Measuring Local Industry Employment Diversity with the Hachman IndexAugust 25, 2020 Have you ever heard the expression don’t put all your eggs in one basket? That saying refers to the notion that diversifying risk is generally a good strategy whether you are producing chicken eggs or are charged with growing your regional economy. Industries behave differently in response to changes in markets, products, or the business cycle. If a regional economy has most of their employment base concentrated in one or just a few industries and that industry or industries experience a downturn, then the whole regional economy is at risk for a downturn as well. Examples of why diversification might be a good economic strategy can be seen dotting many areas of the western United States, such as Bodie, California.
During its boom, Bodie boasted 2,000 buildings, 65 saloons, multiple newspapers and, in classic Wild West fashion, nearly a murder a week. When the gold veins petered out, so did the entire settlement. Today, you can see old pianos, nearly furnished living quarters and kitchen sets, as if no one had touched them in over 100 years. When the town was abandoned during a financial decline, many of its residents quickly left for greener pastures, leaving many of their belongings in their homes. The Bodie State Park sits as a relic to the wild west of an era gone by. It stands as a tale of caution for other communities that rely on just one driver – in this case gold – on which to base their economic prosperity.
As the Oregon Office of Economic Analysis economist Josh Lehner noted in a blog post regarding industry diversity in Oregon, “Industrial diversity, in and of itself, is not necessarily a good or bad. If a region has one big industry, like the energy industry in North Dakota today or the timber industry in Oregon in the 1970s, the region can do extremely well when that one industry is booming. The problem arises when that one key sector is down. Then your regional economy suffers more as there are fewer other types of businesses to drive growth.” But Josh goes on to note, “Overall a more diverse economy is better able to withstand different types of recessions…spreading a region’s eggs across more baskets tends to create more resilience in the long run.”
If the presumption is a diversified local economy is a desirable attribute, how can we measure how diversified a local economy is? One way to measure economic diversity using payroll employment data is called the “Hachman Index.” Frank Hachman with the Bureau of Economic Research at the University of Utah developed a model to measure economic diversity using payroll employment data. The index measures how closely the employment distribution of the analysis region resembles that of the reference region, usually the state or the U.S.
The closer the analysis region’s employment mix matches the reference region’s mix, the higher the value of the Hachman Index. The index has a maximum value of one – meaning the analysis area’s employment mix is exactly the same as the reference region’s mix of industry employment. This analysis uses 2019 Quarterly Census of Employment and Wages (QCEW) employment data for Oregon counties. The index value is affected by the choice of reference area. As state and county employment data are readily available, this analysis focuses on county employment mix relative to Oregon’s mix of employment by industry.
The Hachman Index is calculated as the inverse of the weighted sum of the location quotients, by industry for each analysis county, across all industries. A location quotient (LQ) is the fraction of a county’s employment in a particular industry divided by the fraction of the reference area’s, in this case Oregon’s, employment in the same industry. The LQs are weighted by the share of the county’s employment in a particular industry. Counties with a large share of employment in only a few key industries, which differ substantially from the share of employment for those industries statewide, will have a relatively large weighted sum of LQs and subsequently a relatively low Hachman Index value (since it is the inverse of the weighted LQs). Conversely, counties which more closely reflect the statewide employment distribution will have relatively small weighted LQs and a relatively high Hachman Index value.
In 2019, Lane County boasted the most diverse industry mix of all Oregon counties, followed closely by Clackamas and Multnomah. For the most part, this diversity ranking among counties remains fairly stable over time with a few exceptions. Columbia County was the 21st more diverse economy in Oregon back in 1999. By 2019, it was the ninth most diverse mix of industries among Oregon counties. The top six counties with the most diverse industry mix in 2019 were among most diverse seven counties back in 1999. Benton County leaped from the 26th most diverse economy to the seventh most diverse during those two decades. See below for a little more background about why we saw its industry diversity ranking rise since 1999, there are other factors leading to that rise beyond just simply adding jobs in different industries.
Limitations of the Data
This analysis uses SIC and NAICS data as a time series. That does create some issues not related to true changes in economic diversity, primarily due to non-economic code changes. If you’re in a relatively small county with a relatively big firm or cluster of firms in an industry, and there’s a big decline or closure(s), the county’s economic diversity score can improve because employment becomes more evenly spread among industries, even if the community is not better off as a result.
A good example of a non-economic code change affecting the diversity score is the huge jump in Benton County’s ranking, from 23rd most diverse in 2012 to seventh most diverse in 2019. That’s due in large part to the non-economic switch of universities from state government to local government. In 2012, the LQ for NAICS 611 in state government in Benton County was 14.9. In 2016, that employment moved into NAICS 611 local government, with a much smaller LQ of 3.2. As a result, when you calculate the Hachman Index, the big change of employment to a smaller LQ industry in a relatively small county results in an industry mix that looks less concentrated in education, and therefore more diverse. So even though very little changed, the numbers look like something big changed in Benton County.
Here’s one final thought that Josh Lehner shared in his recent blog post. There are good ways and bad ways for a region to become more diverse. The better way is when an economy adds jobs and new businesses in industries that it may have had few of to begin with. Another way a region can become more diverse is when it loses its specialty industries – an example would be the decline of the timber industry in Oregon. Oregon is now more like the U.S. mostly due to growing jobs in other sectors, but also due to losses in timber industry jobs.