Mobility

In a new working paper, economists Wojciech Kopczuk of Columbia, Emmanuel Saez of UC Berkeley and Jae Song of the Social Security Administration use Social Security earnings administration data to track income and mobility trends in the United States since 1937. Among their conclusions: overall mobility is relatively stable, but there are major differences by gender, with inequality and mobility worsening among men since the 1950s.

Here's the time-series they construct for earnings shares:

And the conclusion section of the paper:

Our paper has used U.S. Social Security earnings administrative data to construct series of inequality and mobility in the United States since 1937. The analysis of these data has allowed us to start exploring the evolution of mobility and inequality over a full career as well as complement the more standard analysis of annual inequality and short term mobility in several ways. We found that changes in mobility have not substantially affected the evolution of inequality, so that annual snapshots of the distribution provide a good approximation of the evolution of the longer term measures of inequality.

However, our key finding is that while the overall measures of mobility are fairly stable, they hide heterogeneity by gender groups. Inequality and opportunity among male workers has worsened along almost any dimension since the 1950s: our series display sharp increases in annual earnings inequality, slight reductions in short-term mobility, large increases in long- term career wide inequality with slight reduction or stability of long-term mobility. Against those developments stand the very large earning gains achieved by women since the 1950s, due to increases in labor force attachment as well as increases in earnings conditional on working. Those gains have been so great that they more than compensate for the increase in inequality for males when focusing on the bottom of the distribution.

Thus, the weakening of social norms and labor market institutions inherited from the post- war years which favored low skilled white male workers59 at the expense of women, Blacks, and top talent has had two important and conflicting consequences for earnings inequality in recent decades. It has allowed women to close a large part of the gender gap, hence improving the position of low earners (especially from a lifetime and upward mobility perspective). However, it may have also strengthened pure market forces which have contributed to increasing sharply the pay of top earners in the US economy.

We would like to develop the present analysis in several ways in future work. First, we plan on investigating in more detail the mechanisms of the surge in top earnings using employee- employer data. This will allow us to examine the industrial composition of top earnings and its evolution. We will also be able to analyze the evolution of the labor market for top earners (such as tenure, turn-over, and earnings changes within jobs and across jobs). Second, our analysis has remained largely descriptive without trying to test formally specific hypotheses or decompose patterns into various factors. Even though the administrative SSA data have relatively few covariates, it should be possible to expand the analysis in that direction.

More importantly, our analysis has lead us to assemble various SSA data that had not been systematically used for research purposes previously. We hope that our broad analysis of inequality and mobility, as well as the new opportunities for joint research pro jects between SSA and outside academic researchers, will encourage further research with these extraordinary SSA earnings data which can cast new light on many different aspects of labor economics in the United States.

As I noted in my previous post, the following claim frames the data in a new Pew Economic Mobilty Project report on immigration:

... the American engine of economic assimilation ... is incorporating a fundamentally different and larger pool of immigrants than it did in earlier generations.

My previous post discussed the problem with the "larger pool" claim, this one takes a look at the claim that today's immigrants are "fundamentally different" than earlier generations of immigrants.

One of the definitions of "fundamental" is "involving all aspects", so one way to interpret the claim that today's immigrants are fundamentally different is that they are different in all aspects from earlier generations. Since today's immigrants are, like their predecessors, members of the human race, this can't be what the Pew report means by fundamentally different.

Perhaps, what is meant by fundamentally different is that today's immigrants are different in some other very important ways. If this is the case, it would have been better for the report to be framed in a way that is more specific—instead of "today's immigrants are fundamentally different"—which is really a partisan talking point in the current contentious immigration debate—it could have said, "today's immigrants are different in these important ways ...."

But even if the report had used more careful and less contentious framing language, the question remains—how different are today's immigrants, and are they different in ways that are relevant to Pew's mission of developing consensus on and depoliticizing the issue of economic mobility?

