96 years of economic insights for Indiana

The IBR is a publication of the Indiana Business Research Center at IU's Kelley School of Business

innovation for business startups sitting around a table with ideas

Assessing regional entrepreneurial readiness

Economic Research Analyst, Indiana Business Research Center, Indiana University Kelley School of Business

Each of the past two industrial revolutions, accompanied by unprecedented technological advance, brought radical change to mankind’s working style and significantly improved mankind’s living. Some say we are marching through a third industrial revolution—the Digital Revolution, brought about by digital computing and communication technology that have made things “easier” for us. Nowadays, mechanical arms and robots have “freed” men from standing in front of assembly lines and working in some hazardous environments. Information and knowledge have never before been so freely and quickly exchanged. With just a few clicks on your phone, your food or groceries can be delivered to your door. You don’t have to go to your doctor’s office to get a checkup. And believe it or not, soon you may have an autopilot for your chauffeur …

Our world is changing. Changes bring together innovation and new opportunities. The old American dream “that you go to college, get a degree, go directly into a well-paying job in your field, work there for 30 years, then get a pension plan and a Rolex when you retire” has faded. The new American dream is “an entrepreneurial journey that is unlike anything that has ever existed before. The barrier to entry to becoming an entrepreneur is literally the lowest it’s ever been.”1

Recognizing that entrepreneurship is an important engine of growth in the economy, economic development practitioners (EDPs) should not rely solely on employment and wage figures to assess their region’s economic well-being, but should focus on promoting a nurturing environment—both physical and cultural—for business growth. Having this in mind, the Indiana Business Research Center (IBRC) developed a timely and comprehensive assessment scheme in the form of the E-primed Entrepreneurship Readiness Index, or simply the E-primed Index. This is just one part of the Regional Economic Development (RED) project funded by the U.S. Economic Development Administration.

What is the E-primed Index?

The E-primed Index leverages and distills a large portion of the RED research and analysis utilizing both conventional and unconventional data. The composite index consists of seven components, each one directly related to economic development and a region’s ability to foster business startups and growth:

  1. Innovation capacity as the entrepreneurship (E-ship) ecosystem support
  2. Social capital as the E-ship enhancer
  3. Inter-industrial linkages as the E-ship propellant
  4. Regional economic cohesiveness as an E-ship attracter
  5. Regional personality and cultural characteristics also as an E-ship attracter
  6. Arts and design and E-ship web resource utilization as signals for startups and business formation
  7. Social media sentiment as signals for E-ship activities and economic growth

The E-ship ecosystem support (component 1) is a repackaging of elements in Innovation Index 2.0—an IBRC legacy product used to assess regional innovative capacity, a key facilitator for growth opportunities. It combines both input measures (such as human capital, high-tech industries and occupations, venture capital, and community banking), as well as output benchmarks (such as gross domestic product per capita, unemployment and poverty rates).

Social capital (component 2) goes beyond the conventional interpretation of social capital as the propensity of residents to participate in community and organizational activities and reaches out for much broader social contents. Its thematic elements include crime; religion, culture and civic behavior; organizations, local institutions and participation; and demographic diversity and fractionalization. Social capital is considered a hidden force that enhances E-ship through reoccurring social interaction in a society.

Industry clusters provide an important regional economic environment that fosters the growth of E-ship. New firms are easier to start with the support of well-organized supply-demand chains in a cluster. Incumbent firms benefit from sharing labor pools and knowledge spreading within a cluster, maintaining their competitiveness and innovation momentum. The inter-industry linkages measure (component 3) looks at two venues of cluster growth: industrial concentration (or strength) and adaptability.

