The Value of Statistics for TBED: Part One

SSTI looks at Milken's State Technology & Science Index

During a presidential election year, perhaps more than any other time, the public is bombarded with statistics: voter surveys, polls, budget deficits, climate change, changes in employment or stock markets or trade, and others. It becomes difficult to determine which figures are true and which have been spun or manipulated for a particular purpose. The value of a statistic should be directly related to its quality. Unfortunately, how widely particular statistics are disseminated and repeated more often determines their acceptance or "truth."

Numerous statistics and rankings are often used in the marketing and promotional efforts of state and local economic development agencies and regional chambers of commerce. As a result, an observer has to look closely at what is being measured when multiple states or communities proudly claim to be first in something. They obviously are not all number one, unless, for instance, City A's claim is on an absolute basis, City B tops the list on a per capita basis, and City C has looked at the data as a percent change or relative to some other selected criterion.

More importantly, stats should play a prominent role in developing and refining public policy for tech-based economic development (TBED). Decision-makers should be interested in determining which programs and policies to implement, what issues need to be addressed, and how well existing TBED efforts are working. Reliable and valid statistical measures should be a part of the discussion.

Two recent papers provide quick case studies to illustrate the good and ugly side of applying statistics to TBED policy development. This week, SSTI examines the Milken Institute's 2004 State Technology & Science Index. Next week, we'll consider "Do Science Parks Generate Regional Economic Growth: An Empirical Analysis of their Effects on Job Growth and Venture Capital?," a working paper from the AEI-Brookings Institute for Regulatory Studies.

Milken's second index has received wide coverage in the general media and economic development community since its release a couple of weeks ago, and it should -- but not because of which particular states come out on top, in the middle or toward the bottom of any of the 75 indicators used in the index. With its new index, the nonprofit and nonpartisan Milken Institute dares to ask if these particular indicators are useful for developing a reliable and predictive model for benchmarking tech-based economic growth on a state level.

The 75 indicators are grouped into four input categories or composites - Research & Development Inputs; Risk Capital and Infrastructure; Human Capital Investment; and, Technology and Science Workforce - and one outcome composite called Technology Concentration and Dynamism. If the model is of any value for TBED policy development, the report contends, then the outcome should be "predictable" or explained based on the four input composites.

Results of a regression analysis and other econometrics used by the Milken Institute found 84 percent of variation in the Technology Concentration and Dynamism composite across states could be explained by movement in the other four composites. In addition, the Index also was "able to explain more than 75 percent of the variations in per capita income of the working age population across the states."

These figures are impressive and suggest the model can provide a useful tool, over time and with further refinement, to help states to assess which elements or measures should be the highest priority. For example, looking at the time series properties of the Index, Milken concludes, "the results indicate that for states with the lowest per capita income, the highest rates of return to investment appear to be in the improvements in broad measures of human capital, such as percent of adult population with a bachelor's degree or greater... At the other end of the scale, there is a long time lag between improvements in the R&D infrastructure and technology outcomes, but they are highly significant. Enhancing the entrepreneurial skills and venture capital networks of a state will garner enormous returns."

The 2004 State Technology and Science Index: Enduring Lessons for the Intangible Economy is available on the Milken Institute website at: http://www.milkeninstitute.org/publications/publications.taf?cat=ResRep&function=detail&ID=304