One way in which today's immigrants are different involves race, ethnicity and national origin. Today's immigrants are much more likely to be Latino or Hispanic and from Mexico or other places in the southern half of the American hemisphere. In my view, this shouldn't really count as a fundamental difference given our national committment to judging people by "the content of their characters, rather than the color of their skin."

Some commentators have argued that the content of the characters of Latino immigrants is fundamentally different from the content of the characters of earlier generations of immigrants. Samuel Huntington, for example, has argued that the settlement of the United States by "Anglo-Protestants" played a crucial role in defining what he sees as a distinct "traditional American identity" and set of values that defines our character to this day. According to this line of reasoning, Latino immigration threatens that identity and our national character.

Whatever one thinks of this argument—I think it's silly—it's certainly not a consensus view, so I hope that's not what the Pew Economics Mobility Project means when it says today's immigrants are fundamentally different.

Even if one buys into the idea that the religious and family values of immigrant groups are relevant to our "national identity" in important ways, there's no reason to think that Latino immigrants are different from earlier groups in ways that are detrimental to that identity. In fact, some conservatives make quite contrary arguments. For example, in a review of Huntington's book Who Are We? The Challenges to America's National Identity, Francis Fukayama concludes that "Hispanic immigrants will help to reinforce certain cultural values like the emphasis on family and work, and the Christian character of American society."

There are a number of grounds for thinking that the United States will assimilate Hispanic immigrants just as it has earlier ethnic groups. Most important is the fact that they are Christian—either Catholic or, to an increasing degree, Evangelical Protestant. When controlling for socioeconomic status, they have stronger traditional family values than their native-born counterparts. This means that culturally, today's Mexican immigrants are much less distant from mainstream "Anglos" than were, say, the southern Italian immigrants or Eastern European Jews from mainstream WASPs at the beginning of the 20th century. Their rates of second- and third-generation intermarriage are much closer to those of other European groups than for African-Americans. And, from Gen. Ricardo Sanchez on down, they are serving honorably today in the U.S. armed forces in numbers disproportionate to their place in the overall population.

If culture and values aren't a relevant source of "fundamental difference", what about education and intergenerational mobility? The fairest comparison of current immigrants with earlier generations would conclude that there are some differences and lots of similarities, depending to some extent on which generations one is comparing. The Pew report correctly notes that the educational levels of current immigrants aren't particularly different from those of earlier generations who entered after the 1965 immigration overhaul. In fact, the current generation is somewhat more likely to have a college or post-graduate degree than earlier generations and more likely than U.S.-born citizens to hold an advanced degree. Given this, it's hard to argue that education is an area of fundamental difference.

Moreover, it's hard to conclude that today's immigrants are falling farther behind educationally than those of some earlier generations. According to Richard Rothstein:

In the first decades of the 20th century, there was a yawning achievement gap between immigrants and the American-born. Even immigrant Jews, whom many now consider yesteryear's "model minority," did worse in school than immigrants do today. Schools took years to lift the children of immigrants up to achievement levels we now take for granted.

According to a 1911 federal report, 80 percent of native white seventh graders of that period in cities like Boston, Chicago and New York made it into eighth grade the next year. But a smaller proportion of Russian Jewish immigrant children, and many fewer Southern Italian immigrants, did so. After eighth grade, Southern Italians made it to high school at only half the rate of their native peers. (For purposes of the report, children were considered immigrants if they or their fathers were foreign-born.)

Twenty years later, things weren't much better. By the end of the 1920's, Italian immigrant high school students were still graduating at only one-quarter the rate of nonimmigrants. This was a much bigger native-immigrant gap than today's.

Immigrant Jews performed less well than native whites, but better than immigrant Italians. ....

....

Standardized exams seemed to confirm immigrants' inferiority. About this time, I.Q. tests became common. In 1919, the median I.Q. score of Italian 10-year-olds in New York City was 84; for native-born whites, it was 109. But when only those students whose fathers were unskilled or semiskilled laborers were compared, Italian and native I.Q.'s were nearly identical.

