The recent work of CEPR’s John Schmitt and Janelle Jones shines a harsh spotlight on the dramatic decline in “good jobs” over the last generation. In Where Have All the Good Jobs Gone?, Schmitt and Jones show that the share of good jobs (defined by an earnings threshold of $18.50/hr and the job-based provision of health coverage and a retirement plan) has fallen—even as the age and educational attainment of the workforce has risen.
The interactive graph below summarizes their findings: Select any combination of demographics (all workers, women, men) in the upper right; and any combination of “good job” elements (earnings, health coverage, retirement plan) in the bottom pane. Start with the earnings threshold for all groups. Here we see decent gains for women, although no more than we might expect given gains in labor productivity and educational attainment over the same span. The share of men at this threshold, by contrast, falls—from about 57.5 percent in 1979 to about 54.5 percent in 2010. Selecting a single demographic underscores the contributing factors. For men, declining pension and health coverage (combined with flat earnings) led to steep decline in the good job share (from 37.5 percent in 1979 to 27.7 percent in 2010). For women, health coverage has fallen less dramatically (from a lower starting point), and pension coverage is pretty flat—and together they have dampened but not erased the gains in earnings (yielding a modest increase in share of good jobs, from 12.4 percent to 21.1 percent).
Sixteen years ago on Wednesday, President Bill Clinton surrendered to House Speaker Newt Gingrich and signed NewtAid, legislation that replaced the Social Security Act's Aid to Families with Dependent Children (AFDC) with a right-wing block grant scheme called Temporary Assistance for Needy Families (TANF). NewtAid/TANF was prematurely lauded as a success before it had even been fully implemented.
It is now clear TANF is a failed program that needs to be overhauled. NewtAid's failure can be seen most simply by comparing the number of children living below federal poverty line in 1992 and 2010. (These years are compared because 2010 is the most recent year we have child poverty numbers for, and 1992 is, like 2010, the first calender year after the end of a recession.)
Number of Children Living in Families with Incomes Below the Federal Poverty Line
1992: 15.3 million
2010: 16.4 million
In sum, just over 1 million more children lived below the poverty line in 2010, more than a decade after TANF's implemetation, than before TANF's implementation in 1992. If AFDC had remained in place and reformed along progressive lines, the number of children living in poverty would be much lower today than it was before NewtAid.
AFDC was far from a perfect program—especially after Reagan-era budget cuts limited the support it provided for working parents—but it was one of the dependable pillars of our system of social insurance (although nowhere near as strong a pillar as Social Security). Instead of strengthening AFDC, NewtAid tore it down and replaced it with a radical conservative scheme that has:
Given states incentives to spend billions of dollars in public funds—funds that had previously been used to promote economic security and opportunity for low-income parents—in an unaccountable and often irresponsible manner. In fact, the bulk of public funds available to states under TANF today are not spent on child care, employment services, or helping families meet basic needs. Instead, states have diverted the bulk of funds to "other services." In the most notorious cases, states have diverted TANF funds to finance unaffordable tax cuts.
Failed to provide adequate information on how states use TANF funds. As GAO has found, we know very little about how states are using 70 percent of the funds provided by the $16.5 billion block grant, including not just "results" but things as basic as how many families are being helped with these funds.
Deeply cut the actual amount of resources available through AFDC/TANF to help struggling, working-class families. Because block grant funding has remained frozen at its 1997 level, its actual value (adjusted for inflation) has fallen by nearly 30 percent.
Of course the HAMP and various derivative programs were complicated. They were made even more so by the fact that mortgage servicers really weren't set up to do modifications. The servicer process was set up to minimize costs by establishing standardized procedures. These outfits knew how to collect monthly payments, send out late notices, start default proceedings, and carry through foreclosures. They had specified payments for each step in this process. Working out loan modifications was not on the list, so it's not surprising that they didn't do a very good job, especially when we factor in complications like second mortgages.
There was an easier route. Early on in the crisis (August of 2007) I proposed the first version of my Right to Rent plan. The basic point was very simple. It gave homeowners the option to stay in their home as renters for a designated period of time following foreclosure (at least 5 years in my ideal world) paying the market rent.
This would accomplish several important goals. First and foremost it would instantly give homeowners facing foreclosure a substantial degree of housing security. If they had kids in school, the rental period would likely be long enough to let them finish. It would also give them some time to get on their feet financially and make arrangements for appropriate housing.
