How diversified is Indiana’s economy?

Professor of Economics, Indiana University Northwest

Indiana's economy became more diversified between 2001 and 2024, as manufacturing accounted for a smaller share of employment.

Manufacturing has long shaped Indiana’s economy, supporting generations of workers, building strong supplier networks and helping communities develop the skills and infrastructure that industrial production requires.

Heavy dependence on a small number of industries also leaves local economies vulnerable when a major plant closes, demand falls or new technologies change how goods are produced. The practical question is how effectively Indiana and its local economies can withstand a downturn in a major industry without widespread employment losses.

This article introduces the Indiana Industry Diversification Index (IIDI) as a way to track how evenly employment is distributed across major industries. The index is calculated annually for Indiana and each of its 92 counties from 2001 through 2024, with a separate regional index for Northwest Indiana or Economic Growth Region 1 (EGR 1), which includes Jasper, Lake, LaPorte, Newton, Porter, Pulaski and Starke counties.

Indiana’s employment base became more diversified during this period, as manufacturing accounted for a smaller share of covered employment and education and health services expanded. County experiences varied substantially, with some counties developing a broader employment base and others becoming more concentrated.

Why measure industry diversification?

Specialization and diversification offer distinct advantages. Specialization can raise productivity by bringing together suppliers, skilled workers, industry knowledge and infrastructure built around one type of business, as Indiana’s manufacturing clusters demonstrate. On the other hand, diversification offers protection. When employment is spread across several sectors, weakness in one industry is less likely to affect the entire regional economy.

A more diversified economy is not automatically a stronger one. A region with an even mix of slow-growing, low-productivity industries may be less competitive than one built around a successful cluster of high-wage firms. The index can rise because new sectors are adding jobs or because a once-dominant industry is shrinking. The IIDI measures employment balance, not economic strength.

Constructing the index

The IIDI adapts an existing Gini-based industry diversification method to Indiana. The analysis uses 12 sectors derived from the Bureau of Labor Statistics’ Quarterly Census of Employment and Wages (QCEW) broad industry groupings: 11 private-sector categories and one category combining federal, state and local government employment. This level of aggregation permits consistent comparisons across counties and years while reducing, though not eliminating, the number of suppressed observations in smaller counties.

For geography r, industry j and year t, the employment share is:

where E is annual average employment and n is the number of industry groups. In this analysis, n equals 12.

The sector shares are ordered from smallest to largest and entered into the following finite-sector-normalized formula:

The index ranges from zero to 100. A value of zero represents complete specialization, with all employment concentrated in one sector, while a value of 100 represents an equal distribution across all 12 sectors. These endpoints are theoretical benchmarks, not targets that regional economies are expected to reach. A higher score indicates a more evenly distributed employment base, but it does not show that the underlying industries are productive, well paid or growing.

One of the 12 sectors, unclassified, holds only a few dozen workers statewide and is exactly zero in roughly seven of every 10 county-year observations. Because it is retained as one of the 12 sectors, the practical ceiling is roughly 91 rather than 100. Recomputing on the 11 populated sectors raises the level by about five points but barely changes the movement over time. Indiana’s index gain becomes 1.9 points rather than 1.7.

Data and industry definitions

The analysis uses annual average employment from the QCEW for 2001 through 2024, a consistent NAICS-based series spanning the early-2000s recession, the Great Recession, the expansion that followed, the 2020 pandemic disruption and the subsequent recovery.

Employment is grouped into 12 broad sectors: natural resources and mining; construction; manufacturing; trade, transportation and utilities; information; financial activities; professional and business services; education and health services; leisure and hospitality; other services; unclassified and government.

The first 11 categories capture private employment in each corresponding QCEW sector. Government is treated differently. It combines federal, state and local government employment across industries and is broader than the public administration category. Public schools, universities, hospitals and other publicly operated establishments are counted under government, while comparable private establishments are counted in their respective private-sector categories.

For the statewide index, county employment is first summed by sector and the index is then calculated from the resulting statewide employment shares. The same procedure is used for EGR 1. Regional employment is calculated as:

where c represents the counties in region r. Individual county scores are not averaged, because averaging would give equal weight to counties with very different employment bases and would not describe the employment distribution of the combined regional economy. Employment that QCEW reports at the state level but does not assign to a county is excluded.

Some QCEW employment figures are withheld to protect information about individual employers. Suppressed values were not counted as zero. For each county, observed sector employment was compared with the published total. When one sector’s employment was suppressed, the difference was assigned to it. When several sectors were suppressed, the difference was distributed using each sector’s employment pattern in nearby years, or statewide sector shares for that year when no adjacent-year information was available. Reconstructed employment accounts for about 1% of statewide employment.