Progress was slow, and assimilation rates for subsequent generations varied by ethnicity. It seems that Jews typically adapted to schools and became fully fluent in English more quickly than other immigrant groups. Italians seemed to do so more quickly than Greeks.

But of immigrant children themselves, even Jews were mostly unsuccessful. The 1911 report found that about half of all Jewish immigrant children around the country were held back in school. (Jews from Poland did particularly poorly; two-thirds were held back.) ....

Is this relevant to today's problems? Yes, because we need reminding that the past provides models of very limited value. Americans now do a better (and more nearly equal) job of educating immigrant children than did Americans of earlier eras. ....

Similarly, in the area of intergenerational mobility the differences between current immigrants and earlier generations are far from fundamental. In a paper published in Economic Journal in 2005, economist David Card notes that, controlling for age, second generation immigrants have "more education than people whose parents were born in the U.S." and that "despite the lower education of their parents, children born to immigrant parents seem to catch up and even surpass the levels of children born to U.S. natives."

Card concludes:

On the question of immigrant assimilation, a major constraint in the U.S. literature has been the absence of true longitudinal data. Nevertheless, I believe that a narrow focus on immigrant earnings is misplaced. Few of the 40 percent of immigrants who come to the U.S. without completed high school education will ever catch up with the average earnings of natives. Most of their U.S.-born children, however, will catch up with the children of natives. Evidence on the intergenerational progress of immigrants’ children is now becoming available, and points to above-average levels of educational attainment, even for children whose fathers had much lower schooling than native-born fathers. The relatively strong educational progress of second generation immigrants, together with the limited evidence of adverse effects on less skilled natives, suggest that the new immigration may not be so bad after all.

Oddly, the Card paper on intergenerational mobility isn't discussed or even cited in the Pew Report, even though Card is one of the nation's leading economists—in 1995, he won the John Bates Clark Prize, considered to be second only to the Nobel economics prize—and the paper speaks directly to the issue of immigrant mobility. Any report that holds itself out as an objective synthesis of research in this area should discuss prominent research on the topic by an economist considered to be a leader in the field.

(As long as I'm on the topic of notable omissions, another paper that should have been cited in the Pew Report is The Improving Educational Profile of Latino Immigrants by B. Lindsey Lowell and Roberto Suro of the Pew Hispanic Center. I guess the Pew Projects don't always read each others stuff.)

Are there some differences between today's immigrants and earlier generations? Yes, but there also are many more similarities, and the relevance of differences that do exist is limited. Finally, it goes too far to claim, particularly in a publication produced by a project that holds itself out as building consensus, that today's immigrants are "fundamentally different" from earlier generations.

A new report from the Pew Economic Mobility Project makes this claim:

... the American engine of economic assimilation ... is incorporating a fundamentally different and larger pool of immigrants than it did in earlier generations.

The claim that the United States is "incorporating a ... larger pool of immigrants that it did in earlier generations" is only true if one defines "earlier generations" restrictively and by "larger pool" means the number of immigrants rather than the percentage of immigrants in the U.S. population.

On the definition of "earlier generations", the key question is whether one limits the comparison to immigrants who came to the United States roughly between the National Origins Act of 1924—an Act designed to limit immigration to "desirable" Western European immigrants by imposing national origin quotas on immigration—and the Immigration Act of 1965, which reformed the quota system to make it less discriminatory, or do those earlier generations include the period before 1924? The Pew paper includes no data from periods prior to 1940, so it's adopting a restrictive definition of earlier generations.

This restrictive definition matters because the percentage of immigrants in the U.S. population was larger in most of the pre-1924 period than it is today. As the U.S. population has increased, the number of immigrants has also increased, except for the period between the 1924 and 1965 Acts when it declined.