Cluster concentration emphasizes the geographic concentration of interdependent industries, connected with input-output linkages, taking advantage of input sharing, knowledge spillovers and labor market pooling. In this category, we constructed the well-known Herfindahl Index and National Average Index—the former measuring the absolute concentration of a cluster and the latter measuring the relative concentration in comparison to the national benchmark. We also included, by the effort of our Pennsylvania State University (PSU) partners, the proximity-adjusted location quotient (PA-LQ), an augmented version of LQ that takes into account both the within-region input-output linkages and the spatial linkages from neighboring regions.2

Adaptability takes a dynamic view on how a region’s industrial profile evolves over time—measured by cluster correlation—and how it adjusts itself in response to extraneous incidents—represented by an industrial structure shift as a result of the Great Recession. The cluster correlation, borrowing the concept of portfolio correlation in finance, measures the average sectoral employment growth correlations within a region. It assesses the degree to which a region’s industrial portfolio follows the national/regional trend. The structure shift, also researched by our PSU partners, measures the distributions of employment changes among local industries relative to the national economy before and after the Great Recession. In other words, it measures how much an industry structure has reorganized.

Historically, firms co-locate to avoid the high cost of transporting goods and services. However, even in the current age where some argue for the “death of distance,” it continues to exist. In fact, urban agglomeration is taking place at an unprecedented pace not before seen even in the Industrial Revolution era. “Being close,” or proximity, in its different forms benefits the regional economy by producing “externalities” that result in increasing returns to scale. Naturally, these benefits accrue to regional entrepreneurs, where it has been shown that entrepreneurship is strongly linked to the geographic concentration of firms, people and even knowledge. Economic cohesiveness (component 4) quantifies proximity in three dimensions:

  1. Proximity to customers and suppliers (input-output proximity)
  2. Proximity to workers (labor market pooling)
  3. Proximity to knowledge (knowledge spillovers measured by patent citation)

Fundamentally, entrepreneurship is a human behavior and is, therefore, affected by certain psychological traits. Behavioral economists, sociologists and other social scientists have demonstrated that culture and values represent a vital set of factors with direct impact on economic development and entrepreneurship. The regional cultural characteristics measure (component 5), an effort by our University of San Diego (UCSD) partners, emphasizes the relationship that regional personality and cultural dimensions have on entrepreneurship. Only metrics with considerable power to predict entrepreneurship were used, and there are 16 domains, such as agreeableness, openness, belief in science and tolerance.

The construct of the above five components relies on conventional large-scale survey data; the next two components utilize unconventional internet streaming data. We value the unconventional data sources as a means to understand the phenomena of E-ship that are often difficult to capture with conventional data.

Bearing in mind the question “What’s the propensity for residents/workers in an area to visit certain websites relevant to starting, running and growing a business?,” the arts and design and E-ship web resource utilization measure (component 6) uses web behavior of three classified E-ship audiences (web users of design, arts and E-ship resources) as a signal for a region’s propensity for entrepreneurship. These unique data were provided by Dstillery, a marketing firm that uses advanced data science techniques to identify audiences for a diverse set of clients based on web behavior distilled from billions of website visits each day.

The majority of the U.S. population uses at least one online social media service, such as YouTube, Facebook, Instagram and Twitter. The content shared on social media provides large-scale, high resolution data on present socioeconomic conditions among their user base. This data set is unique in the sense that it provides unsolicited self-reports from a very large number of individuals. The social media and sentiment (component 7), by the effort of our partners at the School of Informatics of Indiana University, made advances in Natural Language Processing and Complex Adaptive Systems modeling to derive indicators of entrepreneurial conditions—the E-ship sentiment—using geo-located Twitter data.

For a complete list of data items, time frame and sources, please refer to the appendix.

Methodology

Our modeling strategy for building the E-primed Index employed a hierarchical factor analysis. We first collected data, to our best knowledge, for each of the seven theoretical components (listed above) and performed a factor analysis separately for each component to select input measures and form grouping patterns, which essentially aims to discover latent factors that bind the inputs together.3 Second, we constructed an index for each of these components by combining their associated latent factors, weighted by the “importance” (or contribution) of these factors—the highest weight is given to the principal factor, the second-highest to the second factor, and so forth. Last, we performed a factor analysis on the seven component indices to estimate their underlying latent factors, which were then combined, in a similar fashion to constructing the component indices, to form the grand E-primed Index. We scaled the indices to a hundred basis. In general, values higher than 100 are considered above the U.S. average, while lower than 100 are below the U.S. average.

Figure 1 provides a clearer picture. The blue blocks are the theoretical components—the grand E-primed and its seven founding components; yellow blocks are latent factors, upon which the indices were constructed; gray blocks are data items (or inputs) we collected. Here, only the top three most significant inputs are listed. The decimals inside parentheses are the shares of latent factors in relation to their theoretical components and were used as weights for the indices.