Roughly half of the 2,208 county-year observations matched their published totals exactly and none differed by more than three jobs, with the remainder running slightly positive because negative residuals were set to zero. The final dataset covers 92 counties, 24 years and 12 sectors, producing 26,496 county-sector-year observations.

Has Indiana become more diversified?

Indiana became modestly more diversified between 2001 and 2024, although most of the increase occurred before 2011 and the index fluctuated afterward (see Figure 1). The statewide IIDI rose from 49 in 2001 to 50.6 in 2024, a gain of 1.6 points. It climbed steadily through the 2000s, peaked at 51.1 in 2011, then drifted down over the following decade before falling to 50 during the pandemic in 2020 and recovering by 2024.

Covered employment grew over the same period, from about 2.79 million to 3.08 million, a gain of roughly 288,400 jobs, or 10.3%. The state became somewhat more diversified while its employment base expanded, although the small overall index change indicates that Indiana’s broad employment structure remained relatively stable.

Figure 1: Industry diversification in Indiana and Northwest Indiana (EGR 1)

Note: Higher values indicate a more evenly distributed employment structure across the 12 employment sectors. EGR 1 includes Jasper, Lake, LaPorte, Newton, Porter, Pulaski and Starke counties.
Source: Author’s calculations using U.S. Bureau of Labor Statistics, Quarterly Census of Employment and Wages data

Diversification through growth and decline

The increase in the index was driven primarily by the declining share of employment in manufacturing and the expansion of education and health services (see Figure 2). Manufacturing employment fell from 614,600 in 2001 to 521,800 in 2024, reducing its share of covered employment from 22% to 16.9%. Education and health services moved in the opposite direction, adding nearly 179,700 jobs and increasing its share from 11.3% to 16%.

Changes in the remaining sectors were smaller. Professional and business services increased its share from 8.2% to 10.4%, while leisure and hospitality rose from 9.2% to 10.1%. Trade, transportation and utilities was Indiana’s largest broad employment sector in 2024, although its share edged down from 20.3% in 2001 to 19.7% in 2024. Information, financial activities and government also lost employment share. The higher index reflects both service-sector growth and the declining concentration of employment in manufacturing. It does not show that expanding sectors replaced manufacturing jobs with jobs of equal wages, productivity or economic value.

Figure 2: Indiana employment shares by sector

Note: Sectors are sorted by the change in share between 2001 and 2024.
Source: Author’s calculations using U.S. Bureau of Labor Statistics, Quarterly Census of Employment and Wages data

Indiana is not one economy

Indiana’s statewide score masks large differences from one county to another. The average county IIDI increased from 40 in 2001 to 42.5 in 2024, while the median rose from 38.9 to 42.9. County scores also converged. The standard deviation fell from 7 to 6.2 and the interquartile range also decreased from 9.3 to 7.6. The distribution shifted from roughly symmetric to distinctly left-tailed, as a small group of highly concentrated counties fell further behind while the middle of the distribution tightened.

The statewide index sits well above the average county (50.6 as opposed to 42.5 in 2024) because summing counties before calculating shares averages away local concentration. The statewide figure describes Indiana’s combined employment structure. The county average describes the typical county. The distance between them is itself a measure of how uneven the state is internally.

Table 1 shows the five highest and five lowest county diversification scores in 2024. Hamilton County recorded the highest score and Martin County the lowest. Hamilton County also illustrates why the index score and the change must be read together. It held the top rank in both 2001 and 2024 while its score fell by 4.5 points, one of the 10 largest declines in the state. Vanderburgh County slipped from second to seventh. Rankings do not measure economic success. A low score may reflect a strong industry cluster or a single large employer, and a high score does not imply faster growth, higher wages or stronger productivity.

Table 1: Highest and lowest Indiana county diversification scores, 2024

Source: Author’s calculations using U.S. Bureau of Labor Statistics, Quarterly Census of Employment and Wages data

Which counties changed the most?

Randolph County recorded the largest increase in diversification between 2001 and 2024, rising from 37.8 to 52.4 and moving from 55th place statewide to second. Boone County recorded the largest decline, falling from 49.9 to 35 and dropping from seventh place to 83rd (see Figure 3).

The contrast shows why changes in the index must be examined alongside employment growth and sector composition. A rising score can reflect new sources of employment or the decline of a dominant sector; a falling score can reflect greater exposure to one industry or successful specialization in a rapidly growing sector.

Figure 3: Largest increases and declines in county industry diversification, 2001–2024

Note: This figure shows the 10 largest gains and 10 largest declines among the 92 counties. Ohio County is omitted. Between 50% and 80% of its sector employment was estimated from suppressed data prior to 2017, compared with roughly 3% afterward, suggesting that the apparent increase is largely attributable to changes in disclosure practices.
Source: Author’s calculations using U.S. Bureau of Labor Statistics, Quarterly Census of Employment and Wages data

Northwest Indiana: Diversification beyond heavy industry

Northwest Indiana (or EGR 1) has long been identified with steel and heavy manufacturing, but its employment base also includes transportation, healthcare, education, government, construction and business services.

