• August 5, 2019
• Data Visualization

 

America’s cities used to be places where anyone could move to, boost their wages, and live a higher quality of life. No matter a person’s career path, people could find a good, secure job and make more money even after accounting for costs of living. In fact, back in the 1940s, there were far more people moving to cities with lower levels of education to find higher wage work than there were highly-educated workers. This is now no longer the case, as workers with lower levels of education are now leaving cities at greater rates than ever before and highly-educated workers are moving in. This trend is depicted in the graph below from the New York Times.

 

Jobs are clustering into fewer and fewer metro areas in the US

This trend is occurring while jobs are clustering in cities, particularly America’s largest metro areas. In fact, according to a study by the Brookings Institute, New York City, Chicago, San Francisco and Seattle have accounted for 90 percent of the increase in job density experienced by America’s 94 largest metro areas from 2004 to 2015. The remaining 90 largest metro areas saw a total increase in job density of just nine percent in that time frame, with 46 of those areas experiencing a decline.

To make matters worse, the jobs often open to people with lower levels of education (i.e., hospitality, retail, manufacturing, etc.) are clustering in urban areas at far lower levels than other major industries, limiting available opportunities. This takeover of high-wage jobs and the decline of low-wage work in cities has created what MIT Economist David Autor has called “a disproportionate polarization of urban labor markets.”

Costs of living and stagnant wages are forcing low-wage workers out

When high-wage jobs move to cities, the workers who take these jobs don’t just commute in, they move in with them and dominate the housing supply. As the housing supply goes down, the costs of living skyrockets, as many large metro areas across the country are experiencing.

No longer able to afford rent, workers in low-wage industries like those mentioned earlier are often forced to leave America’s big, expensive cities in search of lower costs of living. Research by Peter Ganong of the University of Chicago and Daniel Shoag of Harvard University in 2017 found that “a dramatic decline in the rate of income convergence across states and in population flows” leads to a “redirection of low-skill migration away from high-income places.”

Spatial mismatch of low-wage, hourly workers and employers

When jobs in these industries are available in the expensive cities, employers often have great difficulty finding people to fill them as low-wage workers move further away. This phenomenon is known to economists as “spatial mismatch.”

In February 2019, researchers at the Urban Institute used data from online job marketplace Snag to analyze the distance between hourly job seekers with high school diplomas in low-wage industries and the jobs they applied for in 2017. They compared data from 16 different metropolitan statistical areas (MSAs) across the country, with a focus on the San Francisco Bay Area and Columbus region of Ohio.

In San Francisco, Urban Institute researchers found the Snag employers have a significant shortage of job seekers “within a reasonable distance of their jobs” in a majority of neighborhoods. These are areas in the map below highlighted in dark pink. Very few areas had an oversupply of job seekers and an undersupply of job postings. These are the areas highlighted in gold. This should come as no surprise given the city planning department’s July 2018 report showing that households in San Francisco must bring home $180,000 per year to comfortably afford the median cost of rent (spending no more than 30 percent of income on housing).

While this analysis can be troubling, it still does not paint a full picture of how difficult it is for most to live in the city. Even public servants like teachers, who make between $55,000 and $67,000, are being forced to move outside of San Francisco to afford to live. To avoid lengthy commutes from the suburbs, some are renting out bunk beds in co-living spaces for upwards of $1,200 per month, and others are living in their cars, vans, or RVs despite some taking home six figure salaries. San Franciscans in these situations are not accounted for here.

In Columbus, Ohio there is a different situation. There are a mix of MSAs where Snag jobs outnumber seekers (dark pink areas), but also others where job seekers outnumber available jobs within a reasonable distance (gold areas). Like San Francisco, there are some jobs that are unaccounted for in the study, in this case industrial and logistical distribution jobs which are considered low-wage and are widely available in the area, but are not posted on Snag.

According to the study, these problems are not unique to San Francisco or Columbus, and neither city was on either end of the extreme. Across the 16 MSAs in the study, spatial mismatch was a significant problem. In the graph below, the blue bars represented the percentage of zip codes in an MSA where job posting exceed job seekers within a reasonable distance; the yellow bars represent zip codes where job seekers outnumber job postings within a reasonable distance.

Conclusion

Finding an effective solution to spatial mismatch is incredibly complex. The Urban Institute study identified a few policy proposals to begin to move the needle in connecting hourly workers to employers:

1. Strengthening career pathways
2. Creating housing near jobs and transit
3. Improving access to public transit

Cities across the country are attempting to implement different iterations of these policies, but many are early in their development, with results still to be determined. If one thing is clear, great efforts are being made to make cities fairer places to live for people across the income spectrum. Unfortunately, far more remains to be done and drastic change is necessary to prevent from completely driving low-wage workers away and widening the spatial mismatch gap.