Second, it would prevent a blight of foreclosures from ruining whole neighborhoods. Renters who effectively have long-term leases have almost as much incentive to maintain the property as owners. At the least they would keep homes occupied so that they would not be eyesores and possible havens for crime.
Third, this right would give banks more incentive to find ways to modify mortgages to keep people in their homes as owners. Banks would generally prefer having a foreclosed house free and clear rather than being a landlord for 5 years. By making the foreclosure option less attractive, Right to Rent would make banks more willing to consider alternatives.
Right to Rent would not have required any massive bureaucracy. It could have been handled within the same legal framework in which foreclosures were handled. The main cost would be the payment for appraisers who would determine the market rent on a foreclosed home. This is the same process that appraisers follow when determining the market price of a home for someone seeking a mortgage.
Right to Rent also raised few of the moral hazard or political issues associated with various plans to have the government pay down debt. The former homeowner hardly gets a bonanza in this story. They get a right to stay in their home that they would not have otherwise had, in recognition of the extraordinary circumstances homeowners faced in the peak years of the housing bubble.
I was impressed that many moderate conservatives considered Right to Rent a reasonable solution to the housing crisis such as former Bush administration economist Andrew Samwick and Desmond Lachmon, an economist at the American Enterprise Institute. Even Fox business anchor Neil Cavuto sent me a note saying that he agreed with the idea.
Of course President Obama would have had to get Congress to pass such a measure. Would they have done it in early 2009? Who knows, there was a huge amount of goodwill directed toward the new president, who had sky high approval ratings at the time. Certainly it would have been difficult to push a Santelli type rant-fest against Right to Rent. After all, there were no taxpayer dollars involved.
In this area, as in so many others, the Obama administration was unwilling to really push anything new. Perhaps Right to Rent wouldn't have worked (we can still try it), but it certainly could not have done worse than the alternatives.
CEPR released a new report on Wednesday in conjunction with Rights Action on the circumstances and aftermath related to the May 11 shooting incident in Ahuas, Honduras, involving the Honduran police and DEA agents. Four local members of communities in this part of the Moskitia region were killed in the episode, and four others shot and injured. As was reported by the Associated Press in May, residents of the nearby village of Paptalaya were subsequently besieged by armed men whom residents described as wearing U.S. Army-style uniforms and speaking to each other in English.
CEPR’s Senior Associate for Policy Analysis Alexander Main traveled to Ahuas and Paptalaya, along with Rights Action’s Annie Bird and Karen Spring, and others, to investigate. During their July trip, they interviewed numerous survivors and eyewitnesses to the traumatic events, as well as U.S. Embassy officials and Honduran authorities. They also examined evidence, and talked to legal experts regarding the current progress, challenges and faults with the Honduran government’s delayed and flawed investigation into the incident. Their findings are the basis of the new 54-page report, “Collateral Damage of a Drug War: The May 11 Killings in Ahuas and the Impact of the U.S. War on Drugs in La Moskitia, Honduras.” It provides what is probably the most detailed account of the events so far. Among its key findings:
• U.S. Embassy officials contradict what State Department officials had previously stated about the DEA's role in the operation. Whereas State had said the DEA played a "supportive role only," both the former head of the DEA for Honduras, Jim Kenney, and U.S. Ambassador to Honduras Lisa Kubiske told the report's authors in separate conversations that Honduran police in these operations respond in practice directly to DEA officials. In addition, many eyewitnesses say it was North Americans, in uniforms with US flags on them, who were in the middle of everything, and that it was North Americans who besieged the Paptalaya village, holding residents at gunpoint and assaulting some of them. This would also contradict the “supportive role only” description.
This week, the Department of Labor (DOL) posted an invitation to states to apply for federal funds to promote work-sharing (officially called "short-time compensation") programs. The total amount available is almost $100 million, with the largest amount -- over $11.5 million -- available for California (see this chart for how much each state could get).
Only states that have work-sharing programs that fit the new federal definition in the Middle Class Relief and Job Creation Act of 2012 (Act) can apply for these grants. The Act places states into 3 categories:
states with existing work-sharing programs that fit the new definition
states with existing programs that don't conform to the new definition
states that don't have work-sharing programs
The federal grants to promote work sharing are divvied up for 2 purposes:
1/3 for implementation or improved administration of their programs -- such as upgrading processing systems
2/3 for promotion and enrollment activities -- such as outreach to and education of employers about work-sharing
Since work sharing is voluntary on the part of employers, publicity and outreach by states is key to improving participation rates. These grants should help jumpstart such efforts. DOL provides a handy application checklist and sample quarterly progress report, along with other useful information about applying. If states fail to take advantage of these grants and the federal reimbursements, they'll be leaving significant funding on the table, at a time of tight state budgets.
