Is Education the Cure for Poverty?

Economists may disagree a lot on policy, but we all agree on the "education premium" -- the earnings boost associated with more education. But what role can education play in a realistic antipoverty policy agenda? And what are the limits of that role?

First, it depends on whether you're talking about children or adults, and schooling versus job training. And second, the extent to which education is rewarded depends on what else is going on in the economy.

As Greg J. Duncan's companion piece (page A20) suggests, investment in early childhood has immense benefits. And at the other end of the schooling spectrum, college graduates' wage advantage over those with only a high-school diploma went up dramatically in the 1980s and early '90s. But the premium that high-school graduates enjoy over dropouts has been flat for decades. In 1973, high-school grads earned about 15.7 percent more per hour than dropouts, 15.9 percent in 1989, 16.1 percent in 2000, and 15.5 percent last year. And for adult workers, the historical record for job-training programs is pretty dismal, though more recent initiatives -- with their focus on more carefully targeting training for local labor markets -- show much more promise.

Nobody doubts that a better-educated workforce is more likely to enjoy higher earnings. But education by itself is a necessary insufficient antipoverty tool. Yes, poor people absolutely need more education and skill training, but they also need an economic context wherein they can realize the economic returns from their improved human capital. Over the past few decades, the set of institutions and norms that historically maintained the link between skills and incomes have been diminished, particularly for non-college-educated workers. Restoring their strength and status is essential if we want the poor to reap the benefits they deserve from educational advancement.

What Research Shows

Julie Strawn of the Center for Law and Social Policy, reviewing an extensive sample of basic education and training programs, concluded that education alone is much less successful in raising employment and earnings prospects than education combined with a strategy of focused job training (with an eye on local demand), "soft skills," and holding out for quality jobs.

One study found that a year of schooling raised the earnings of welfare recipients by 7 percent, the conventional labor economics finding. But given that many of these workers entered the job market in the $6- to $8-an-hour range back in the 1990s, you're talking about moving families closer to the poverty line, not pushing them significantly above it.

Strawn reports that when education is combined with multidimensional job training, readiness, and a quality job search, the returns more than double. One Portland, Oregon, program resulted in a 25 percent increase in earnings, a 21 percent increase in employment, and a 22 percent reduction of time spent on welfare (all compared with a control group that didn't get the services).

This finding makes intuitive sense: Programs that combine general education with training specific to both the individual and his or her local labor market work better than ones that fail to combine these activities. (They're also more expensive, but you get what you pay for.) Yet to get to the nub of the strengths and limits of education and poverty reduction, we need to go back to first principles and think about how they interact with the realities of the political economy.

Education is only a partial cure for poverty because of all the other recent changes in the labor market. At least half of the inequality increase has taken place within groups of comparably educated people, and since 2000 that proportion has been increasing. Income-inequality data show that the concentration of income in 2005 is the highest it has been since 1929. Yet research that Lawrence Mishel and I conducted shows that since the late 1990s, the college wage premium has been flat. In real terms, college wages were up less than 2 percent from 2000 to 2006. Even among the highly educated, only some are getting ahead, and lots aren't.

In short, we are not living in a meritocracy, where we can reliably count on people being fairly rewarded for their improved skills. So we need additional mechanisms in place to nudge the invisible hand toward outcomes that are more meritocratic and just.


Education is a supply-side policy; it improves the quality of workers, not the quality or the quantity of jobs. A danger of overreliance on education in the poverty debate is that skilled workers end up all dressed up with nowhere nice to go.

Some economists contend that faster rates of technological advance require ever more highly skilled workers, and that demand shifts lead to low wages for the low skilled. But our work at the Economic Policy Institute suggests that while technological changes have always been an important factor in the labor market, the rate of change now is no greater than in the recent past. Technological change is one of the reasons we've doubled the share of college grads but continued to see their unemployment rates in the 2 percent range -- we produce and absorb a lot of college grads.

Our economy, however, is still very much structured to produce lots of low-wage jobs. In fact, according to the occupational projections by the Bureau of Labor Statistics, the low-wage sector of our economy will be the source of much job growth over the next decade. The American economy will continue to employ significant numbers of retail salespersons, waiters and waitresses, food-prep workers, home health aides, maids and housekeepers, etc. Of the 30 occupations adding the most jobs to our economy, those requiring the least training make up half of the total.

The question, thus, is not whether jobs for those with only high-school degrees or even some college will exist or be plentiful in our future (they almost certainly will be); the question is whether the quality of these jobs will help reduce or reinforce working poverty.

In our most recent version of "The State of Working America," we borrow a technique from economists Sheldon Danziger and Peter Gottschalk for analyzing the roles played by multiple determinants of poverty. Their method parses out the roles of race, family structure, economic growth, and inequality, and we add the role of education.

As the chart on the right shows, family poverty rates did not fall much between 1969 and 2000, because major factors were offsetting one another. Improved education lowered family poverty by almost 4 percentage points, a considerable effect. But economic growth and inequality had considerably larger effects. Growth in the overall economy lowered poverty rates by 5.7 points, while inequality raised it by 5.1 points. Family structure added 3 points to family poverty rates over these years, and race added 1 point.

Decompositions of this type are far from definitive; they tend to hold one factor constant and see how things change, then do the same for another factor, etc. But in this case, the results are demonstrative of the main point regarding education in the poverty debate: It's an important part of the story, but it's not the whole story, or even the most important part.


Demand -- the extent of overall growth, how taut the labor market is -- matters, as does the extent and nature of inequality, as does the quality of jobs. In the late 1990s, poverty fell to historic lows for those with the lowest education levels, including African Americans and single mothers. Did skills rain from the heavens? Did employers suddenly shed their advanced-skill requirements? Of course not. It was good old-fashioned full employment forcing employers to bid wages up to get -- and keep -- the workers they needed. And yes, this interacted with welfare reform and a significant expansion of work supports, like the Earned Income Tax Credit, subsidized health and child care, and the minimum-wage increase.

In fact, one could be forgiven for thinking that, except for some of the punitive aspects of welfare reform, we briefly got poverty reduction right during the late 1990s. The one-two punch of full employment and expanded work supports worked to meet the expanding labor supply with even faster growing labor demand, and the subsidies helped to close part of the gap between what people earned and what they needed.

But notice how all of this is unwinding in the 2000s. Unemployment is low, but other indicators -- such as labor-force participation and real wage trends -- suggest we're not yet at full employment; there's been no expansion of work supports, and even some retrenchment of supports such as the State Children's Health Insurance Program and child care, policies clearly associated with helping the working poor get ahead. The outcome has been predictable and depressing, especially in contrast to the progress we made in the 1990s.

And if education is one key antipoverty strategy, then programs demanding that beneficiaries "work first" often sacrifice the promise of increased returns to education and training on the altar of take-any-job. This approach is not only stingy; it's also shortsighted, as it threatens to diminish the likelihood that those who want to "play by the rules" will realize their economic potential.

Helping the poor receive more education is part of the answer. Whatever their skill level, workers need a context wherein they can be rewarded for their skills, where the benefits of the growth they help to create flow freely their way. This means having a set of protections, institutions, regulations, and social norms in place to keep the greedy fingers of inequality from picking the pockets of the working poor.

You need to be logged in to comment.
(If there's one thing we know about comment trolls, it's that they're lazy)