As the Great Recession recedes in our rearview mirrors, it is becoming ever more apparent that innovative activities are determining an increasing proportion of the long-term economic growth of cities, metropolitan areas, regions, states, and even nations. Look no further than the Milken Institute's "Best-Performing Cities 2015" report for evidence that 13 of the top 25 metropolitan areas were technology centers. Among the leaders were San Jose, Seattle, Denver, Austin, and Raleigh — all with important high-technology clusters. Or look to California, where, despite a reputation for high taxes and regulatory hurdles, technology and innovation have fueled stellar economic performance since 2010.
The State Technology and Science Index (STSI) endeavors to benchmark states on their science and technology capabilities and broader commercialization ecosystems that contribute to company growth, high-value-added job creation, and overall economic growth. We view the STSI as a measure of a state's innovation pipeline. The index isn't intended to be a measure of immediate economic impact, but rather to demonstrate that the return on science and technology assets will accrue in future years. Along with deep human capital, individuals who recognize entrepreneurial opportunity and have the knowledge and skills to develop it are among the strongest assets a geographical area can have in today's innovation-based economy.
The STSI's 107 individual indicators are sorted into five composites: Research and Development Inputs, Risk Capital and Entrepreneurial Infrastructure, Human Capital Investment, Technology and Science Workforce, and Technology Concentration and Dynamism.
The efficacy of the State Technology and Science Index in explaining variations in economic performance is strong and robust. We performed several regression analyses of the various relationships between the STSI and real high-tech GDP per capita and overall real GDP per capita of the working-age population. This adjusts for differences in population structure across states. States with higher birthrates, such as Utah, have high levels of young dependents.
The results demonstrate that STSI scores explain 75 percent of the variation in real high-tech GDP per capita of the working-age population across states. Figure ES1 graphically displays this relationship. It shows the actual 2014 values of real high-tech GDP per capita of the working-age population relative to those explained (predicted) by the association with the STSI. As shown in Figure ES1, Oregon and Washington have the largest unexplained variance. This is caused by the high concentration and large value of output per employee in the electronic components industry in Oregon (Intel) and in software employment in Washington (Microsoft). If we were to adjust for both, the overall explanatory power approaches 90 percent.
The findings show that for each 10 percent increase in the STSI score, real high-tech GDP per capita of the working-age population rises by 14.6 percent. An alternative interpretation is that at the mean STSI score, a 10-point increase translates into a 28.1 percent rise in real high-tech GDP of the working-age population. Additionally, a separate investigation found that 62 percent of the variation in real GDP per capita of the working-age population is reflected in movements in a state's STSI score. It's hard to imagine anything else with a higher return on investment.
The states with the weakest innovation assets and ecosystems for starting and growing innovative firms face a bleak future unless changes are made. West Virginia, Arkansas, Mississippi, Kentucky, and Louisiana make up the bottom five in this year's STSI. They are the least knowledge-intensive and their residents exhibit weak entrepreneurial skills. All of them have undertaken efforts to change their position in technology and science but have had limited success.
West Virginia had the biggest drop in STSI scores from 2002 to 2016, falling 4.3 points. Combined with a decline in coal production, the state's economic performance is in serious jeopardy. Although its decline was smaller, Kentucky, another coal state, faces similar challenges. Louisiana slipped as well. The good news is that Arkansas rose 5.2 points. Based on our statistical model, Arkansas' real high-tech GDP per capita of the working-age population is 6.3 percent higher than it would have been if its STSI scores had remained at 2002 levels. Mississippi has witnessed a small gain since 2002.
A renewed commitment to making investments in research, entrepreneurship, and human capital are necessary. We hope that the STSI aids in focusing the attention of these states and others on the need to invest and improve their current and future residents' economic fortunes.