Nuclear Power

• “Why Nuclear Power Companies Build Nuclear Reactors on Fault Lines,” by J. Mark Ramseyer. June 2011. SSRN #1874869.

In this crisply written paper, Mark Ramseyer describes what happened in Fukushima, Japan on March 11, 2011. On that day, a 9.0 magnitude (Richter scale) earthquake produced a devastating 14–15 meter tsunami that crashed against four Japanese nuclear plants outside the city and set off a national radiation emergency. (Elsewhere along the Japanese coast, the tsunami reached heights of 38 meters.) The water flooded the plants’ backup diesel generators, resulting in loss of coolant to the reactors. Given that tsunamis of equivalent height have hit Japan three times since 1896, the 2011 disaster was predictable. So why were the plants built in such a dangerous area?

Ramseyer argues that nuclear power is fairly unique in its ability to generate damages that easily exceed the net worth of its owners. In this case, a company with a net worth of around $32 billion has caused damages on the order of $60–70 billion. This occurs because all risks created that exceed the net worth of the company are “free” from the perspective of its owners because they suffer no loss beyond the destruction of the value of their shares.

In theory, regulation of private plants can internalize the excessive external costs problem, but “good” regulation is itself a public good and the companies will take much more interest in the regulatory process and its results than the public.

Would government ownership of nuclear plants result in safer operation? Ramseyer argues no because unequal taxation, which would be used at least partly to fund nuclear power, leads to most voters receiving the benefits while a small minority of taxpayers pay the bills through the progressive tax system.

Public Sector Pension Underfunding

• “How Much Do Public School Teachers Value Their Retirement Benefits?” by Maria D. Fitzpatrick. Stanford Institute for Economic Policy Research.

Public sector employees typically receive a relatively large fraction of their income in the form of deferred defined-benefit compensation. While private employees earn $1 in compensation through employer-provided retirement benefits for every hour of their work, teachers earn $3. This is a particularly worrisome tendency because state pension funds have been underfunded since before the decline in financial markets during the Great Recession.

Maria Fitzpatrick asks how teachers’ valuation of deferred compensation compares to the cost of providing it. If teachers have a strong preference for current rather than future compensation, then the possibility exists that states can purchase back their future obligations, improve their finances, and make the employees better off.

The literature reports that, in general, retirees are willing to exchange half of their Social Security benefits for a lump-sum payment even if that payment were only 75 percent of the present discounted value (PDV) of the expected benefits. Similarly, in the military drawback program of the early 1990s, most of the separatees selected a lump-sum payment over a retirement annuity worth twice as much in present value.

Fitzpatrick uses a change in the Illinois teacher retirement system enacted in 1998 to estimate the value that teachers place on current versus deferred compensation. Illinois teachers were given the opportunity to purchase an upgrade to their pensions after the 1998 law passed. The price was 1 percent of salary as of 1998 per year of service. The benefit for a teacher with 20 years experience would be an 8.2 percent per year increase in retirement pension in each year of retirement. So for a one-time payment of 20 percent and immediate retirement, the payback period would be less than 3 years. Fitzpatrick calculates that 99 percent of teachers would receive a rate of return of more than 7 percent for a pension upgrade.

But even though the returns for this upgrade were generous, the author estimates that teachers were willing to pay only a fraction (less than 20 percent) of its cost. That is, teachers were willing to pay less than $2 for $10 of future benefits in present value. For example, the average price of the enhanced plan offered to employees with 25 years of experience in 1998 was $15,245 while the expected costs of providing them with the extra retirement benefits if they all purchased the upgrade would have been $94,166. Given the data on the upgrade purchases, the author concludes that the typical teacher in Illinois public schools values future compensation at just 17 cents on the dollar. Fitzpatrick suggests that states could dramatically reduce their currently unfunded pension obligations if they offered a future benefits buy-back option at the rate of 20 cents on the dollar, more than the rate (17 cents) at which employees value them.

Hygienists vs. Dentists

• “Battles among Licensed Occupations: Analyzing Government Regulations on Labor Market Outcomes for Dentists and Hygienists,” by Morris M. Kleiner and Kyoung Won Park. November 2010. NBER#16560.

One would think that the decline of unions in the private sector (see “The Rise and Decline of Unions,” Summer 2007) implies greater freedom in the labor market. Morriss Kleiner, however, has documented that as the labor force has become more educated and white-collar, unions have been replaced by occupational licensure (“A License for Protection,” Fall 2006). For instance, dentists and dental hygienists are both licensed in all 50 U.S. states. Generally, dentists dominate the state licensing boards that, among other duties, advise the legislature on the services hygienists can perform relative to dentists and the kind of supervision of the former by the latter. Some states, however, allow greater autonomy for hygienists. As of 2007 seven states allowed hygienists to be self-employed without the direct oversight of a dentist. What effects does this freedom have on wages and employment?