The Labor and Employment Relations Association is accepting proposals for "stimulating, creative, and controversial panels, symposia, workshops and papers" related to the theme of the organization's 65th annual meeting, "The Future of Work." Submissions can be from different disciplines – including but not limited to economics, sociology, political science, labor and employment law, industrial relations, and human resource studies – and different stakeholder perspectives, including investors, managers, employees, policymakers and unions.
Complete details about topics and how to submit a session proposal, paper or poster abstract can be found on the LERA website or contact
with questions. The deadline is October 5. The annual meeting will be held in St. Louis, MO, June 6-9, 2013.
The graph shows that over the last three decades the share of college-educated workers with what we define as a good job --one that pays at least $37,000 a year, has employer-provided health insurance, and an employer-sponsored retirement plan-- has actually declined slightly.
The graph also shows that over the same period, the drop in good jobs was even steeper for those with less education. But, what we and most of those commenting have focused on is the decline in good jobs experienced even by the best-educated part of the workforce. This group has “done the right thing” by the labor market, but is, nevertheless, now less likely to have a good job than was the case back in 1979.
Given the interest in “good job” trends by education, we thought we'd produce similar graphs showing the trends in the share of workers that separately meet each of the three underlying criteria. These graphs give a fuller picture of the forces driving the fall in good jobs within the education categories.
Within the economics profession, the standard explanation for rising inequality over the last three decades is a shortage of college-educated workers. Technological progress, the story goes, has increased demand for highly skilled workers of the kind that colleges produce. But, young people have not been going to college in sufficient numbers to meet that demand. The result, these economists argue, is that the wages of the third or so of the workforce with a college degree have pulled ahead of the rest, creating a widening gap between the top and bottom of the income scale.
Data that Janelle Jones and I assembled for a recent CEPR report (pdf), however, raise serious doubts about this commonly held view. We analyzed trends since 1979 in “good jobs,” which we defined as one that pays at least $37,000 per year, has employer-provided health insurance, and an employer-sponsored retirement plan. We found that, overall, the share of jobs that meet these three criteria has declined since 1979 --despite a more than 60 percent increase in GDP per person.
But, what was even more surprising was that the share of workers with a college degree or more who held a good job also fell. As the figure below illustrates, in 1979, 43.2 percent of college-educated workers met our threshold; by 2010, the share had dropped to 40.5 percent. The decline is even more remarkable when you consider that the share of workers with advanced degrees (M.A., J.D., M.D., Ph.D. or similar) almost doubled over the same period, from 6.5 percent in 1979 to 11.8 percent in 2010.
If inequality is being driven by our inability to keep up with rapid technological change, wouldn't we expect to see the share of college-educated workers with good jobs rising, not falling? Of course, workers with less education fared even worse over the same period. But, the story the figure above tells is one where something is consistently pulling the bottom out of the labor market, not one where something is pulling the top away --at least not the top as defined by broad education categories
As Janelle and I emphasize in our report, the real culprit is the systematic decline in the bargaining power of workers --reflected in a drop in the inflation-adjusted value of the minimum wage, a collapse in the unionization rate in the private-sector, the deregulation of previously well-paying industries, the privatization of state and local government jobs, a series of business-biased trade deals, a dysfunctional immigration system, poor enforcement of already weak labor standards, and high unemployment.
College-educated workers do consistently better than those with less education. But, even workers with a college degree have not been able to avoid the shift in the balance of power away from workers and toward their employers.
This week, I've looked at trends over the past 30 years in health insurance and retirement plan coverage for workers. But, what has happened to good jobs overall? In a report released this week, John Schmitt and I calculate the share of the workforce in good jobs by gender from 1979 through 2010 (the most data available).
Overall, the share of workers with a “good job” by our definition fell from 27.4 percent in 1979 to 24.6 percent in 2010, but the graph below shows large differences by gender. For women, the share in good jobs grew almost continuously, from 12.4 percent in 1979 to 21.1 percent in 2010. Over the same period, however, the men's share fell almost 10 percentage points, from 37.4 percent to 27.7 percent.
