Skip to main content

Industries, Job Growth, and Poverty Trends

The shares of a county’s employment that are in each major industry classification are correlated with the county's poverty rate. Employment shares in healthcare and public administration, for example, are positively correlated with poverty rates, while employment shares in professional services and construction are negatively correlated with poverty rates. In this analysis, we examine some of the changes in these correlations in recent years. We will also look at the changes in industry employment that have accompanied changes in county poverty rates.

Between 2007 and 2012, the percentage of Americans living in poverty increased from 12.5 percent to 15 percent. Both poverty rates and poverty growth rates vary a lot across counties. One-quarter of the US population lives in counties in which poverty rates increased 4.3 percentage points or more between 2007 and 2012. Another quarter of the population lives in counties where poverty rates increased by 1.8 percentage points or less. Less than 5 percent of Americans live in counties that experienced unchanged or declining poverty rates.

Figure 1: Estimated County Poverty Rate: 2012
Figure 2: Change in County Poverty Rate: 2007 - 2012

Employment in certain industries could be correlated with poverty for many reasons. One possibility is that the industry serves a clientele that is disproportionately poor. The share of employment related to social assistance and welfare administration could be higher in high-poverty counties because workers in this industry manage the public programs designed to assist families in poverty. In the data, this correlation is indeed positive and significant.

Another possibility is that industries that employ low-skilled, low-wage workers might have employees whose household income remains below the poverty thresholds. A higher share of employment in these industries would therefore be associated with higher rates of poverty. Two low-paying industries, agriculture and accommodation and food service, display significant positive correlations between their shares of employment and the poverty rate. Accommodation and food service’s correlation has risen from a value close to zero in 2007. (For median pay by industry, see this BLS news release.)

Following similar reasoning in the opposite direction, counties with higher shares of employment in high-wage industries should have a smaller portion of their residents in poverty. Several high-wage industries do display negative correlations with poverty, including finance, information, management, and professional services.

Still, a number of the correlations between employment share and the poverty rate await an alternate explanation. Among the industries with the highest positive correlations with poverty are healthcare, public administration, utilities, and education. These industries’ median wages rank mid to high among all industries. And shares in the low-wage retail and arts industries display negative correlations with the county’s poverty rate.

Figure 3: Correlation between Industry Share of Employment and Poverty Rate

The broad categories of mining and manufacturing display some of the smallest overall correlations with poverty in 2012. However, disaggregating these categories reveals interesting trends and variation.

As we have seen in the Fourth District, the expanding oil and gas extraction industry has been a source of relatively strong employment growth in an otherwise slow recovery. In 2007, all three subcategories of mining were positively associated with poverty. However, by 2012 all three subcategories of mining employment had dramatically decreased their correlation with poverty. The decline in the correlation for oil and gas was larger than that for the mining subcategory that includes coal.

Correlation between Mining Subcategory's Share of Employment and Poverty Rate

Mining subcategory Correlation
2007 2012
Mining (excluding oil and gas) 0.095* 0.053*
Oil and gas extraction 0.116* 0.048*
0.129* 0.034

Note: An asterisk indicates significance at the .05 level. Observations are 3,132 counties. Calculations are weighted by population.
Sources: County Business Patterns; Census of Governments; American Community Survey; and authors' calculations.

There are pronounced differences among the 21 subcategories of manufacturing as well. Employment shares related to wood and textiles are positively correlated with poverty at the county level. In contrast, fabricated metals and machinery are negatively correlated with poverty. Computer makers are categorized in manufacturing even if much of their production is offshore. These firms, too, are generally located in lower-poverty areas. One of the notable changes in manufacturing since 2007 is that the positive correlation between poverty and petroleum and coal products manufacturing has declined. The shift in this mining-related manufacturing is similar to the poverty-correlation declines discussed above in the mining industry.

Correlation between Manufacturing Subcategory’s Share of Employment and Poverty Rate

Manufacturing subcategory Correlation
2007 2012
Wood products 0.152* 0.129*
Apparel 0.129* 0.129*
Food 0.126* 0.105*
Textile mills 0.086* 0.077*
Paper 0.064* 0.056*
Textile product mills 0.046* 0.055*
Beverage and tobacco products 0.021 0.036*
Leather and allied products 0.019 0.028
Petroleum and coal products 0.036* 0.018
Furniture and related products 0.019 0.016
Transportation equipment 0.022 0.015
Nonmetallic mineral products 0.033 0.015
Primary metals 0.025 0.009
Plastics and rubber products 0.017 0.002
Electrical equipment, appliances, and components 0.008 —0.015
Chemical —0.008 —0.018
Machinery —0.009 —0.048*
Fabricated metal products —0.021 —0.050*
Miscellaneous —0.075* —0.056*
Printing and related support activities —0.124* —0.111*
Computer and electronic products —0.227* —0.213*

Note: An asterisk indicates significance at the .05 level. Observations are 3,132 counties. Calculations are weighted by population.
Sources: County Business Patterns; Census of Governments; American Community Survey; and authors' calculations.

If we focus in on the counties that had the largest and smallest increases in poverty, we can see interesting differences in their job growth. Counties with the greatest increases in poverty experienced job losses in construction that average over 2 percent of their 2007 labor force. Their equivalent losses in manufacturing were over 1.5 percent. Two industries in which job gains were similar in the worst- and best-performing counties were education and healthcare.

Figure 6: Average Change in Industry Employment for Countries with Largest and Smallest Increases in Poverty: 2007 - 2012

The counties that kept their poverty rates down are distinguished by smaller job losses or larger job gains in every category. The top-performing counties had notably higher job growth in accommodation and food service and mining. The counties that experienced small increases or declines in their poverty rate added more than twice as many workers in social assistance and welfare compared to the counties that had the greatest increases in poverty rates.

Upcoming EventsSEE ALL

  • 05.29.14

    2014 Inflation, Monetary Policy, and the Public

    Highlighting research and advances in data requirements for macroprudential policy, systemic risk measurement, and forecasting tools.

  • 06.25.18

    International Journal of Central Banking Conference

    CALL FOR PAPERS The International Journal of Central Banking (IJCB) is organizing its 2018 annual conference, "Ten Years after the Global Financial Crisis: What Have We Learned about Ensuring Financial Stability?" The conference will be hosted by De Nederlandsche Bank (DNB) in Amsterdam, The Netherlands, on June 25-26, 2018, and is being organized by Tobias Adrian, Harrison Hong, Luc Laeven, and Loretta Mester.