North Carolina’s Innovation Reconstruction Imperative

North carolina’s Innovation Reconstruction Imperative

Powered by SAS analytics and the minds of iei
and the north carolina board
of science, technology & innovation

More and more, innovation drives economic prosperity.  What factors or inputs are most important for achieving overall economic growth and well-being in an innovation-driven global economy? Recently, IEI collaborated with the North Carolina Department of Commerce’s Office of Science, Technology & Innovation (OSTI) to explore the question, using software and technical expertise provided by SAS.

We found that the following three key innovation-related variables are statistically significant in predicting per capita GDP, average annual pay, and per capita personal income:

  1. Post-secondary educational attainment
  2. Proportion of workers in high-tech industries
  3. Proportion of workers in science and engineering occupations across the economy

It is imperative, therefore, that North Carolina focus on boosting outcomes in these three areas to drive future economic gains.

North Carolina’s current innovation ranking is unexceptional.  According to OSTI’s Tracking Innovation: North Carolina Innovation Index 2013, which evaluates performance across 38 innovation-related indices, North Carolina ranks 24th among U.S. states.

We must do better, and our new study points the way. The following short video helps illustrate our findings.   (A more detailed discussion of our research approach appears below the viewer.)

Details on our research approach:

We wanted to know which indices examined in the Tracking Innovation report would have the largest impact on three target variables that reflect economic growth and well-being:

  1. Per Capita Gross Domestic Product (GDP)
  2. Average annual pay
  3. Per Capita personal income

To gain that insight, we built statistical models that help predict differences and changes in the above-mentioned target (or dependent) variables in response to differences and changes in several input (or independent) variables. Independent variables used in the analysis consisted of measures of educational attainment, research and development activities, science and engineering occupations, high-tech occupations, exports, venture capital, cost of living, manufacturing GDP, and indicators of entrepreneurial activity.

We were able to find reliable, comprehensive data for all 50 states for the years 2000–2011. Within this timeframe, some variables occasionally had missing values; in those cases, missing values were imputed statistically before building the predictive models. We binned each of the three target variables into quartiles. North Carolina ranked at the low end of the second quartile from the top (i.e., near the middle of all states) for all three target variables.

To understand how North Carolina could move from its current ranking into the top quartile, we identified the drivers for each of these quartiles in terms of the independent variables described above. We achieved that by building a predictive model for each of the three target variables. Specifically, we built a Decision Tree model using SAS® Visual Statistics to determine the drivers for the binned target variables (quartiles). Simply put, a decision tree is a set of rules that splits the data into groups; the best split is the one that best separates the data into groups, where a single class (quartile, in this case) predominates in each group.

We wanted to know which factors tracked in the Tracking Innovation report would have the largest impact on three target variables that reflect economic growth and well-being:

  1. Per Capita Gross Domestic Product (GDP)
  2. Average annual pay
  3. Per Capita personal income

Based on the Decision Tree model, we found the following three key variables to be statistically significant in predicting per capita GDP, average annual pay, and per capita personal income:

  1. Post-secondary educational attainment
  2. Proportion of workers in high-tech industries
  3. Proportion of workers in science and engineering occupations across the economy


RETURN TO 2015 FORUM next steps PAGE

 

Resources & Credits:

Visit IEI’s Emerging Issues Commons to check out North Carolina county-level data on:

Other essential resources:

Tracking Innovation: North Carolina Innovation Index (2013)
SAS Visual Analytics
SAS Visual Statistics