The main reason for the improvement for women was the dramatic increase in their labor force participation. The labor force participation rate for women increased from 50.9 percent in 1979 to 59.3 percent in 2007, before the Great Recession drove the rate for men and women back down. (Over, the same period, the rate for men decreased from 77.8 percent to 73.2 percent). Despite the increase in women's attachment to the labor market, however, women were still much less likely to be in a good job in 2010 than men were in the same year.
According to our definition, a good job has three components: it pays at least $18.50 per hour, offers health insurance at least partly paid by the employer, and provides some kind of retirement plan. As the overall share of good jobs declines, women and men are meeting in an awful middle.
The official poverty measure has numerous flaws. Most notably, what I'll call the "old PM" has defined deprivation down since the 1960s because it has not kept pace with changes in typical living standards in the United States. As a result, the old PM’s poverty threshold is far too low as a measure of the resources Americans need to meet basic needs in today's economy.
The new Supplemental Poverty Measure, what I'll call the "new PM", is often touted as providing a more accurate picture of poverty in the United States than the old PM. However, the poverty threshold established by the new PM is not adequate as a measure of basic needs. In fact, in many parts of the United States, the new PM threshold will actually be lower than the current threshold. Thus, as Mark Levinson of the SEIU wrote recently in The American Prospect, the new PM “doesn’t fix the fundamental problem" with the old PM.
That said, the new PM seems to improve on the old PM in other important, albeit less fundamental ways. In particular, it counts the Earned Income Tax Credit and various other important benefits that are not currently counted as income by the old PM, and subtracts certain work expenses and health care expenses that people are able to pay out of pocket. This certainly provides a more accurate calculation of total (post-tax, post-benefit) income minus out-of-pocket spending on child care and health care, but does it provide a more accurate picture of poverty?
Not necessarily, particularly when we look at child poverty. To see why not, it is useful to compare how using the new PM changes the picture of poverty among children and the elderly. As the table below shows, using the old PM, the child poverty rate (22.5 percent) is more than twice as high as the elderly poverty rate (9 percent). But using the new PM, the rates look very similar because the child poverty rate drops by more than 4 points, while elderly rate jumps by nearly 7 percentage points. In essence, according to the new PM, child poverty is significantly lower than we thought, while elderly poverty is much higher. And, comparing the two groups, children are only slightly worse off in terms of having their basic needs met than the elderly.
My post yesterday showed trends over the last three decades in employer-provided health insurance coverage, based on analysis John Schmitt and I did for a recent CEPR report. We also examined trends over the same period for employer-sponsored retirement plans.
The graph below shows that share of workers with a retirement plan has zigzagged over the last 30 years, falling from 51.7 percent in 1979 to 45.5 percent in 1987 year, before rising to 50.8 percent in 2001, before falling again to 45.5 percent by 2010.
The pattern differed by gender. Since 1979, the share of men with a retirement plan has fallen about 10 percentage points, while the rate for women has remained essentially unchanged. Since the early 2000s, women have been more likely than men to have any type of retirement plan, a substantial reversal of the pattern in the 1980s and 1990s.
For a CEPR report released yesterday, John Schmitt and I calculated the share of US workers that have employer-provided health insurance (where the employer pays at least part of the premium).
As the graph below shows, since 1979, the share of workers with coverage has fallen 12.9 percentage points, to just 56.8 percent by 2010 (the most recent data available). The decline in coverage has been much more drastic for men than for women. Of course, this is nothing for women to feel good about; in part, this is just because we women had a shorter distance to fall. In 1979, 62.6 percent of female workers had coverage and by 2010 that number had decreased 7.3 percentage points to 55.3 percent. For men, the share with coverage fell 16.4 percentage points, from 74.5 percent in 1979 to 58.1 percent in 2010.
Since the CPS does not measure the quality of employer-sponsored health-insurance coverage, these numbers understate the deterioration in coverage. In 1979, copays, deductibles, annual, and lifetime limits, for example, were rare.
Our new report puts these changes in health-care coverage in the context of a broader deterioration in “good jobs.” We define a good job based on three simple criteria: an earnings threshold at about $18.50 an hour, an employer-sponsored retirement plan, and employer-provided health insurance. Despite enormous increases over the last three decades in the productive capacity of the U.S. economy, we find that the share of good jobs has actually declined. One of the most important contributors to this decline was the fall in employer-provided health insurance.
