Methodology: 50-State Competitiveness Dashboard
Methodology
Our 50-State Competitiveness Dashboard provides insights into where Massachusetts stands in the national economic landscape through indicators of state competitiveness. These indicators are aggregated into six different indices: Fiscal Stability and Public Management, Talent and Workforce, Tax Environment, Growth and Innovation Climate, Cost of Doing Business, and Quality of Life. Each consists of a weighted combination of datasets. The Council updates the competitiveness indices based on the availability of new source data and produces reports bi-annually based on a computation of the most up-to-date data available. The methodology for computing the competitiveness indices is given below.
Data Pre-Processing
Missing Values: The metrics in the MATTERS system often have data available for different years. If a value for some metric is missing for a given year, the closest previous value is used in the computation of the index ranking. If there is no previous value, the closest possible value is used.
Normalization: The values of metrics in the system vary greatly. Some are percentages that only vary by a few tenths of a point, some are numbers in the millions representing populations, or Gross State Product. The values must be normalized so that large values do not dominate. We use a standard z-transform to scale and center the data. For each metric, we subtract the mean value across all states and divide by the standard deviation. This results in each metric having a mean of zero and standard deviation roughly equal to one, making the different metrics comparable.
Inverted Trends: Typically, when looking at trends, high values are considered better than low values. However, for some metrics, the opposite is true. For instance, a low unemployment rate is a more favorable policy outcome than a high rate. Negative coefficients correct for data with inverted trends.
MATTERS Index Computation
Each index is computed by combining a set of datasets (or metrics) {${m_1, m_2 …m_n}$} as a weighted sum. The coefficient ${c_{i}}$ for each dataset determines its weighting in the sum. The general formula for each index is:
The metrics used and their weights are given in the tables below.
| Metric | Weight |
| Fiscal Balance | 30% |
| State Reserve Fund Level | 20% |
| State Debt Level | 20% |
| Medicaid as a Share of States’ State-Funded Budgets | 10% |
| Unfunded Pension Liability | 20% |
| Metric | Weight |
| Science and Engineering Degrees as % of Higher Education Degrees Conferred | 20% |
| Technology Employment as % of Total Employment | 25% |
| Bachelor’s Degree Holders in Workforce | 20% |
| Relocation of College-Educated Adults | 15% |
| Job Openings and Labor Turnover Survey (JOLTS) | 20% |
| Metric | Weight |
| State and Local Tax Burden Per Capita in $ | 10% |
| Top State Corporate Income Tax Rate | 40% |
| State Personal Income Tax Rate | 30% |
| Property Taxes as a Share of Housing Value | 20% |
| Metric | Weight |
| Annual Gross State Product Growth Rate | 20% |
| National Institutes of Health (NIH) Awards by State | 20% |
| Patents Awarded per 1,000 Individuals in Science and Engineering Occupations | 20% |
| Venture Capital Invested – Dollar Total and per Deal | 30% |
| Rate of New Employer Business Actualization | 10% |
| Metric | Weight |
| Average Family Health Insurance Premium | 25% |
| Median Earnings | 50% |
| Unemployment Insurance Premium per Employee | 10% |
| Retail Price of Electricity | 15% |
Quality of Life
| Metric | Weight |
| Education Quality | 30% |
| Housing Affordability | 45% |
| Average Commute Time | 10% |
| Healthcare Quality | 15% |