The region’s diversification index increased from 50 in 2001 to 50.7 in 2024, a gain of 0.7 points, reaching its high point of 51.3 in 2005 and its low point of 49.6 in 2015. Northwest Indiana began the period with a slightly more diversified employment structure than the state as a whole, by about 1.1 points. Indiana moved ahead in 2008 and maintained that lead through 2020. Northwest Indiana regained a small advantage in 2021. By 2024, the two index values were nearly identical, with the region ahead by 0.1 points.

Total covered employment in the region changed very little, increasing from 318,700 jobs in 2001 to 320,500 in 2024, a gain of 1,800 jobs, or 0.5%. That small overall change masks substantial shifts among industries (see Figure 4). Manufacturing employment declined from 58,400 jobs to 44,900 and government employment fell by approximately 8,900 jobs, while education and health services added 15,200 jobs and professional and business services increased by 6,600. Leisure and hospitality also expanded. Trade, transportation and utilities remained the region’s largest broad employment sector, with employment almost unchanged.

These changes left Northwest Indiana more evenly balanced across sectors, but did not produce broad-based employment growth.

Figure 4: Northwest Indiana (EGR 1) employment shares by sector

Note: EGR 1 includes Jasper, Lake, LaPorte, Newton, Porter, Pulaski and Starke counties. Sectors are sorted by the change in share between 2001 and 2024.
Source: Author’s calculations using U.S. Bureau of Labor Statistics, Quarterly Census of Employment and Wages data

The regional total also hides important differences among the seven counties. Newton County recorded the largest increase, rising 13.8 points, followed by Starke County at 9.6. Porter County rose 2 points, placing it fifth among Indiana counties in 2024. Jasper County’s score was essentially unchanged, Lake and LaPorte counties recorded modest declines and Pulaski County fell 1.3 points. Structural change has unfolded at different speeds and through different industries across the region.

Using the index

A low or declining score can signal that a region is becoming more dependent on a small number of sectors. That does not mean those industries should be weakened. It means regional leaders should watch their competitiveness, technological direction and exposure to outside shocks. Because the index can be updated each year, it also offers a practical way to ask whether a given investment broadens the employment base or deepens dependence on a sector that already dominates.

Diversification and job growth do not always move together. A region can become more diversified while losing employment, or more concentrated during a period of rapid growth. The IIDI is best read alongside total employment, employment growth, the largest sector’s share, average annual wages, industry location quotients and measures of employment volatility. The kinds of industries expanding also matter. Growth in sectors that pay well, support technological learning or serve outside markets may contribute more to long-term development than an even spread of employment for its own sake.

Limitations

The IIDI measures how evenly employment is distributed across sectors. It does not measure wages, productivity, output, innovation or export performance. Additionally, 12 broad categories conceal differences within industries. Manufacturing includes primary metals, pharmaceuticals, motor vehicles and much more, so two counties with the same manufacturing share may face very different strengths and risks.

QCEW data are reported by place of work, showing where jobs are located but not where workers live or how easily residents can reach those jobs. They also do not fully capture self-employed workers. Results for counties with heavy suppression warrant particular caution, because reconstructed employment can move the index when disclosure practice changes from one year to the next. Finally, the index does not identify an ideal employment structure. It does not suggest that every sector should account for the same share of employment.

Conclusion

Indiana became modestly more diversified between 2001 and 2024, with the statewide IIDI rising from 49 to 50.6 as education and health services and professional and business services expanded and manufacturing’s share of employment declined. The change was not uniform. Some counties developed a more evenly distributed employment base while others became more concentrated, and county scores converged even as the distance between the most and least balanced counties remained wide. Northwest Indiana also became slightly more diversified, but total regional employment changed very little because service-sector gains largely offset losses in manufacturing and government.

A rising index can accompany employment growth, employment decline or a redistribution of existing jobs. The IIDI should be used to identify changes in employment concentration, not to rank economic success or prescribe an equal distribution of jobs across industries. Its contribution is to show where dependence on a small number of sectors is increasing and where the employment base is becoming more balanced.

References

• Regional Economic Research Institute. “Industry Diversification Project.” Lutgert College of Business, Florida Gulf Coast University. https://www.fgcu.edu/cob/reri/idp/.
• Gini, Corrado. 1921. “Measurement of Inequality of Incomes.” The Economic Journal 31 (121): 124–125. https://doi.org/10.2307/2223319.
• U.S. Bureau of Labor Statistics. Quarterly Census of Employment and Wages. https://www.bls.gov/cew/.