conversableeconomist

Government Workers: It\’s Not the Wages, It\’s the Benefits

On average, government worker are paid more than private-sector workers, at both the federal level and at the state and local level. But the comparison is not apples-to-apples. On average, government workers have much higher levels of education and experience.  When adjusting for levels of education, it turns out that average wages are very similar for government and private-sector workers. The real advantage that government workers have in their compensation is that they receive noticeably better benefits–especially in their pensions and health benefits after retirement.

The most recent article of my own Journal of Economic Perspectives has an article by Maury Gittleman and Brooks Pierce called \”Compensation for State and Local Government Workers.\”   In January, the
Congressional Budget Office published \”Comparing the Compensation of Federal and Private-Sector Employees.\”  Here,  I\’ll first summarize some of the main evidence from these studies, and then list some of the main reasons why studies that compare pay of government and private-sector workers can reach such different results.

Numbers of government workers in context

Here\’s the CBO summarizing patterns in total government employment in recent decades (footnotes and references to figures omitted): 

\”For the past 30 years, the number of civilians employed by the federal government has hovered around 2 million people. During that period, federal employees have accounted for a declining share of the total U.S. workforce, because employment by the private sector and by other levels of government has grown along with the economy. In 1980, when about 79 million people worked in the private sector and 13 million worked for state or local governments, federal employees made up 2.3 percent of the workforce. By 2010, private-sector employment had reached 111 million and employment by state and local governments had reached 20 million.\”

Higher Education and Skill Levels for Government Employees

Here are some illustrations of the education differences between government and private-sector workers. From Gittleman and Pierce, here\’s a table showing average pay and education levels for state government employees,  local government employees, and private-sector employees. For example, 39% of private-sector workers have only a high school education, or less education than that. Among state government employees, only 18% have a high school education or less. Conversely, about 10% of private-sector workers have a post-graduate degree, compared with 29% of state government workers.

Here\’s some evidence from the CBO, making the same point  about federal employees. In their data sample, 41% of private sector workers have a high school degree or less, compared with 20% of federal employees. Conversely, 10% of private sector worker have a post-graduate degree, compared with 21%  of federal employees.

Similar Pay, Dissimilar Benefits

When comparing pay for state and local government employees, Gittleman and Pierce write:

\”Government workers are much more likely to be offered health insurance and retirement plans, and are more likely to enroll in  such plans if offered. In addition, public sector plan structures tend to offer more
comprehensive coverage. Public sector health plans tend to require lower employee contributions and have higher employer premiums, and are more likely to come bundled with supplemental dental, vision, or prescription drug plan components. … [T]he costs per hour worked for the various benefits collected are
much greater in the public sector (about $14) than in the private sector (around $8). Spending on health insurance in the government ($4.30 at the state level and $4.56 at the local level) is more than double that in the private sector ($2.14), while expenditures on retirement and savings are more than triple ($3.18 and $3.37 versus $1.00). … Paid leave is also more generous in government, more than double the private sector level in state government and more than 50 percent higher in local government.\”

Overall, Gittleman and Pierce conclude: \”After controlling for skill differences and incorporating employer costs for benefits packages, we find that, on average, public sector workers in state government
have compensation costs 3–10 percent greater than those for workers in the private sector, while in local government the gap is 10–19 percent.\”

In comparing the wages, benefits, and total compensation of federal workers, the CBO report finds:

\”Overall, the federal government paid 2 percent more in total wages than it would have if average wages had been comparable with those in the private sector, after accounting for certain observable characteristics of workers. …  On average for workers at all levels of education, the cost of hourly benefits was 48 percent higher for federal civilian employees than for private-sector employees
with certain similar observable characteristics … The most important factor contributing to differences
between the two sectors in the costs of benefits is the defined-benefit pension plan that is available to most federal employees. … Overall, the federal government paid 16 percent more in total compensation than it would have if average compensation had been comparable with that in the private sector, after accounting for certain observable characteristics of workers.\”

Why do comparisons of pay for government and private-sector workers reach varying conclusions?

1) Some comparisons don\’t adjust for education or skill level of the workers, which means that the comparison will inevitably find that government-sector workers are paid much more on average. But such comparisons are silly. It\’s a bit like saying that–surprise!–those with a college degree earn more than those without a college degree.

2) Some comparisons look only at wages and skip benefit. Such comparisons leave out a main advantage for government workers, and will tend to find that they are paid about as much as private sector workers.

3) Employees in larger firms tend to be paid more than employees in smaller firms. There are various theories for why this pattern holds true. Perhaps large employers do a better job of screening employees to get those with higher productivity. Perhaps in a large firm there is space for a more specialized division of labor, which leads to more productivity and higher pay for workers in those organizations. Perhaps in large organizations, pay becomes a little more detached from productivity, and workers can claim a larger share of the cash flow running through the organization. Ultimately, the question here is whether a worker who moved between the government and the private sector has a type of skill and expertise that would also lead to higher pay in the private sector–or not. If you adjust for size of employer, then it will look as if government employees should be paid more–because they work for a large employer. If you don\’t make such an adjustment, then pay for government workers will look relatively higher. The CBO study, for example, found that if one adjusts for size of employer (and education), federal employees get wages that are 2% above private sector workers, but if you don\’t adjust for size of employer, then federal employees get wages 9% above those of private sector workers.

 4) A larger share of public-sector workers are unionized, and we know that unionized workers are paid more than other workers with equivalent skills. Again, the essential question here is whether a being unionized represents a skill set that the worker would take with them into private-sector employment, or not. I think it\’s more plausible to say that being unionized isn\’t a a portable \”skill set.\” If a study adjust for unionization, it will tend to find that government workers deserve to be paid more.

