Indiana evidence on the employment beta: a simple tool for evaluating the desirability of targeted sectors.

• Author: Dufrene, Uric

Beta is a term widely used and accepted in the world of investments and corporate finance to measure volatility. Portfolio managers use beta to select securities of a desired risk level and financial analysts to estimate a project's required rate of return and the cost of capital. Portfolio theory suggests that an equity asset with a beta of more than 1 exhibits more volatility than the market, and an equity asset with a beta of less than 1 exhibits less volatility than the market.

The concept of beta can be applied to targeted industry analysis. (1) Berry and Blackwell propose that economic developers consider the employment beta of the industry when conducting targeted industry selection. The authors indicate that industries with higher betas exhibit greater swings in employment than industries with lower betas. Therefore, high beta industries can contribute to greater social upheaval due to the greater sensitivity of its employment to overall changes in national employment. Berry and Blackwell suggest that, other things equal, economic developers should prefer high growth/low beta industries to high growth/high beta industries. From a social costs standpoint, employment in low beta industries is not as sensitive to changes in national employment and thus provides more certainty in local employment levels.

This article does not recommend that a region shun or discourage the growth of any particular industry, as those decisions are best made locally. However, given scarce resources among communities, the employment beta concept is one way to focus economic development efforts.

As a starting point for economic developers across Indiana, employment growth and employment betas for supersectors for Indiana metro regions are provided (see Figure 1).

[FIGURE 1 OMITTED]

Beta estimates require monthly data and those are more readily available at the supersector level. (2) Obviously, it is possible for an industry within the sector to exhibit a different growth/beta pattern from the overall sector. Therefore growth/beta estimates provide initial information that might motivate a closer analysis of a specific industry targeted.

Interpreting Employment Betas

A beta of 1 implies that the respective supersector moves in parallel fashion with national employment. So, for example, if the nation's employment increases by 1 percent, then one can expect employment in that particular supersector to increase by 1 percent on average.

A beta higher than 1 implies that employment in the supersector is more volatile than changes in employment at the national level. Suppose a supersector has a beta equal to 4. A 1 percent change in employment at the national level would produce a corresponding 4 percent change in employment in the respective supersector. Hence, supersectors with high betas are more sensitive to changes in national employment. High beta supersectors produce more jobs when national employment is on the upswing. When national employment is on the decline, however, these supersectors lose more jobs.

A beta of less than 1 implies that employment in the supersector is less volatile than changes in employment at the national level. A supersector with a beta of 0.5, for example, indicates that for a 1 percent change in employment at the national level, employment in the supersector will change by only 0.5 percent. Low beta supersectors have less sensitivity to changes in the overall economy. Therefore, when national employment is increasing, employment in the supersector increases at smaller rates. Conversely, when national employment is decreasing, local employment declines are smaller than national changes.

The Calculations

Monthly non-seasonally adjusted employment data from the Bureau of Labor Statistics are used to calculate employment growth and employment betas.

Growth: Average annual growth measures for metro total nonfarm payrolls and metro supersectors are calculated using average fourth quarter employment figures. The average annual rate of growth is calculated using the annual compound growth rate by using fourth quarter average employment in 1990 as the beginning value and fourth quarter average employment in 2006 as the ending value.

Betas: Monthly employment data are used to calculate employment betas. In this context, the monthly percentage change in employment is analogous to the monthly percentage return in a typical equity security calculation.

In a typical security beta calculation, ordinary least squares (OLS) is used to regress security returns against returns in a market portfolio. In calculating employment betas, the "market portfolio" is national total nonfarm payrolls. Therefore, to calculate an employment beta, one can use OLS and regress monthly changes in employment at the local level against monthly changes in employment at the national level. This is done for total nonfarm payrolls in each metro area and for each supersector within each metro area. The coefficient on the independent variable, percentage change in monthly national total nonfarm payrolls, represents the employment beta. (3)

[FIGURE 1 OMITTED]

Beta Stability: The finance literature has debated the issue of beta stability. Is the beta stable for a security over different business cycles? The evidence suggests that the individual stock betas do change over time and portfolio betas exhibit more stability over time than individual securities. (4) A supersector is, in essence, a portfolio of the individual industries; so the use of supersector data provides economic developers with betas that are potentially more...