The Surprising Disappearance of Labor Unions in Rural America

  • Union elections per worker were surprisingly high in rural areas during the 1960s and 1970s; the decline of unions was much more rural in nature than previously assumed
  • In rural areas, sectoral employment shifts were only minorly more impactful than in the nation as a whole.
  • Urban migration played virtually no direct role in the total decline of unions.

The labor movement in the United States was once a powerful force in the economy. The prevalence of unions, however, has been in retreat for many decades. Data from http://unionstats.com/ show that the share of private sector workers that are organized has fallen from 35% in 1955 to 6.5% today. What accounts for this drastic decline in labor unionization? Were all regions of the US impacted the same?

To answer these questions, we analyze data on union elections – the democratic process by which workers organize in the US. This flow of union formation is a good proxy for overall union prevalence. Like union membership, the number of union elections have been in dramatic decline since the 1970s. Interestingly, when we geocode these elections, and categorize them by their urban/rural locations, we find that the decline is much more severe in rural areas than in urban areas (we follow the Census definition that a county is urban if it contains 50,000 people or more).

Moreover, unionization was very common in rural counties during the 1970s. Although most elections happen in urban places, the per capita numbers tell a different story. We are surprised to find that there were around 4 times more elections per 10,000 workers in rural counties than in urban counties. Yet, there was also a much steeper decline in rural places. With such a high concentration and impact in rural counties, these findings suggest that labor union decline was much more rural in nature than previously supposed.

Zooming in on just Colorado, we confirm the pattern. Figure 2 shows a heat map of total change in elections per worker since 1975. Darker colors indicate larger declines in union elections per worker. Counties shaded yellow had no union elections in either year (or had slight growth). Bent, Park, Jackson, and Lincoln counties stand out as having the starkest disappearance of union formation—all notably rural.

Digging deeper, we engage the possibility that the decline is due to sectoral or regional employment shifts. Economists have theorized that unions faded because of the decline of manufacturing and the rise of services. To the extent that labor unions were concentrated in rural areas with manufacturing work, this theory could explain our differential trends. Similarly, as the US shifted to a more serviced-based economy with centralized work environments, workers relocated to urban areas where job growth was more robust. In essence, we consider that the union dynamics might be driven by restructuring of types of employment and how that varies between rural and urban areas.

To see if the decline of rural unions is due to changing economic activity across sectors or geographic regions, we use a shift-share decomposition method, which determines the portion of union election decline that may be attributed to a shift in employment shares. This method mathematically isolates changes in elections per worker into three terms:

  • The ‘Across’ term isolates the decline in elections due to changes in employment across sectors/regions. It shows how much of the total decline would have happened if election activity within each sector/region had stayed at 1975 If employment shifts are important, this term will be large.
  • The ‘Within’ term isolates the decline in elections due to changes in employment within sectors/regions. It shows how much of the total decline would have happened if sector shares/regional shifts had stayed at 1975 If changes within sectors/regions are key, e.g., the taste for unions changes within sectors/regions, this term will be large.
  • The ‘Covariance’ term represents the dynamic interaction between sectoral/regional shifts (across term) and the structural intra-sectoral/intra-regional shifts (within term). If there is a strong interaction effect when elections are declining both across and within sectors/regions, e.g., the employment shift aligns with changes in taste for unions, this term will be large.

When we performed this decomposition for the nation as a whole, we found that 41% of unionization decline comes from a sectoral shift; when we restrict to just elections in rural counties, the across sector decline only bumps up a bit to 45%. Additionally, we find that in the regional decomposition, i.e., measuring the role of urban migration, the across component is very small, indicating that across-region employment shifts had virtually no role in the decline of labor union elections, as shown in Figure 3.

Discussion and Conclusion

Given that the results show the decline in unionization is largely rural in nature and driven by within county factors, what might these factors be?

  • One possibility is that there are structural issues with the system of The election process is long and onerous; perhaps it has become impossible to navigate for rural workers.
  • Alternatively, employers in remote areas may be especially effective at preventing workers from unionizing. It is well documented that employer resistance has increased since the 70s.
  • Finally, rural workers may have developed a strong distaste for

Discovering the extent to which each of these explanations is at play is incredibly important. Workers in rural regions may bear severe negative effects of deunionization, such as low pay and harsh working conditions. They are also likely to be very isolated and vulnerable to market power. Should a worker feel undervalued or mistreated by their employer and wish to seek better conditions on the job market, they are unlikely to have as many options as their urban counterparts without long commutes. This lack of competition shifts bargaining power decidedly toward employers. And since rural regions often specialize in key sectors for the economy, including mining, logging, and agriculture, deunionization may have broader effects on productivity for the economy as a whole.

We conclude that rural union election decline is primarily due to forces within rural regions, such as aggressive firms, union funding being pooled into urban centers, or other factors unique to rural labor markets. Ultimately, it is not a story of population migration, and it is only partially a story of sectoral employment shifts. Further research should focus on determining whether the change is the result of shifts in taste, or the result of an increasingly anachronistic election process.

 

Scroll to Top