5) One problem in comparing government and private sector jobs is that the jobs themselves can be so different. For example, think of jobs in the education sector: Private sector jobs in this area tend to be either with preschool children or with college students and adults, while public-sector jobs in this area tend to be with K-12 students. Conversely, most jobs in sales or manufacturing are in the private sector, with very few in the public sector. If the researcher tries to adjust for the exact kind of job, comparing public and private-sector workers become difficult or even impossible. But if the researcher doesn\’t adjust in some way for the sector of the economy, you end up comparing workers who are in potentially quite different industries. An in-between approach here is to adjust for broad sectors–like \”education\” or \”services\”–but not to try to adjust for highly specific job categories.

6) Pay scales are more compressed in government. As emphasized in the CBO study, workers with lower levels of skill are on average paid more in government work than in the private sector, but worker with higher levels of skill are on average paid less. An overall comparison between all government and all private-sector workers will miss this distinction.

7) Job are more than wages and benefits. For example, government jobs in the Great Recession and its aftermath have tended to be more secure than private-sector jobs, in the sense of a lower chance of being laid off and a lower chance of pay cuts. Jobs have other characteristics, too. Certain jobs pose greater health risks, like mining and manufacturing in the private sector, or police and firefighters in the civilian public sector. I\’m not aware of studies that make a serious effort to value and adjust for these kinds of factors.

In short, when looking at a study comparing government and private-sector pay, run down this checklist of
seven points and it will tell you something about whether the study is likely to be leaning in one direction or another.

conversableeconomist

For-Profit Higher Education

It has been common for some years for politicians and educators to vow that America will greatly increase the proportion of students attending college. For example, in a speech to Congress on February 24, 2009, President Obama said: \”I ask every American to commit to at least one year or more of higher education or career training.  This can be community college or a four-year school; vocational training or an apprenticeship.  But whatever the training may be, every American will need to get more than a high school diploma.  And dropping out of high school is no longer an option.  It’s not just quitting on yourself, it’s quitting on your country – and this country needs and values the talents of every American.  That is why we will provide the support necessary for you to complete college and meet a new goal:  by 2020, America will once again have the highest proportion of college graduates in the world.\”

But how will the United States structure and pay for this expanded college attendance? For example, one policy approach would be to plan for dramatic expansion of enrollment in existing state colleges and universities, but with very tight state budgets, this isn\’t happening to any great extent. Instead, the answer that seems to be evolving, without ever really being enunciated and debated, is that the federal government will finance a dramatic expansion of higher education through an expansion of student loans, and because of limits on the number of slots at existing public colleges and universities, many students will take those loans to the for-profit higher education sector. In the Winter 2012 issue of my own Journal of Economic Perspectives, freely available on-line courtesy of the American Economic Association, Deming, David J., Claudia Goldin, and Lawrence F. Katz discuss the tradeoffs of this choice in \”The For-Profit Postsecondary School Sector: Nimble Critters or Agile Predators?\”

As the authors point out, the time since 2000 is \”a period when enrollment in the for-profit sector tripled while enrollment for the rest of higher education increased by just 22 percent. The solid dark line shows
that the fraction of fall enrollments accounted for by the for-profits increased from 4.3 percent in 2000 to 10.7 percent in 2009.\” They point out that \”almost 90 percent of the increase in for-profit enrollments during the last decade occurred because of the expansion of for-profit chains,\” where a \”chain\” is defined as an institution that operates across states or has more than five branches within a state.\” 

Along with the flexibility to expand enrollments, for-profit higher education has shown considerable flexibility in teaching groups not well-served by traditional higher education. \”African Americans
account for 13 percent of all students in higher education, but they are 22 percent of those in the for-profit sector. Hispanics are 11.5 percent of all students but are 15 percent of those in the for-profit sector. Women are 65 percent of those in the for-profit sector. For-profit students are older: about 65 percent are 25 years and older, whereas just 31 percent of those at four-year public colleges are, and 40 percent of those at two-year colleges are.\” In addition, for-profits are typically non-selective institutions, requiring only a high school diploma or a GED certificate.


For-profit institutions also have been quite flexible in providing the kinds of career-oriented classes that many students want: \”For-profit programs typically are not meant to prepare students to continue to another form of higher education, as is the case with most community colleges. Rather, the for-profits almost always offer training for a vocation or trade…. Although 5 percent of all BAs are granted by for-profit institutions, 12 percent of all BAs in business, management, and marketing are. Other large for-profit  BA programs are in communications (52 percent of all BAs in communications are granted by for-profits), computer and information sciences (27 percent), and personal and culinary services (42 percent). … Among AA degrees just two program groups—business, management, and marketing, and the health professions—account for 52 percent of all degrees. In the BA group, the business program produces almost 50 percent of the total. Among certificates granted in the Title IV for-profit sector, health professions and personal and culinary services account for 78 percent of certificate completers.\”




Of course, there is controversy over for-profit higher education. The big firms that dominate the sector are, well, for-profit, and they tend to pay their executives well and their faculty not-so-well. A high proportion of students at for-profit institutions are financing their studies with debt, and if they don\’t complete the degree–which can be a big problem at some for-profit institutions–the students are still on the hook for that debt. Total student debt now exceeds the total amount of credit card debt, and will soon top $1 trillion. In their well-balanced essay, Deming, Goldin, and Katz discuss these issues.