Our model shows that the regional entrepreneurial readiness is driven by three underlying forces: ecosystem support, cluster growth and economic cohesiveness, and regional cultural characteristics.

Of these three, the ecosystem plays a more vital role than the other two, as evidenced by its highest weight (40%). Each one of these underlying forces governs a subset of the seven components (for example, ecosystem support is made of component 1, 2, 6 and 7). Each component is also driven by its own underlying forces. For example, innovation capacity (II 2.0) is comprised of growth potential, regional economic wellness, and business formation and productivity, among which the growth potential is the dominant factor accounting for 44% of innovation capacity.

Figure 1: Hierarchical structure of the E-primed Index

Diagram showing the index components. Full details available in the appendix.

Source: Indiana Business Research Center

How to use E-primed

Our goal was to create an easy-to-compare method of assessing the readiness of a region’s business environment. The E-primed Index makes it easy for EDPs to learn the overall position of their region. It also allows them to diagnose the region’s weakness and strengths through the sub-components. All indices will be available this summer on StatsAmerica (www.statsamerica.org).

To illustrate the E-primed Index, let’s consider the example of Morgan County, Indiana (see Figure 2). Morgan County’s E-primed is 99.3, slightly below the U.S. average. But this doesn’t tell us much in isolation. What factors are contributing to this score? Its scores for the three E-primed factors—ecosystem, cluster growth and cohesiveness, and cultural characteristics—are 96.1, 101.8, and 100.5, respectively. Clearly, the ecosystem factor is acting as a drag on the headline E-primed. If digging further into this factor, we see that the drag doesn’t come from innovation capacity, whose index score is above the average at 101.9, but rather from E-ship audience web activities (score at 90.3) and social capital (94.8). We further disaggregate social capital into three underlying factors, social assistance (95.1), membership (84.5) and religious engagement (91.5), and it becomes clear that membership is particularly lacking under this category. Membership includes features such as voters’ participation, number of nonprofit organizations and nonfamily households, etc. A low score could mean that Morgan County’s voter participation is low, or there aren’t many nonprofit organizations. In other words, what seems to be missing for the county is community engagement, and this is an area that EDPs might look into for improvement.

Figure 2: Partial E-primed for Morgan County, Indiana

Diagram showing the 3 factors along with the components of the Ecosystem Support factor. Described in prior paragraph.

Source: Indiana Business Research Center

Discussion

The following discussion will focus on the grand E-primed Index and the cluster growth component, where policy interventions are most likely to take place.

Figure 3 presents the geographic distribution of the E-primed Index among Indiana counties. Not surprisingly, counties within large metropolitan areas, such as Marion, Hamilton and Elkhart counties, dominate smaller counties in terms of entrepreneurial readiness. Hamilton County earned first place with a score of 109.3. It sits just north of Indianapolis and is packed with financial institutions, life science and information technology industries. Given that high-tech innovative capacity is the number one factor driving regional E-ship, Hamilton County well deserves its place. Other large “winners,” such as Marion (health care and bio-pharmaceutical manufacturing), Elkhart (RV capital of the world), Tippecanoe (Purdue University), Hendricks (logistics), Bartholomew (home of Cummins and Toyota) and Monroe (Indiana University’s main campus), are all leading in innovation either in knowledge creation or high-tech production—or both. We also see “spillovers,” as higher-score regions tend to cluster together.

Figure 3: The E-primed Index for all Indiana counties

Map: 14 counties = more than 103; 27 counties = 100.1 to 102.9; 21 counties = 99 to 100; 30 counties = less than 99.

