Economists everywhere are scrambling to determine how today’s weak jobs report impacts the strong recovery story of 3.2 million jobs created over the last year. But when it comes to millennials in the labor force, the monthly numbers are only a small part of the story. That’s why I’ve done some number crunching to see what’s really going on with my generation.

My research highlights two factors that are holding millennials back: too many are not completing college, and too many that do have skills that are not well-matched to labor market demands.

When it comes to young workers, aged 25-34, the gap in labor force participation for those with and without a degree is now roughly 10 percentage points – and the gap is widening.* The chart below illustrates this stark reality – having a college degree could make the difference in whether or not millennials find a job.

However, my research also shows that in today’s labor market, having a college degree may not be enough. That’s because, in addition to completing college, the economic prospects of millennials depends on having high-wage skills employers demand.

Since the recovery began in 2009, college graduates’ outcomes have diverged. Some have seen great success in the economic recovery, while others have floundered at the expense of their less educated peers. I call this phenomenon the “Great Squeeze,” and I have previously written on it here. That real average annual earnings for young college graduates fell by 12 percent over the last decade reinforces this divergence between workforce success and underemployment.

It turns out that what you study matters, as not all graduates are struggling. Graduates in high-skill, high-demand fields such as computers and mathematical occupations, for example, are doing just fine. The most recent Conference Board data shows the ratio of unemployed workers to advertised jobs for computer and mathematical occupations is just 0.17.

The skills mismatch helps in part to explain why too many college graduates find themselves underemployed well after graduation. Our higher education system has not adjusted to the changing shape of the labor market, one where job creation is focused at the high and low end of the skills spectrum.

That’s why it is not obvious that while some postsecondary credential is necessary, a college degree for everyone is the right fix. Instead, these charts suggest we need to look outside status quo higher education, to encourage more pathways into the workforce that provide young people with the skills employers demand.

*Note: Few in the aged 25-34 cohort are enrolled in school, and both men and women with a high school diploma or some college, no degree had significantly lower labor force participation rates than college graduates.

For young Americans, Washington, D.C. may have more to offer than government jobs and free museums.  It may also provide more opportunities to get a raise than any other top 10 U.S. city.

According to my analysis of new Census data on young Americans, Washington, D.C. was the only top 10 U.S. city, by population, where young workers saw an increase in real median earnings since 2000 (in addition to having the highest real median earnings overall).  Young workers in every other top 10 U.S. city experienced sizable declines.  More than half – six – of the top 10 cities saw declines greater than 10 percent, with young workers in Miami and Atlanta enduring real declines of more than 15 percent.

The table below shows the 2000-2013 change in real median earnings for young Americans age 18-34 working full-time by major U.S. city, where 2013 is measured as a five-year average over 2009-2013. Data for 2000 comes from the 2000 Census long form, and data for the 2009-2013 five year average comes from the American Community Survey 2009-2013 five-year estimate.

Median Earnings for Young Americans Aged 18-34 in the Ten Most Populated Metro Areas

That real median earnings increased in Washington, D.C. while falling elsewhere might help explain why the nation’s capital has become an increasingly popular place to be for young people. The number of 18-34 year-olds living in the D.C. metro area has increased by 19 percent, or 226,000, since 2000, compared to one percent increases in Chicago and New York, and 9 percent nationally. Washington, D.C. also has a high share of employment dependent on the federal government, and a highly educated youth population, both of which may have been less affected by the economic downturn. (Houston, however, was the top city for youth inflows in spite of falling real median earnings, which saw its 18-34 year-old population increase 25 percent since 2000.)

Still, falling real median earnings across the board outside Washington, D.C. suggests the underlying issues affecting young workers is not solely about educational attainment or geography. Other major cities with a higher than average share of young college graduates, such as New York and Chicago, also experienced a decline in real median earnings. This is consistent with my previous research, which shows falling real average earnings for young college graduates at a time when many are questioning the value of a college degree.

Overall, the sharp decline in real median earnings for young workers is troubling. It suggests young Americans continue to face strong financial headwinds during their professionally formative years. Moreover, it could hinder young people’s ability to fully participate in the greater economy long-term. That has significant implications for politicians on both sides of the aisle, especially Democrats who care about creating a more convincing pro-growth agenda.

The latest jobs numbers, along with new research from the Federal Reserve and Brookings, reaffirms what I’ve been writing for some time: the Great Squeeze in labor force participation is hitting the young and least educated the hardest. Further, the conclusion that this drop is a structural problem bolsters my argument that both a slow-growth economy and a workforce skill mismatch are to blame, instead of simply higher rates of school enrollment. This has big implications for what policies will – and won’t – fix the problem.

The new joint Brookings-Federal Reserve study takes a deep dive into the troubling fall in the labor force participation rate for young people aged 16-24 since the mid-1990s. The study concludes that:

“some crowding out of job opportunities for young workers [is] associated with the decline in middle-skill jobs and thus greater competition for the low-skilled jobs traditionally held by teenagers and young adults”

I’ve been writing about this for two years – calling this phenomenon the “Great Squeeze.” The premise of the Great Squeeze is simple: the slow-growth economy, coupled with a skills mismatch, is forcing more college graduates and experienced professionals to take lower-skill jobs for less pay. This is hitting those with less education and experience the hardest – young people, who are being forced down and out of the labor force.

That’s why we still see historically high numbers of young people neither enrolled in school nor in the labor force, particularly during the summer. In fact, the latest numbers for July show that more than 8.1 million people aged 16-24, 4.9 million of whom were teenagers, were neither enrolled in school nor in the labor force. This is 1.8 million more young people than in July 2000, and still 1.3 million more than in July 2007.

Importantly, the new Brookings-Fed paper makes it clear that most of this problem is structural – that is, it is a long-term problem as opposed to a temporary effect of the Great Recession. This can certainly be seen in the latest data, where the labor force participation for teenagers not enrolled in school during July has dropped from 67 percent in 2000 to 50 percent in 2014.

The structural nature of the Great Squeeze has significant implications for policy. First, it suggests that some of the problem stems from employers not creating enough middle-skill jobs. In other words, the slow-growth economy of the last decade has left a large amount of young college graduates underemployed. That calls for a pro-growth, pro-investment agenda, which we will outline in a forthcoming PPI paper.

Second, it suggests there is a workforce skill mismatch, particularly for young underemployed college graduates. This will not be solved by maintaining the current postsecondary education system, or funneling everyone into four-year college degrees. New research also out from the Fed demonstrates that a Bachelor’s degree is not the right investment for everyone, with a quarter of college graduates earning the same salary as those with a GED. Instead, we need more public-private partnerships in higher education, and viable, employer-driven alternative pathways into the workforce.