But to repeat my earlier point, we have apparently made a social decision that a combination of student loans and for-profit institutions is the primary method by which the United States will raise college admissions. I would like to see a competitive response to the for-profits from public-sector non-profit higher education. I\’d love to see the public sector aggressively increasing non-selective enrollments, offering on-line classes and flexible meeting times for nontraditional college students, using technology and nontenured lecturers aggressively to hold down costs, and expanding certificate and degree programs with a focus on what is demanded in the market. This kind of institutional change is undoubtedly difficult.  But with a few local exceptions, the public higher education sector is reacting too much much like Kodak when that company was first confronted with low-cost competition for film and then with the change to digital photography–and the firm was too slow to adapt. 

I know that for-profit higher education has its warts and flaws. But so far, the not-for-profit higher education sector has not shown that it is serious about being flexible or entrepreneurial in way that can meet the goal of expanding college enrollment.

 

conversableeconomist

Same Income, Varying Taxes: ERP #4

This is the fourth of four posts based on figures from the 2012 Economic Report of the President. For the first post and an overview, start here.

Amid the complexity and confusion of the U.S. income tax code, it\’s quite possible for people with similar levels of income to pay widely varying level of tax. For illustration, consider the table. The rows of the table divide up the U.S. income distribution into fifths, or quintiles. The last row shows results for the top 1% of the income distribution.  For each quintile–and for the top 1%–the columns of the table then tell about the distribution of taxes for that group.

For example, if one looks at the distribution of average tax rates for the bottom quintile, households at the 10th percentile of that distribution have a federal tax rate of -13.7% (that is, they receive refundable tax credits from the government). In the bottom quintile of the income distribution, those at the median pay 5.4% of income in federal taxes. (These calculations include income taxes and payroll taxes.)

Or look at the top 1%. Given the distribution of federal taxes for that group, the household at the 10th percentile of tax payments for this group pays 8.7% of income in federal taxes (presumably due to substantial tax-free investments, or perhaps to carrying forward losses from a previous tax year that count against income in this year). However, a household in the top 1% of the income distribution and the 90th percentile of the tax distribution for this group pays an average federal tax rate of 34.6%.

At least for me, there is something of a tendency when looking at tables like this one to feel as if some of those with high income are paying too little, and some of those with low incomes are paying too much. And maybe that quick reaction is correct. But the tax code has so many rules and provisions and exceptions and situations, that it\’s s also possible that if I knew the actual details of some of these taxpayers, the outcome would seem fairly reasonable to me.

The broader point here is that when a tax code becomes enormously complex and lengthy, it is also going to allow the possibility of considerable variation in taxes paid even for those with similar incomes. Even if all the individual provisions of such a tax code are defensible (an enormous \”if\”!), the tax code as a whole is likely to end up appearing arbitrary and unfair.

conversableeconomist

Job Market Churn is Slowing: ERP #3

This is the third of four posts based on figures from the 2012 Economic Report of the President. For the first post and an overview, start here.

The U.S. job market has long been famous for its \”churn\” — that is, the simultaneously large inflows and outflows out of jobs which suggest a fluid and adjustable labor market. Thus, it\’s disturbing to observe a long-term trend toward less churn in the U.S. labor market. Here\’s a figure using the Business Dynamics Statistics from the Bureau of Labor Statistics:

Here\’s commentary from the Economic Report of the President: \”The rates of both gross gains and gross losses have been declining over time. Whereas, on average, 18.2 percent of private-sector jobs in the 1980s were newly created positions in startups or expanding firms, gross job gains fell to 16.8 percent of total private-sector employment in the 1990s and to 15.8 percent between 2000 and 2009 (Figure 6-3). Similarly, gross job losses were slightly more than 16.2 percent of overall private-sector employment in the 1980s but fell to 14.9 percent in the 1990s and then remained largely the same between 2000
and 2009. These secular declines also are apparent when one focuses more narrowly on startups.\”

Here\’s a similar pattern from another source: quarterly data from the Business Employment Dynamics (BED) survey.

What explains this drop in job churn over time, and is it a cause for concern? The report says (citations omitted): \”Now that researchers have documented the long-term secular slowdown in job gains and losses, the underlying reasons for the slowdown and its implications for the future of the U.S. economy are fast becoming the subject of an active debate. One possible reason for the slowdown in job reallocation is the aging of the population. Older workers may be less likely to become entrepreneurs, and research has documented a positive correlation between worker age and job tenure.\”

An aging workforce probably is part of the explanation. But one also wonders if there isn\’t another dynamic at work: for a variety of reasons, it may be getting harder to start up a business in the United States, and harder to be an employer. In turn, workers perceive fewer outside opportunities, and become more likely to stick with their present job. Or perhaps the U.S. labor market is becoming less fluid and adjustable in other ways.

conversableeconomist

Why Wasn\’t the Risk of a Housing Price Decline Taken Into Account? ERP #2

This is the second of four posts based on figures from the 2012 Economic Report of the President. For an overview and the first post, start here.

When I give talks about the causes of the recession, people often shake their heads in disbelief at the thought that few investors were taking the risk of a housing price decline into account. I think this disbelief is a case of 20:20 hindsight. Investors didn\’t take the risk of a national housing price decline into account  because they hadn\’t seen anything like it before. Here\’s a figure showing comparing the housing price declines nationwide during the Great Depression,  and then comparisons with more local housing price declines in Boston in 1989 and in California in 1990. Sure, investors knew that housing prices could nosedive in a local market, and some were even braced for the possibility of a modest housing price decline nationwide. But for the country as a whole, predicting in 2005 or 2006 that national housing prices would drop much farther and faster than during the Great Depression would have been a prediction outside all historical experience. I admire the clairvoyance of the few who truly saw it coming, but I can\’t reasonably blame those who didn\’t.