At a minimum, a document produced by a project that purports to "broaden the current, sometimes politicized, debate over income inequality to achieve greater consensus about the state of economic mobility in America" should avoid presenting data in ways that narrow, politicize, and, uh, disconsensify the debate by failing to do something as simple as providing the whole picture for the 20th century as well as the relevant historical context. But by excluding the historical data from the period before the 1924 immigration law went into effect (and not explaining that the 1924 law is now widely viewed as racially motivated)¹ and not putting immigration trend numbers into the context of the overall U.S. population, the report narrows debate and avoids consensus.

As additional context it also would have been helpful to compare the U.S. immigration rate to those of other well-off nations. Conventional wisdom has it that the United States is an outlier compared to other nations, but increased migration is a global trend. Immigrants comprise more than 15 percent of the population in more than 50 countries (by comparison they're around 11 percent in the US according to the 2000 Census) and account for large proportions of population and employment growth in most developed countries.

If the Pew Mobility Project had wanted to produce a balanced report on immigration, it would have been smart to 1) not have a partisan in the immigration debate be the sole author of the report (the author is a nice partisan for a conservative, but a partisan nonetheless, for an example of a more balanced report that involved partisans from differing camps, rather than just a single representative of one, see this one); 2) have a balanced group of immigration researchers involved as reviewers/commenters on the publication. Only one of the researchers listed in the report's acknowledgments, the controversial George Borjas, a favorite academic of the anti-immigration camp, actually specializes in the field of immigration. Among the obvious choices to balance the influence of Borjas, those prominent researchers who either specialize in immigration or have produced key immigration-related research include: Alejandro Portes, David Card, Donald Hernandez, Audrey Singer (the immigration specialist at Brookings, who oddly doesn't appear to have been consulted at all, even though the author of the Pew Report is also at Brookings), Jeff Passel, Michael Fix, Richard Alba, etc.

The problem here isn't that the Pew immigration report falls outside of the standards of the kind of research advocacy literature that organizations taking sides in policy debates produce. It's of the same quality as research advocacy literature produced by the Heritage Foundation, CBPP, or Brookings. Rather, it's that the Pew Mobility Project says that it's holding itself to a higher standard, and that's a standard that this publication fails to meet. So, either they should change the publication, or change the standard they say they're holding themselves to. Personally, I'd be fine with the latter.

I'll take on the second half of the claim, that today's immigrants' are "fundamentally different"—a claim commonly made in regards to Jewish, Italian, and other southern and eastern European immigrants in earlier periods of our nation's history—later in Part 2 of my reponse.

Duke's Jacob Vigdor summarizes his recent research on the impact of relocation for families after Hurricaine Katrina:

....

The survey data obtained through the CPS paints a portrait much bleaker than some press coverage. Evacuees, on the whole, suffered a serious negative shock to their labour supply and earnings. These effects were concentrated in the subset of evacuees who failed to promptly return to the greater New Orleans area after the storm. Moreover, for this group of long-term evacuees, there are little to no signs of improvement over the twelve-month period from November 2005 to October 2006. Some evacuees may be doing quite well, but most are doing even worse than they had been in the New Orleans area prior to the hurricane. The fortunes of long-term evacuees appear to be almost completely unrelated to the characteristics of the metropolitan areas where they relocated.

Few may have genuinely believed that individuals, particularly poor residents of depressed urban areas, could land in a new city, dust themselves off, and productively join the workforce. But the evidence of little to no progress, even fifteen months after initial evacuation, starkly reveals the importance of local ties and information networks in the employment outcomes of people like the Katrina evacuees.

....

Some good news from CEPR on the future of the Survey of Income and Program Participation:

One week after 446 social scientists signed a letter urging Congress to continue funding the U.S. Census Bureau’s Survey of Income and Program Participation (SIPP), the Bush Administration has reversed its decision to eliminate the program.

Bush's FY08 budget called for the elimination of the SIPP. Now that it has decided to continue the program, it is unclear how it will be funded. Unless the Bush administration modifies its budget request, Congress may not be able to allocate funds for the SIPP because President Bush has vowed to veto any appropriations above his budget requests.

