidemiology of the virus, its interaction with political, economic, and cultural arrangements that affect its spread, and the differing attitudes, time horizons, and belief systems that influence the spread of the disease.” Those are all positively affected by market signals. Though the authors don’t mention it, this is exemplified by the two breakthrough mRNA vaccines, one created by American biotech firm Moderna and the other brought to market in part by American pharmaceutical giant Pfizer. Contrast their timeliness, safety, and effectiveness with vaccines developed and administered in far more coercive China and Russia.

“It is clear,” the authors write, “that the institutional frontier of economically free democracies is a bundle of institutions that have costs on certain margins…. Understanding these tradeoffs is the first key to analyzing any pandemic.”