The evidence does offer some hints that the decline in U.S. housing prices may be just about over. For example, one measure of a price bubble is to look at the ratio of housing prices to rents. When this ratio rises sharply, then it may be a signal that prices are getting out of line. But that ratio has now fallen back to pre-crisis levels. Similarly, the ratio of mortgage value per home-owning household has trended up over time, as the economy has grown. During the housing price bubble, that trend launched like a rocket, but now it has been flat for a few years, and is not too far out of line with the longer-term trend.

At the most basic level, the national average of actual housing prices does seem to have levelled out since early 2009. Indeed, futures markets that were predicting a continued drop in housing prices as of January 2009 have seen housing prices hold up better than expected at that time.

conversableeconomist

The Relatively Mild U.S. Financial Recession: ERP #1

I always enjoy looking through the annual Economic Report of the President, but I confess that I impose a couple of rules. I focus almost entirely on the figures and tables, and how they are discussed in the text. I ignore all economic projections for the future, and all comments about specific policies of the current administration. At least for me, this approach is useful in stripping away the politics, and focusing instead on some vivid facts and analysis. I\’ll offer four posts today using figures from the 2012 ERP:

  1. The Relatively Mild U.S. Financial Recession
  2. Why Wasn\’t the Risk of a Housing Price Decline Taken Into Account?
  3. Job Market Churning is Slowing
  4. Same Income, Varying Taxes

The Great Recession has been brutally deep, and the aftereffects seem likely to persist for at least five years after it officially ended in June 2009 (by the dating of the National  Bureau of Economic Research). But in the context of financial recessions in other countries, the U.S. experience actually doesn\’t look so bad. Here\’s a table comparing the behavior of real GDP across 14 recessions associated with financial crises. The average peak-to-trough decline is a drop of 10.2%; in the U.S., the decline was 5.1%. The average length of these recessions was 6.6 quarters; in the U.S., peak-to-trough was 6 quarters.


The rise in U.S. unemployment rates has been similar to that in other financial crisis in this comparison group, but believe it or not, somewhat less prolonged. The next table shows the rising U.S. unemployment rate over time from the business cycle peak compared with the average of the other countries in the comparison group. The figure after that shows a country-by-country comparison of the total rise in the unemployment rate.

Even the pattern of the U.S. economic recovery, sluggish as it has been, has basically followed the time profile of the 14 comparison countries.

Sometimes people talk about the depth of the Great Recession as if it really couldn\’t have been any worse–as if the very depth of the recession and the sustained proves that macroeconomic policy to counter the recession was necessarily ineffective. The conclusion does not necessarily follow, of course. To me, these comparisons offer some (admittedly impressionistic) evidence that the monetary policy steps taken by the federal government–the huge budget deficits on the fiscal side, along with  the near-zero federal funds interest rates and quantitative easing on the monetary side–did have beneficial effects. The Great Recession and its aftermath have been gruesome, but without the aggressive fiscal and monetary policy response, it could have been even worse.

conversableeconomist

Winter 2012 Journal of Economic Perspectives

The Winter 2012 issue of my own Journal of Economic Perspectives is now up on the web. Courtesy of the American Economic Association, this issue and indeed back issues of the journal all the way back through 1994 are freely available on the web. I\’ll be blogging about some of these papers over the next week or so, but for now, here\’s the table of contents and and abstract for each article.

Symposium: Energy Challenges

\”Is There an Energy Efficiency Gap?\” by  Hunt Allcott and Michael Greenstone

Many analysts of the energy industry have long believed that energy efficiency offers an enormous \”win-win\” opportunity: through aggressive energy conservation policies, we can both save money and reduce negative externalities associated with energy use. In 1979, Daniel Yergin and the Harvard Business School Energy Project estimated that the United States could consume 30 or 40 percent less energy without reducing welfare. The central economic question around energy efficiency is whether there are investment inefficiencies that a policy could correct. First, we examine choices made by consumers and firms, testing whether they fail to make investments in energy efficiency that would increase utility or profits. Second, we focus on specific types of investment inefficiencies, testing for evidence consistent with each. Three key conclusions arise: First, the evidence presented in the long literature on the subject frequently does not meet modern standards for credibility. Second, when one tallies up the available empirical evidence from different contexts, it is difficult to substantiate claims of a pervasive Energy Efficiency Gap. Third, it is crucial that policies be targeted. Welfare gains will be larger from a policy that preferentially affects the decisions of those consumers subject to investment inefficiencies.
Full-Text Access 

\”Creating a Smarter U.S. Electricity Grid,\” by Paul L. Joskow

This paper focuses on efforts to build what policymakers call the \”smart grid,\” involving 1) improved remote monitoring and automatic and remote control of facilities in high-voltage electricity transmission networks; 2) improved remote monitoring, two-way communications, and automatic and remote control of local distribution networks; and 3) installation of \”smart\” metering and associated communications capabilities on customer premises so that customers can receive real-time price information and/or take advantage of opportunities to contract with their retail supplier to manage the consumer\’s electricity demands remotely in response to wholesale prices and network congestion. I examine the opportunities, challenges, and uncertainties associated with investments in \”smart grid\” technologies. I discuss some basic electricity supply and demand, pricing, and physical network attributes that are critical for understanding the opportunities and challenges associated with expanding deployment of smart grid technologies. Then I cover issues associated with the deployment of these technologies at the high voltage transmission, local distribution, and end-use metering levels.
Full-Text Access 