A letter sent yesterday from Representative Wm. Lacy Clay, Chairman of the House Oversight and Government Reform Subcommittee on Information Policy, Census, and National Archives, to Census Bureau Director Charles Louis Kincannon, requests a cost estimate for the survey. According to the letter, “it appears that an increase in the Commerce- Justice- Science Appropriations bill will be necessary to fund any new SIPP panel in FY 2008.”

So, SIPPY isn't out of the woods yet when it comes to adequate funding, but the Administration's reversal is a good sign.

Over 400 researchers signed a letter urging Congress to maintain funding for the important Survey of Income and Program Participation, otherwise known to researchers as SIPP, and to kids all over America as SIPPY, the lovable longitudinal data set and easy-to-grasp cup for toddlers. Well, he's not known yet to kids all over America, but it's just a matter of time ....

I can't help myself—I'm compelled to say a bit more about this:

The flexibility of the American labor force seems to be one reason that recessions have become less frequent and unemployment is less of a problem here than in Europe, notes Jason Furman, a leading Democratic economist. In this country, fast-growing companies can hire new workers without worrying that they are making a 30-year commitment.

Furman is a smart guy and has done good progressive work on issues like Social Security, so I expect more from him than the kind of generalized dissing of Europe that one regularly hears from conservative economists, dissing that counts on conventional wisdom and Europhobia rather than facts for its resonance. There's good and bad in European economic policies and outcomes, just as there is good and bad in American economic policies and outcomes. But the advantages of US flexibility are less black and white than Furman paints them. Two facts are particularly relevant to this debate:

First, employment rates between the US and Europe aren't all that different. As John Schmitt and Dean Baker have recently shown, the employment rate among prime-age workers in the EU is 78.2, compared to 79.3 in the US, not exactly a big difference. Among prime-age men, there's basically no difference. Nordic nations like Sweden and Denmark have higher employment rates than the US.

Moreover, Baker and Schmitt note that these comparisons overstate the difference because they don't take incarceration rates into account:

Since incarceration rates in the United States are five to ten times higher than they are in European countries ... the exclusion of the non-institutional population from the comparison of employment rates has a much larger effect on the U.S. employment rates than it does on those in Europe. As we have demonstrated elsewhere, the effect of excluding the non-institutional population from estimates of the employment-to-population rate raises the overall U.S. employment rate for the population 16 and over by 0.9 percentage points; the effect for men, who make up about 90 percent of the U.S. prison population, is 1.3 percentage points ....

Second, and particularly damning in my view, income mobility is lower in the United States than in nearly all other wealthy European nations, as the chart below shows. (Based on OECD data, the chart comes from an excellent piece by John Schmitt and Ben Zipperer that examines US-Europe differences on a range of indicators.)

Furman's overgeneralization is also problematic for strategic reasons. Nobody I know of is arguing for the adoption of the French Labour Code in toto—although it's probably looking pretty good right now to the former Circuit City employees who were fired because they had the teremity to stick around until they were paid" too much" (in the $12.00 an hour sense, not in the $2.5 million sense, which is how much Circuit City's CEO makes.) But a lot of us think that policies to improve job quality and strike a better balance between work and family life—policies like paid family leave and paid sick leave that are a core part of almost all of the European economic and social models—would be a good idea, even though there is no question that they reduce employer flexibility. Unless Furman opposes paid family leave, paid sick leave, minimum wages, and other forms of labor market regulation, it would be better for him to forgo generalizations that support the case of people who do.

Along with all of the other changes on the site, you’ll notice the new button and tab for The Mobility Agenda.

For the past year, Shawn, Rachel, and I—with research assistance from Sarah Sattelmeyer and Amelia Dietrich—have designed and developed this new initiative with the goal of strengthening the economy with better jobs. The Center for Community Change provided crucial support for the Agenda by hosting us through the development and implementation year.

