Recent Research: Identifying peer states for technology-based economic development

While competition between states over business incentives and headquarters attraction is often derided, new research published in the Journal of Technology in Society suggests that competition in technology-based economic development is hardly a zero-sum game. In Persistent peers and the rhetoric of state economic competition, author David Schwarzkopf from Bentley University uses cluster analysis to track state progress across 53 TBED measures published by the NSF through its Science & Engineering Indicators series. Over a 12-year study period, Schwarzkopf finds that all 50 states improved on more than half of the variables used, with more than 60 percent of the states moving in the same direction on 80 percent of the measures. He argues that although states clearly compete, more focus is needed on how each state is making progress while also working to improve on their deficiencies.   

In the study, Schwarzkopf calculates the average annual change for each state across each of the 53 variables studied. Measures include data on higher education, workforce, financial R&D inputs, and others. Based on changes over time, the author constructs “self-organizing maps” to group states on how they cluster together at the beginning and end of the study period. The phrase “persistent peers” refers to states that are grouped by similarity at the beginning of a measurement period or states that remain similar through the end of that period. A “really persistent peer” group remains similar at both the beginning and the end of the period.

A map of the 12 “really persistent peer” clusters is below.

As a note, the order of these groups does not imply a difference across performance. Nor do groupings imply similar practices or policies enacted by states. Instead, the cluster analysis allows for states to have a more accurate picture of which other states rank similarly across a variety of TBED measures and over a period of time.

Schwarzkopf suggests there are two distinct advantages to knowing one’s peer states. First, it gives a state a better idea of who their peers are when they are evaluating their own progress on TBED measures. Second, it may provide clues about which state’s practices may be worth pursuing, adopting, or adapting.

There are several areas identified for supplementary research. First, more is needed to understand why non-geographically proximate states may cluster together. While it might make sense that New Hampshire and Vermont are peers, why is Minnesota also in that group? There is also an interest in what effect state initiatives have on TBED rankings, and whether some initiatives are more beneficial than others. Similarly, a better understanding of other TBED characteristics, like the number of universities or the presence of federal labs, would better inform these peer groups.

Using these same measures, a future study could look at peer groups across the subcategories within the TBED measures examined, such as higher education, R&D outputs, or workforce.  This would give states an even better understanding of similar states, and perhaps provide more insight on particular strengths and weaknesses.

Schwarzkopf concludes by noting the reality that economic development progress is frequently framed through the lens of competition. However, he suggests, technology-based economic development should not be perceived as a zero-sum game of chasing the leader. Instead, states should focus on where they are making progress, and where their deficiencies still exist.

A table showing the 12 “really persistent peer” groups can be found below.