\”Prospects for Nuclear Power,\” by Lucas W. Davis

Nuclear power has long been controversial because of concerns about nuclear accidents, storage of spent fuel, and how the spread of nuclear power might raise risks of the proliferation of nuclear weapons. These concerns are real and important. However, emphasizing these concerns implicitly suggests that unless these issues are taken into account, nuclear power would otherwise be cost effective compared to other forms of electricity generation. This implication is unwarranted. Throughout the history of nuclear power, a key challenge has been the high cost of construction for nuclear plants. Construction costs are high enough that it becomes difficult to make an economic argument for nuclear even before incorporating these external factors. This is particularly true in countries like the United States where recent technological advances have dramatically increased the availability of natural gas. The chairman of one of the largest U.S. nuclear companies recently said that his company would not break ground on a new nuclear plant until the price of natural gas was more than double today\’s level and carbon emissions cost $25 per ton. This comment summarizes the current economics of nuclear power pretty well. Yes, there is a certain confluence of factors that could make nuclear power a viable economic option. Otherwise, a nuclear power renaissance seems unlikely.
Full-Text Access 

\”The Private and Public Economics of Renewable Electricity Generation,\” by Severin Borenstein

Generating electricity from renewable sources is more expensive than conventional approaches but reduces pollution externalities. Analyzing the tradeoff is much more challenging than often presumed because the value of electricity is extremely dependent on the time and location at which it is produced, which is not very controllable with some renewables, such as wind and solar. Likewise, the pollution benefits from renewable generation depend on what type of generation it displaces, which also depends on time and location. Without incorporating these factors, cost-benefit analyses of alternatives are likely to be misleading. Other common arguments for subsidizing renewable power—green jobs, energy security, and driving down fossil energy prices—are unlikely to substantially alter the analysis. The role of intellectual property spillovers is a strong argument for subsidizing energy science research, but less persuasive as an enhancement to the value of installing current renewable energy technologies.
Full-Text Access 

\”Reducing Petroleum Consumption from Transportation,\” by Christopher R. Knittel

The United States consumes more petroleum-based liquid fuel per capita than any other OECD high-income country—30 percent more than the second-highest country (Canada) and 40 percent more than the third-highest (Luxembourg). The transportation sector accounts for 70 percent of U.S. oil consumption and 30 percent of U.S. greenhouse gas emissions. Taking the externalities associated with high U.S. gasoline consumption as largely given, I focus on understanding the policy tools that seek to reduce this consumption. I consider four main channels through which reductions in U.S. oil consumption might take place: 1) increased fuel economy of existing vehicles, 2) increased use of non-petroleum-based, low-carbon fuels, 3) alternatives to the internal combustion engine, and 4) reduced vehicle miles traveled. I then discuss how these policies for reducing petroleum consumption compare with the standard economics prescription for using a Pigouvian tax to deal with externalities. Taking into account that energy taxes are a political hot button in the United States, and also considering some evidence that consumers may not \”correctly\” value fuel economy, I offer some thoughts about the margins on which policy aimed at reducing petroleum consumption might usefully proceed.
Full-Text Access

\”How Will Energy Demand Develop in the Developing World?\” by Catherine Wolfram, Orie Shelef and Paul Gertler

Over the next 25 to 30 years, nearly all of the growth in energy demand, fossil fuel use, associated local pollution, and greenhouse gas emissions is forecast to come from the developing world. This paper argues that the world\’s poor and near-poor will play a major role in driving medium-run growth in energy consumption. As the world economy expands and poor households\’ incomes rise, they are likely to get connected to the electricity grid, gain access to good roads, and purchase energy-using assets like appliances and vehicles for the first time. We argue that the current forecasts for energy demand in the developing world may be understated because they do not accurately capture growth in demand along the extensive margin, as low-income households buy their first durable appliances and vehicles. Within a country, the adoption of energy-using assets typically follows an S-shaped pattern: among the very poor, we see little increase in the number of households owning refrigerators, vehicles, air conditioners, and other assets as incomes go up; above a first threshold income level, we see rapid increases of ownership with income; and above a second threshold, increases in ownership level off. A large share of the world\’s population has yet to go through the first transition, suggesting there is likely to be a large increase in the demand for energy in the coming years.
Full-Text Access

Symposium: Higher Education

\”The For-Profit Postsecondary School Sector: Nimble Critters or Agile Predators?\” by David J. Deming, Claudia Goldin and Lawrence F. Katz

Private for-profit institutions have been the fastest-growing part of the U.S. higher education sector. For-profit enrollment increased from 0.2 percent to 9.1 percent of total enrollment in degree-granting schools from 1970 to 2009, and for-profit institutions account for the majority of enrollments in non-degree-granting postsecondary schools. We describe the schools, students, and programs in the for-profit higher education sector, its phenomenal recent growth, and its relationship to the federal and state governments. Using the 2004 to 2009 Beginning Postsecondary Students (BPS) longitudinal survey, we assess outcomes of a recent cohort of first-time undergraduates who attended for-profits relative to comparable students who attended community colleges or other public or private non-profit institutions. We find that relative to these other institutions, for-profits educate a larger fraction of minority, disadvantaged, and older students, and they have greater success at retaining students in their first year and getting them to complete short programs at the certificate and AA levels. But we also find that for-profit students end up with higher unemployment and \”idleness\” rates and lower earnings six years after entering programs than do comparable students from other schools and that, not surprisingly, they have far greater default rates on their loans.
Full-Text Access 

\”Student Loans: Do College Students Borrow Too Much–Or Not Enough?\” by Christopher Avery and Sarah Turner

Total student loan debt rose to over $800 billion in June 2010, overtaking total credit card debt outstanding for the first time. By the time this article sees print, the continually updated Student Loan Debt Clock will show an accumulated total of roughly $1 trillion. Borrowing to finance educational expenditures has been increasing—more than quadrupling in real dollars since the early 1990s. The sheer magnitude of these figures has led to increased public commentary on the level of student borrowing. We move the discussion of student loans away from anecdote by establishing a framework for considering the use of student loans in the optimal financing of collegiate investments. From a financial perspective, enrolling in college is equivalent to signing up for a lottery with large expected gains—indeed, the figures presented here suggest that college is, on average, a better investment today than it was a generation ago—but it is also a lottery with significant probabilities of both larger positive, and smaller or even negative, returns. We look to available—albeit limited—evidence to assess which types of students are likely to be borrowing too much or too little.
Full-Text Access 