In the states that allow hygienists to be self-employed and perform relatively broadly defined services, hygienists have about 10 percent higher earnings and 6 percent higher employment growth. Dentists in those states have 16 percent lower hourly earnings and 26 percent slower employment growth. On the other hand, the transfer of income from hygienists to dentists in those states where regulation limits hygienists’ occupational freedom is estimated to be approximately $1.34 billion per year.

The authors are able to estimate the magnitude of the deadweight loss assuming that there is no reduction in the quality of services provided to patients in states with fewer dentistry regulations. The output loss from licensing is found to be between $540 million and $680 million per year, with an additional $80 million per year in losses from the restrictions that allow hygienists to be employed only by dentists. The combined loss estimate is approximately $620–$750 million per year. In other words, there is an approximately 1 percent annual reduction in the output of dental services for those states that required dentists’ supervision of dental hygienists.

Occupational regulation distorts free-market outcomes. It redistributes wealth from lower-skilled occupations to higher-skilled occupations and imposes efficiency losses on society.

FTC v. Intel

• “Does Antitrust Enforcement in High Tech Markets Benefit Consumers? Stock Price Evidence from FTC v. Intel,” by Joshua D. Wright. May 2011. SSRN #1739786.

On December 16, 2009 the Federal Trade Commission filed a complaint against Intel alleging that its loyalty discounts program violated antitrust law. The FTC argued that the computer chip maker’s loyalty discounts prevent its rivals from achieving minimum efficient scale to compete effectively. Intel’s practices gave computer manufacturers incentive to purchase almost all of their microprocessor and graphic processor units from Intel. Once rival chip producers fail, conjectured the FTC, Intel would raise prices to extract surplus from the captured chip buyers. Intel and the FTC settled the complaint in August 2010 without Intel admitting that it had committed any anticompetitive acts.

The evidence demonstrates that neither Intel’s principal rival nor consumers were harmed by Intel’s pricing behavior.

Because the predictions of economic theory on the effects of loyalty discounts are inconclusive, government investigations rely on empirical analysis. The case of Intel is unusual in terms of the volume of available data: their discount program had been in place for about 10 years. Wright uses two methods to estimate the competitive effects of Intel’s conduct. First, he follows the conventional approach of looking at the traditional antitrust metrics — market share and profit margins — and second, he uses an alternative approach that is based on financial market information. Neither of the two methods supports the hypothesis that Intel’s behavior harmed consumers.

The conventional approach tests the prediction that if Intel’s practices are anticompetitive, then Intel’s market share would increase and the market share of its rivals would decrease. During the period when the Intel discounts were in place, AMD, Intel’s main competitor, did not experience a decline in market share. In addition, AMD’s financial statements showed that the company increased its output and invested heavily in capacity expansion.

Financial market data analysis makes clear that AMD’s stock performed well from 2001–2002 to 2006, a period when Intel was providing its buyers with loyalty discounts. The available evidence is not sufficient to attribute the decline in AMD share performance after 2006 to Intel’s practices. Moreover, when both firms’ performance is tested against the market’s performance starting in 1999, Intel’s cumulative abnormal returns actually trend downward, which is not the case for AMD until around 2007.

The evidence demonstrates that neither Intel’s principal rival nor consumers were harmed by Intel’s pricing behavior. The antitrust action by the FTC had no empirical basis.

Effects of Gifted and Talented Educational Tracking

• “Is Gifted Education a Bright Idea? Assessing the Impact of Gifted and Talented Programs on Achievement,” by Sa A. Bui, Steven G. Craig, and Scott A. Imberman. May 2011. NBER #17089.

On March 13, 2010 the Obama administration released its proposal for reauthorization of the No Child Left Behind Act (NCLB). Some criticize NCLB because of its emphasis on students who perform below grade level. But the Elementary and Secondary Education Act (ESEA) (which was renamed NCLB in its reauthorization under President Bush) includes the Jacob Javits Gifted and Talented Students Education Act passed in 1988. The Gifted and Talented program (GT) created by the act counterbalances some of the NCLB attention to low-ability students.

Not much is known about the effectiveness of GT programs. This paper attempts to estimate the effects of these government educational programs on high-ability students.

As with all educational policy studies, the econometric problem is to separate the effects of underlying unobserved student characteristics, such as motivation, from the causal effects of participation in a GT program. The correlation of unobserved characteristics with attendance and achievement likely produces an overestimate of the true impact of a GT program on achievement.