The following highlights CEPR's latest research, publications, events and much more.
CEPR on the Minimum Wage July 24 marked the third anniversary of the last increase in the federal minimum wage, and CEPR marked the occasion with several op-eds and blog posts. The current federal minimum of $7.25 an hour comes to just over $15,000 a year for a full-time worker. CEPR Co-Director Dean Baker wrote this piece for Truthout and discussed it (MP3) on Sirius XM Radio’s Stand Up! with Pete Dominick. CEPR Director of Domestic Policy Nicole Woo penned this one for the U.S. News and World Report, and CEPR Senior Economist John Schmitt weighed in on the CEPR Blog.
CEPR has long argued that the current minimum wage is too low. Several news articles on the anniversary - including this one in McClatchy - cited CEPR’s work. (For a full listing of CEPR research on the topic click here)
CEPR on the Mexican Elections CEPR Co-director Mark Weisbrot had this op-ed published in the New York Times that highlighted Mexico’s failed economic policies in the wake of the recent elections. Mark also discussed credible allegations of vote-buying and other irregularities that affected the elections in thisGuardian column, and described how problems in the 2006 elections provide important context for understanding these elections and the popular skepticism and protest against them. CEPR’s work related to Mexico’s elections, especially our June paper, “The Mexican Economy and the 2012 Elections", was cited in media coverage such as this article in The Nation and this op-ed in Foreign Policy.
I don’t think the average non-economist appreciates just how much richer and more productive the U.S. economy is today than it was three decades ago. For the typical American, the large increase in economic inequality has masked most, if not all, of the progress.
Janelle Jones and I prepared the table below for a CEPR report (pdf) released today. The table assembles several indicators that all demonstrate a substantial rise in the country’s productive capacity. The share of workers with a four-year college degree or more increased from about one-fifth in 1979 to over one-third in 2010. The median age of the workforce increased by seven years. While we don’t have ideal data, the amount of physical capital (buildings, machinery, equipment, etc.) per worker is about 50 percent higher today than in 1979; and more than 60 percent of workers now use a personal computer on the job, compared to essentially zero in 1979.
The workforce today is more experienced, much better educated, and working with more –and better– capital. Largely as a result, GDP per capita was 63 percent higher in 2010 than it was in 1979.
Mike Spector has a great piece in today’s WSJ explaining how private equity firm Sun Capital retained ownership of Friendly’s after taking the iconic ice cream parlor chain into bankruptcy. Normally, owners lose their investment in a bankruptcy. But Sun arranged for another of its affiliates to provide a loan to keep Friendly’s operating while in bankruptcy, and so became its major creditor as well as its owner. As Spector noted, ‘ That put Sun first in line to be repaid in a bankruptcy, allowing the buyout shop to reacquire Friendly's with a $75 million "credit bid"—essentially using debt owed it as currency to bid for the company.” Sun retained ownership of Friendly’s but with fewer liabilities – including getting rid of its employees’ pensions, which it off loaded onto a government agency when Friendly’s declared bankruptcy. This is not the first time a company owned by Sun has declared bankruptcy, only to emerge from bankruptcy still owned by Sun. Sun has used this tactic in other cases as well. The PE firm managed to pull off the same deal with Anchor Blue, Big 10 Tires, and Fluid Routing.
This post is in response to a recent segment on NPR's Planet Money in which a panel of economists, which included Dean Baker, made recommendations for a dream presidential economic platform. Dean writes that disagreements between the five economists on the panel should be noted, as the resulting fake presidential candidate "will have to do a bit more work to get my vote, even if I did help to design the platform."
I do feel there were some important aspects of these issues that listeners may not fully appreciate that I would like to lay out.
First, while I fully endorse the view expressed in the segment that a tax deduction for employer provided health care makes no sense abstractly, there is a historical basis for this deduction that makes it difficult to change. Workers, and especially unionized workers, have often explicitly given up higher wages for better health care benefits. If they were to lose employer provided health care benefits, there is no guarantee that their wages would rise by a corresponding amount. While all good economists believe that there is trade-off between wages and benefits, that does not mean that the trade-off is always one-to-one and immediate.
I would be worried that if we were to eliminate the health care deduction in an environment like the current one, in which high unemployment has badly weakened workers' bargaining power, it would result in a net reduction in workers' compensation. In my view, that can't be a good thing at a time where we have already seen such a large upward redistribution of income.