\”American Higher Education in Transition,\” by Ronald G. Ehrenberg

American higher education is in transition along many dimensions: tuition levels, faculty composition, expenditure allocation, pedagogy, technology, and more. During the last three decades, at private four-year academic institutions, undergraduate tuition levels increased each year on average by 3.5 percent more than the rate of inflation; the comparable increases for public four-year and public two-year institutions were 5.1 percent and 3.5 percent, respectively. Academic institutions have also changed how they allocate their resources. The percentage of faculty nationwide that is full-time has declined, and the vast majority of part-time faculty members do not have Ph.D.s. The share of institutional expenditures going to faculty salaries and benefits in both public and private institutions has fallen relative to the share going to nonfaculty uses like student services, academic support, and institutional support. There are changing modes of instruction, together with different uses of technology, as institutions reexamine the prevailing \”lecture/discussion\” format. A number of schools are charging differential tuition across students. This paper discusses these various changes, how they are distributed across higher education sectors, and their implications. I conclude with some speculations about the future of American education.
Full-Text Access 

Articles

\”Compensation for State and Local Government Workers,\” Maury Gittleman and Brooks Pierce

Are state and local government workers overcompensated? In this paper, we step back from the highly charged rhetoric and address this question with the two primary data sources for looking at compensation of state and local government workers: the Current Population Survey conducted by the Bureau of the Census for the Bureau of Labor Statistics, and the Employer Costs for Employee Compensation microdata collected as part of the National Compensation Survey of the Bureau of Labor Statistics. In both data sets, the workers being hired in the public sector have higher skill levels than those in the private sector, so the challenge is to compare across sectors in a way that adjusts suitably for this difference. After controlling for skill differences and incorporating employer costs for benefits packages, we find that, on average, public sector workers in state government have compensation costs 3-10 percent greater than those for workers in the private sector, while in local government the gap is 10-19 percent. We caution that this finding is somewhat dependent on the chosen sample and specification, that averages can obscure broader differences in distributions, and that a host of worker and job attributes are not available to us in these data. Nonetheless, the data suggest that public sector workers, especially local government ones, on average, receive greater remuneration than observably similar private sector workers. Overturning this result would require, we think, strong arguments for particular model specifications, or different data.
Full-Text Access 

\”Recommendations for Further Reading,\” by Timothy Taylor
Full-Text Access

conversableeconomist

Are the New Auto Fuel Economy Standards For Real?

Politicians are predisposed to like a technology standard, like the Corporate Average Fuel Economy (CAFE) standards for automobile miles-per-gallon, as a way of holding down petroleum use. After all, it sounds a lot better to voters than enacting a gasoline tax or a carbon tax! Pass a law that better-mileage cars will be phased in over the next decade or two, and politicians can boast of their great achievement –sidestepping the fact that promised aren\’t achievements and rules are made to be changed.

Thus, when I heard about the plans for a dramatic increase in CAFE standards, I was skeptical. In the most recent issue of my own Journal of Economic Perspectives, Christopher R. Knittel discusses various aspects of  \”Reducing Petroleum Consumption from Transportation.\” As Knittel writes: \”A new CAFE standard in place for 2011 seeks to increase average fuel economy to roughly 34.1 miles per gallon by 2016. The Environmental Protection Agency and Department of Transportation are currently in the rule-making process for model years 2017 and beyond, with President Obama and 13 automakers agreeing to a standard of 54.5 MPG by 2025.\” Knittel provides evidence to back up my skepticism about the past use of CAFE standards, but he also argues that those future standards–unbelieveable as they may at first appear–are technologically achievable.

Back in 1975, against a backdrop of a dramatic rise in oil prices and concern over dependence on imported oil, the U.S. enacted the Corporate Average Fuel Economy (CAFE) law, requiring that over the average of their cars sold by each company, the average had to start at 18 miles-per-gallon, and then rise to 27 mpg by 1985. Higher gasoline prices provided a strong inducement for people to buy these more fuel-efficient cars, but when gasoline prices dropped in the mid-1980s, the CAFE standards stagnated. Here\’s a figure from Knittel showing how CAFE standards flattened out after 1985–and also showing the planned increase to take place.



The lack of any increase in the CAFE standards was only part of the story. Knittel explains: \”[T]wo features of the original CAFE standards reduced their effect. First, sport-utility vehicles were treated as light trucks, and thus could meet a lower miles-per-gallon standard than cars. Perhaps not coincidentally, in 1979 light trucks comprised less than 10 percent of the new vehicle fleet, but this share rose steadily
and peaked in 2004 at 60 percent. Second, vehicles with a gross vehicle weight of over 8,500 pounds, which includes many large pickup trucks and sports-utility vehicles, were exempt from CAFE standards.\”

Taking these factors together, actual fuel economy for the U.S. fleet of cars hasn\’t been rising much, although it has edged up in the last few years with higher gasoline prices. My own interpretation is that the CAFE standards effectively became nonbinding–that is, they weren\’t pushing anyone to buy a different car than they otherwise would have purchased, and they weren\’t adding to fuel economy. Here\’s the data:

So, is it technologically possible meet the future increase in miles-per-gallon standards? Knittel argues \”yes.\” He points out: \”By world standards, these [currently existing] miles-per-gallon standards are not aggressive. After accounting for differences in the testing procedures, the World Bank estimated that the European Union standard was roughly 17 MPG more stringent in 2010 than the U.S. standard …\”