The study takes two principal approaches: First, the authors exploit a regression discontinuity (RD) design to compare the performance of two student groups: those who scored just above the well-defined cutoff of eligibility to the GT program, and those who scored just below the cutoff. Students are selected into the program on the basis of an index score that combines achievement tests, a nonverbal ability test, grades, teacher recommendations, and socioeconomic status. The objective of the regression discontinuity analysis is to estimate the effect of enrollment in a GT program among similarly gifted students, some of whom randomly do not attend a GT program. Second, the authors exploit the randomized nature of the enrollment of talented students into a premier GT magnet school. They compare performance of the students who win the lottery and attend one of these schools to the students who lose the lottery and either attend a neighborhood GT program in the district, a magnet school based on a different specialty, or a charter school. This structure allows for estimation of the effect of a premium magnet school on the talented students’ achievement.

The authors obtained the administrative records in a large urban school district in the American Southwest for the 2007-08 to 2009-10 academic years. Their sample for the RD analysis includes approximately 2,600 students in the 7th grade cohort who were evaluated for GT in 5th grade. For the lottery analysis, the sample consists of 542 students who applied for admission in one of the district’s oversubscribed magnet schools. The authors perform numerous robustness tests to ensure that their results are unbiased and consistent. For example, they test for a possibility that teachers manipulate student evaluations to compensate for the low test scores of applicants.

The researchers make two important findings: Regression discontinuity analysis showed that students who were exposed to GT curriculum for 1.5 years in middle school demonstrated no significant improvement in achievement as compared to their peers who just barely missed the cutoff. In fact, their preferred regression specification shows declines in reading, math, and social science scores, although the results were not statistically significant. Other specifications report similar results, so the authors are able to rule out positive effects of the GT programs. Another interesting finding is that although the GT students got better peers, they did not have better teachers assigned to them. The lottery-based study that compared students in the two premier GT magnet schools to other GT students also showed little improvement in overall 7th grade achievement with the exception of science scores, despite such clear advantages as higher-quality teachers and peers.

NCLB does emphasize the results of the least successful students and some have voiced concern about the implications of this emphasis on the progress of the gifted. The presumption behind such concern is that programs for GT students have positive results that would suffer from reduced emphasis and funding. At least in one large urban school district, GT programs do not appear to have any effects at all on student achievement.

Health Care Reform

• “The Health-Related Tax Provisions of PPACA and HCERA: Contingent, Complex, Incremental, and Lacking Cost Controls,” by Edward A. Zelinsky. June 2010. SSRN #1633556.

In last winter’s issue of Regulation, David Hyman argues (“In Medicine, Money Matters”) that the current structure and performance of the U.S. health care system is the result of incentives created by its (fee-for-service and Medicare) payment systems. Rather than realigning those incentives, the Patient Protection and Affordable Care Act (PPACA) and the Health Care and Education Reconciliation Act of 2010 (HCERA) broaden access to the existing flawed system and put off reform into the future.

In this paper, Edward Zelinsky describes the excise tax on “high cost” health plans, the Medicare tax on investment income, the increase in the Medicare tax rate for high-income earners, the small-employer health insurance and premium-assistance tax credits, and tax-enforced individual and employer mandates. Zelinsky arrives at a conclusion similar to Hyman’s: health care reform increases spending and demand for health care largely within the existing system. The tax provisions are complex, friendly to the status quo, and do not control health care costs.

A central feature of the new federal health care law is the mandate that individuals purchase and employers offer health insurance. If noncompliant, an individual must calculate and pay an excise tax penalty, which is determined through a non-trivial four-step procedure. The process is complicated by a number of exemptions, typically income- or employment-related, that have to be determined through elaborate formulas. In 2014, eligible taxpayers will be able to claim a premium-assistance tax credit. The amount of the credit for individuals whose household income exceeds the poverty level but is below 400 percent of the poverty level is determined monthly through a complex procedure. Together, the individual mandate and subsidies impose significant compliance obligations on taxpayers, primarily those of modest means, and add to the IRS enforcement burden.

The employer mandate has similar effects. It obligates every “large employer” to provide its workers with an affordable plan that offers no less than minimum coverage as established by law. The status of “large employer” is determined yearly while the number of full-time employees is determined monthly. The employer is penalized on a monthly basis if it does not offer affordable medical coverage to all its full-time employees and their dependents. If fined, the employer pays the penalty calculated according to the number of employees, even if only one of the workers’ insurance conditions triggers the penalty. In addition, compliance with the affordability provisions by employers requires the employer to know employee adjusted gross income, which means the employer must know workers’ deductions, not just gross income — which is not possible under the current system.

Besides the compliance requirements, the most important cost a firm will face is the decision to hire its 50th employee. That will trigger reclassification from small to large employer and subject the firm to the mandate.

Ironically, according to Hyman and Zelinsky, health care reform preserves many of the perverse incentives of the current health care system and is more incremental than either its proponents or opponents acknowledge. The tax on “Cadillac” health care insurance plans, for example, is scheduled to take effect in 2018. Why should we expect a future president and Congress to let that tax go into effect?

PPACA and HCERA exacerbate U.S. health care problems through both complexity and incrementalism. The tax provisions postpone politically difficult decisions rather than undertake serious health care cost control.