Nearly one-third of all jobs pay around $10 an hour or less, often providing no employment benefits and little flexibility. Even though the U.S. is among the wealthiest nations in the world, workers in these jobs are paid less than workers who hold similar jobs elsewhere.

The last decade has seen some progress on advancing a number of well-known policies to improve job quality by boosting the minimum wage and expanding publicly subsidized employment benefits, like child care and wage subsidies such as the Earned Income Tax Credit. Similar progress has been made to address education and advancement strategies to prepare workers for skilled jobs.

However, when one worker advances out of a low-wage job and another worker takes it, the job does not change. Across the nation, state and local stakeholders are experimenting with a host of new initiatives to improve low-wage jobs. These innovative ideas are far less well known and aren’t commonly incorporated into an anti-poverty agenda.

We will identify and promote new ideas and strategies to improve jobs in this large and growing segment of our labor market.

For more information about The Mobility Agenda, including findings from our national scan for ideas and our recent roundtable in Illinois, click on The Mobility Agenda button or tab. And let us know what you think—the search for new ideas goes on!

Two days after the election, the New York Times published the news that “Victorious Democrats Vow Cooperative Approach on Taxes and the Economy”. Well. Bollocks.

My friends and family keep asking me if I am excited, celebrating, partying up a storm, and so on. I am not.

Everyone assumes that inclusionist economic policies I’ve advocated over the years stand a chance of implementation in the next Congress. Well, maybe – some of those ideas. But is it a new day for equitable economic policy? I think not…not just yet.

Of course it’s good news that so many of the President’s worst ideas are now buried deep and going nowhere. Plans to further reduce taxes on the wealthy (by eliminating the estate tax) and kill the universal retirement system (by privatizing Social Security) are not moving in the 110th Congress.

But, I know we aren’t likely to see true progress on economic fairness and inclusion just because both houses have new leadership. Read more...

Published in 1999, No Shame in My Game: The Working Poor in the Inner City by Harvard anthropoligist Katherine Newman is one of the best books written in the last decade on poverty and low-wage work. Here's Newman's explanation (from the book's preface) of why she decided to write the book:

It seemed to me that social scientists eager to put the urban poverty problem back on the national agenda had focused almost all their attention on people who were outside of the labor market .... This focus on the jobless poor has created a gap in our understanding of inner-city economies. The preoccupation with ghetto dwellers who don't work has convinced policy-makers and most Americans that no one else is left in places like Harlem. Sixty-nine percent of families living in central Harlem have at least one worker. Yet in surveying most books on poverty over the last decade or so, one would have to dig pretty deep before finding very much information about the working poor, people who toil year-round and either fail to pull above the poverty line or struggle to make ends meet just above it. Having convinced ourselves that welfare dependency and joblessness represent the true face of American poverty, the policy agenda of the 1990s as pressed by conservative politicians became very easy to sell.

Michael Negron has an excellent post over at TPM Cafe that hits on many of the "beyond the poverty debate" points that we've made before on this site (see our resource page on "reframing the poverty debate").

Long-term poverty, writes Jacob [Hacker in his new book The Great Risk Shift], is rarer than we think. He cites a finding in a recent book on poverty by Mark Rank: less than one tenth of Americans experience five straight years of poverty during adulthood. In discussing child poverty, Jacob writes that the concluding that one in five children live in poverty doesn’t fully capture the economic insecurity behind this figure. The fact is that more than half of American children spend at least a year in poverty by the time they reach eighteen. Over half of Americans will spend at least a year in poverty between the ages of twenty and seventy-five. This figure does not hinge on the inclusion of cash-poor college students; if people under the age of 25 are excluded from the calculation the likelihood of experiencing poverty by seventy-five is still nearly fifty percent. For comparative purposes, people in their forties today have more than a 36% chance of ending up in poverty, nearly three times greater than the average forty year-old in the 1970s.