Moreover, Knittel has carried out a series of studies looking at technological progress in cars, and the tradeoffs between weight, engine power, and fuel efficiency. He finds: \”On average, a vehicle with a given weight and engine power level has a fuel economy that is 1.75 percent higher than a vehicle with the same weight and horsepower level from the previous year. …In the medium run, automakers can adjust vehicle attributes by trading off weight and horsepower for increased fuel economy. In Knittel (2011), I find that reducing weight by 1 percent increases fuel economy by roughly 0.4 percent, while reducing horsepower and torque by 1 percent increases fuel economy by roughly 0.3 percent.\”

By Knittel\’s calculation, getting from the new-car average fuel economy standard of 29 mpg in 2010 to 34.1 mpg in 2016 is do-able. If technological progress continues to improve mileage by 1.75% per year, and ways are found to reduce weight and engine power by about 6%, the standard for 2016 is achieveable.

But what about that planned standard of 54.5 mpg by 2025? Knittel explains that the number is somewhat inflated: \”Taken literally, it would require fundamental changes to rates of technological progress and/or the size and power of vehicles. The 2025 number is a bit misleading. In the law, the 54.5 miles-per-gallon standard is based on a calculation from the Environmental Protection Agency based on carbon dioxide tailpipe emissions. It also includes credits for many technologies including plug-in hybrids, electric and hydrogen vehicles, improved air conditioning effifi ciency, and others. On an apples-to-apples basis, Roland (2011) cites some industry followers that claim that the actual new fleet fuel economy standard in 2025 is more like 40 miles per gallon. Achieving 40 miles per gallon by 2025 is certainly possible. At a rate of technological progress of 1.75 percent per year, 40 miles per gallon requires additional reductions in weight and engine power of less than 7 percent.\”

But although the planned mileage standards do appear–to my surprise–technologically feasible, it remains to be seen whether they are politically feasible, and also whether they are even a sensible public policy idea.

On the political side, the U.S. political system found a way for most of the last three decades to have fuel economy standards on the books as a matter of law and public relations–but to have standards with very little bite. Let\’s see whether the fuel economy standards planned for the future actually cause some real changes in the U.S. auto fleet, or whether they are quickly riddled with exceptions.

But at a deeper level, it\’s not even clear that fuel economy standards are a good policy idea. Knittel explains: \”At a basic level, it focuses on the wrong thing—fuel economy instead of total fuel consumption. CAFE only targets new vehicles and leads to subsidies for some vehicles. Finally, CAFE pushes consumers into more-fuel-efficient vehicles without changing the price of fuel, leading to more miles traveled. The empirical size of this last effect, known as “rebound,” is a matter of ongoing research,
but to the extent that rebound occurs, it necessarily leads to greater congestion, accidents, and criteria pollutant emissions relative to the status quo.\” A considerable body of economic research suggests that if your policy goal is to reduce petroleum consumption, a gasoline tax or a carbon tax accomplishes the goal at a far lower social cost than fuel economy standards–although for politicians the explicitness of that cost seems to make it a nonstarter.

For more discussion of this topic, I recommend \”Automobile Fuel Economy Standards: Impacts, Efficiency, and Alternatives,\” by Soren T. Anderson, Ian W. H. Parry, James M. Sallee, and Carolyn Fischer, in the Winter 2011 issue of the Review of Environmental Economics and Policy.  The publisher has made article freely available here.
 

conversableeconomist

The Big Decline in Housing Segregation

Edward Glaeser and Jacob Vigdor document and discuss \”The End of the Segregated Century: Racial Separation in America\’s Neighborhoods, 1890-2010,\” in Civic Report #66 written for the Manhattan Institute for Policy Research.  
Here\’s a headline graph. Take the two most common measures of residential segregation, the \”dissimilarity index\”  and the \”isolation index\” (both explained further in a moment). Apply them to the 10 largest American cities using Census data The pattern that emerges is a large increase in residential segregation  from about  1910 to 1950, segregation remaining at that  high level from about 1950 to 1970, and then a sharp decline in residential segregation from 1970 up through 2010.

Glaeser and Vigdor summarize the pattern in this way: \”Segregation has declined steadily from its mid-century peak, with significant drops in every decade since 1970. As of 2010, the separation of African-Americans from individuals of other races stood at its lowest level in nearly a century. Fifty years ago, nearly half the black population lived in what might be termed a “ghetto” neighborhood, with an African-American share above 80 percent. Today, that proportion has fallen to 20 percent.\”
How is residential segregation measured? 
\”Residential segregation can be measured in a variety of ways. The most common method is to form an index that summarizes the level of segregation in a metropolitan area on a scale from zero, where every neighborhood is just as diverse as the entire region, to 100, where individuals of different races never share neighborhoods. Indices differ according to their coding of intermediate situations … Some indices require more detailed geographical data than others, with the most sophisticated using census information collected on a block-by-block basis. 
\”This report focuses on two measures—the dissimilarity index and the isolation index—both of which have a long history in social-scientific writing on segregation. The two measures together adequately summarize segregation, being highly correlated with more sophisticated indices, while being simple enough to calculate that even data from the late nineteenth century are sufficiently rich to permit their computation.\”
 What is intuitive interpretation  of a dissimilarity index? 
\”The dissimilarity index measures the extent to which two groups are found in equal proportion in all neighborhoods. It can be interpreted as the proportion of individuals of either group that would have to change neighborhoods in order to achieve perfect integration. It is the most commonly used segregation measure, first introduced into the sociology literature shortly after World War II.\”
 What is the intuitive interpretation of an isolation index? 
\”Dissimilarity is not a perfect measure. Consider the following scenario. There are two equal-size neighborhoods in a city: one is 100 percent white; and the other is 98 percent white and 2 percent black. According to the dissimilarity index, this city is fairly segregated, since about half of the black residents would need to move in order to achieve perfect integration. In an important sense, though, the black residents are not isolated—after all, they live in a neighborhood that is 98 percent white.
 