Note: Median and mean are both 100.
Source: Indiana Business Research Center

Figure 4 just looks at the counties with less than 65,000 residents. Even for small/midsize counties, size still matters: Midsize counties (population over 25,000) exhibit higher levels of entrepreneurial readiness compared to smaller counties. The spillover effect is evident here: Midsize counties in northeast and central Indiana—Noble, DeKalb, LaGrange, Whitley, Boone, Clinton and Jackson counties—those close to the E-ship active basins also tend to have higher E-primed scores themselves. These regions are likely to benefit from their big and “wealthier” neighbors through inter-industry linkages. Several southwestern counties—Posey, Dubois and Gibson—that are auto-manufacturing hot spots also earned high scores. Small counties with a pillar industry, such as Newton (landfill), Spencer (Holiday World amusement park), Randolph (burial casket manufacturing) and Carroll (pork manufacturing), also show good standing.

Figure 4: The E-primed Index for Indiana’s small/midsize counties

Map: 12 counties = more than 102; 31 counties = 99.1 to 102; 10 counties = 98 to 99; 14 counties = less than 98.

Note: Counties shown have a population less than 65,000 as of 2017.
Source: Indiana Business Research Center

The maps in Figure 5 present the geographic distribution of the cluster growth index and its latent factors among Indiana counties. Two characteristics of cluster growth emerge: specialization and adaptability (it is explained by the sectoral growth correlation within a region and its labor reorganization in response to the Great Recession. Regions exhibiting larger positive sectoral growth correlations indicate a better integrated economy, and they also tend to be more adaptive to changing economic conditions, such as the Great Recession). There is an inverse relationship between specialization and adaptability, as is seen by contrasting color shades of part (b) and (c) of Figure 5. Thus, the “too-big-to-change” story seems to apply to industry too.

A general sentiment is that counties exhibiting low strength in cluster growth—they either do not have specialized industries or have gone through little change after the economic shock—are among the least “prepared” regions for business growth (there is a great overlap of lighter-hued areas on both Figure 5a and Figure 3 of the E-primed Index); more adaptive counties on the other hand show stronger E-ship readiness (there is a great overlap of darker-hued areas on both Figure 5c and Figure 3). This is especially true for counties in central Indiana. Some counties are ranked higher in terms of both specialization and adaptability (for example, Elkhart, Dubois and Decatur), and these counties also have greater potential for business growth. However, for many smaller counties, specialization alone might be the key to business success.

Figure 5: The cluster growth indices among Indiana counties

a) Cluster growth index

Map: 18 counties = more than 103; 16 counties = 101.1 to 103; 22 counties = 99.1 to 101; 17 counties = 97 to 99; 19 counties = less than 97.

Note: Median and mean are both 100.
Source: Indiana Business Research Center

b) Specialization index

Map: 19 counties = more than 103; 15 counties = 99.1 to 103; 17 counties = 97.1 to 99; 22 counties = 96 to 97; 19 counties = less than 96.

Note: Median and mean are both 100.
Source: Indiana Business Research Center

c) Adaptability index

Map: 20 counties = more than 107; 14 counties = 103.1 to 107; 16 counties = 100 to 103; 25 counties = 96 to 100; 17 counties = less than 96.

Note: Median is 100 and mean is 101.
Source: Indiana Business Research Center

One might question: Is this E-primed Index really a good indicator for entrepreneurship capacity? And first of all, what is entrepreneurship? Is anyone self-employed (such as an Uber driver or someone holding a side job mowing lawns) an entrepreneur? The key element of E-ship lies on “novelty” (or innovation)—whether that’s a new product or service, a new location, or expanding to a new market. That said, the percentage of self-employed in a region is generally not a good measure for E-ship, which itself is difficult to measure precisely. Therefore, to examine the validity of the E-primed Index, we looked at the total number of FDI jobs announced during 2012-2016 and used it as a proxy for E-ship activities since studies have found a positive linkage or impact between FDI and E-ship of local businesses.4

In Figure 6, bubbles represent a county’s proportion of total FDI jobs from 2012 to 2016. Counties with larger bubbles received more FDI than counties with smaller bubbles. Across all Indiana counties, there is a great alignment between the E-primed Index and FDI jobs—regions showing greater business growth potential had received more FDI. This is also true for the small/midsize counties.

Figure 6: The E-primed in relation to FDI

a) All counties

Map with FDI jobs shown as an overlay, ranging from 0 to 6,798: 14 counties = more than 103; 27 counties = 100.1 to 103; 21 counties = 99.1 to 100; 22 counties = 97 to 99; 8 counties = less than 97.