The isolation index is designed to distinguish this sort of scenario from one where neighborhoods have dramatically different racial character. It measures the tendency for members of one group to live in neighborhoods where their share of the population is above the citywide average. In this hypothetical example, black residents live in a neighborhood that is 2 percent black, which is just 1 percentage point higher than what would be expected under perfect integration. The isolation index would therefore be on the order of 1 percent, rather than 50 percent.\”

This paper is more about documenting the patterns than about inquiring into underlying causes, but Glaeser and Vigdor do offer an overview of causes.

In their telling, the dramatic rise in residential segregation from 1910 to about 1950 arose because it was a time of huge migration from rural areas to cities in the United States, and when African-Americans participated in this migration, they were met with a combination of white hostility and legal restrictions that pushed them into highly segregated neighborhoods. The decline in residential segregation over the last 40 years is a combination of factors: reductions in the legal and social barriers that enforced residential segregation; outflow of the African-American population from highly segregated neighborhoods; and in some parts of the country, inflow of Hispanic and Asian immigrants to neighborhoods that had been segregated before.

High levels of residential segregation have been an enormous and legitimate social concern, so the decline in those levels is worth noting and welcoming. But as Glaeser and Vigdor point out, the good news comes with a mournful undertone for those who remember some of the high hopes for housing desegregation back in the 1960s:

\”The 1960s were the heyday of racial segregation. During those years, segregation seemed a likely cause of many of the troubles afflicting African-Americans. Segregation was so enormous, and so unfair, that it seemed to create a separate and unequal experience for African-Americans everywhere. During those years, the fight against housing segregation seemed to offer the possibility that once the races mixed more readily, all would be well.

Forty years later, we know that this dream was a myth. There is every reason to relish the fact that there is more freedom in housing today than 50 years ago and to applaud those who fought to create that change. Yet we now know that eliminating segregation was not a magic bullet. Residential segregation has declined pervasively, as ghettos depopulate and the nation’s population center shifts toward the less segregated Sun Belt. At the same time, there has been only limited progress in closing achievement and employment gaps between blacks and whites. … While the decline in segregation remains good news, far too many Americans still lack the opportunity to achieve meaningful success.\”

conversableeconomist

Don\’t Know Nothing About Military Strategy

I don\’t know anything about issues like the appropriate number of soldiers in the armed forces, or which weapons systems are really needed, or where forces should be based around the world. But I do know something about federal budget numbers. As the political debate unfolds over appropriate levels of defense spending in the years ahead, here are some historical and international perspectives.

Measured as a share of GDP, U.S. defense spending is down substantially from the figures of 10% or more that often occurred in the 1950s and 1960s, and down from the 6% reached during the Reagan defense build-up of the 1980s. Although defense spending as a share of GDP has nudged up since September 11, 2001, it was 4.7% of GDP in 2011.

Defense spending also doesn\’t dominate the federal budget as it once did. Back in January 1961, in President Dwight Eisenhower\’s farewell address, he warned of the dangers of the \”military-industrial complex.\” But at that time, defense spending was still almost 10% of GDP and more than half of total federal spending. Indeed, defense spending was as much as 70% of all federal spending back in the early 1950s (and higher than that at the peak of WWII). But for the last two decades, defense spending has dropped to about 20% of all federal spending.

While U.S. defense spending at relatively low levels, historically speaking, both relative to GDP and relative to total federal spending, it remains high relative to spending by other countries. Here\’s a table showing that U.S. defense spending is more than 40% of the world total, and that U.S. defense spending comfortably exceeds the sum of defense spending by the next 10 largest spenders.

The U.S. economy is the largest in the world, and it also spends one of the greatest shares of that economy on defense. By my count, only eight countries in the world (for which SIPRI has data) spend a larger share of their GDP on defense than does the U.S.:  Saudi Arabia, 11.2%; Chad, 6.2%; Georgia, 5.6%; Iraq, 5.4%; Israel, 6.3%; Jordan, 6.1%; Oman, 9.7%; and UAE, 7.3%.

Of course, spending isn\’t the only variable that matters in national defense: strategy, diplomacy, ideology, economic ties, even personal cross-border ties can affect the likelihood and extent of instability. Defense spending can be quite good at projecting certain kinds of power, but not especially useful at blocking a biological or nuclear weapon that fits in a panel truck or even a large suitcase. That said, these sorts of numbers cut both directions in the debate over levels of defense spending. Those who favor reductions in defense spending over time might take note of the fact that we haven\’t been living in Eisenhower\’s world for some time, and U.S. defense spending has a smaller share of the economy and of federal spending than the historical norm. Those who favor higher defense spending might take note of the fact that the U.S. is far and away the largest defense spending nation now–and that many of the other largest spenders are our allies.

For both sides, I\’m always interested not just in hearing argument about \”more\” or \”less,\” but about what is enough. If you prefer cutting defense, how low would you go before you would say \”enough\”? If you prefer increasing defense, how high would you go before you would say \”enough\”? If someone can\’t explain their answer to that question, I suspect that underneath their show of confident certainty, they don\’t really know any more about weighing the costs and benefits of military spending than I do.

Thanks for Danlu Hu for putting together the U.S. defense spending figures over time from the Historical Tables of the President\’s Budget for 2013.