Source: Indiana Business Research Center

b) Small/midsize counties

Map with FDI jobs shown as an overlay, ranging from 0 to 2,076: 12 counties = more than 102; 13 counties = 100.1 to 102; 18 counties = 99.1 to 100; 24 counties = 94 to 99.

Source: Indiana Business Research Center

Conclusion

The E-primed Entrepreneurship Readiness Index provides an easy-to-compare assessment tool for practitioners. The comparison can be made cross-sectionally with regional peers to learn a region’s relative standing, and it can also be made over time to assess the improvement of the region itself once a time series is available. The hierarchical structure of indices also provides practitioners a useful diagnostic tool, so they can explore the various components and sub-components to see where the weaknesses are and what actions can be taken to address them.

In the case of Indiana, we have shown that counties with higher innovative capacity (packed with advanced knowledge and high-technology industries) are more prepared for business growth. A lesson taken from this? Maybe it is unnecessary for all lost jobs to come back to our regions … maybe what we need is to focus on “better” jobs—jobs that can nurture long-term business growth and strengthen regional innovative momentum. To prepare for such jobs and business opportunities, a better prepared workforce seems to be the key. We have also shown that adaptive counties are more ready for new business opportunities. Thus, losing jobs to automation is not something to be scared of. Rather, not realizing automation is replacing man labor is something to be worried about.


This article was prepared by Ping (Claire) Zheng at the Indiana Business Research Center using Federal funds awarded to the Trustees of Indiana University and as a sub-component under award number ED17HDQ3120040 from the U.S. Economic Development Administration, U.S. Department of Commerce. The statements, findings, conclusions, and recommendations are those of the author(s) and do not necessarily reflect the views of the Economic Development Administration or the U.S. Department of Commerce.

Notes

  1. Quotes from “Why America Is Still the Land of Opportunity” by Bedros Keuilian, www.entrepreneur.com/article/310208.
  2. Location quotient (LQ) is another population tool of quantifying concentration as compared to a benchmark case (in our example, the nation).
  3. The basic assumption of factor analysis is that, for a collection of observed variables, there are a (smaller) set of underlying variables—called latent factors—that can explain the interrelationship among those variables. For example, a business-major student takes a variety of courses on accounting, finance, marketing, business strategy, business policy, etc. All of these subjects are meant to prepare the student for two sets of skills: quantitative (analytics) and verbal (leadership), which would be considered the latent factors.
  4. The theoretical justification has two merits: 1) the demand creation effect from FDI—demand for local products and services; 2) the role of vertical (inter-industry) linkages between local and “foreign” firms for technology spillovers.

References

  • Engbers, T., Thompson, M., & Slaper, T. (2016). Theory and measurement in social capital research. Social Indicators Research, 132(2), 537-558.
  • Rentfrow, P. J., & Jokela, M. (2016). Geographical psychology: The spatial organization of psychological phenomena. Current Directions in Psychological Science, 25(6), 393-398.
  • Slaper, T. (2017). The long view: New data and methods in regional economic development. Indiana Business Review, 92(4). Retrieved from www.ibrc.indiana.edu/ibr/2017/outlook/longview.html
  • Slaper, T., Bianco, A., & Lenz, P. (2018). Digital vapor trails: Using website behavior to nowcast entrepreneurial activity. In 2nd International Conference on Advanced Research Methods and Analytics (CARMA 2018) (pp. 107-113). Valencia: Universitat Politècnica de València.
  • Slaper, T., Does, T., Egan, P., Ortuzar, G. & Strange, R. (2016). Driving regional innovation: The Innovation Index 2.0. Retrieved from www.statsamerica.org/ii2/reports/Driving-Regional-Innovation.pdf
  • Stangler, D., & Bell-Masterson, J. (2015). Measuring an Entrepreneurial Ecosystem. Kauffman Foundation Research Series on City, Metro and Regional Entrepreneurship. Retrieved from www.kauffman.org/what-we-do/research/2015/03/measuring-an-entrepreneurial-ecosystem
  • Tian, Z. (2013). Measuring agglomeration using the standardized location quotient with a bootstrap method. The Journal of Regional Analysis & Policy, 43(1):186–197.