Association for Public Policy Analysis & Management

In the absence of strong national action, the past 20 years have witnessed a great deal of innovation by state and local governments on climate and energy policies. However, practitioners, policy-makers, and other stakeholders have been poorly served by the traditional tools of policy analysis, which include models generally designed for national or multi-state geographical scales. These tools cannot account for the impacts from energy policies that vary within states and across municipalities. This paper presents an econometric energy-economy model aimed at addressing this gap. Using the state of Indiana as an illustration, we demonstrate how our approach contributes to policy analysis and may be adapted for other states and small-scale jurisdictions.

The Indiana Scalable Economic and Energy Model (IN-SEEM) is a data-intensive, dynamic, simultaneous econometric model that shows the relationship between the energy system and the economy at two geographic scales: the state level and for sub-state, multi-county regions. Using data from the BEA, BLS, Census, NOAA, and the EIA—along with an allocation model that estimates energy consumption at the sub-state level—the model describes the energy sector using prices and consumption of electricity, natural gas, and motor gasoline across residential, commercial, industrial, and transportation end-uses. Employment and earnings in ten economic sectors, as well as gross state and regional product, unemployment, and non-wage income, comprise the economic component of the model. The “scalable” aspect of IN-SEEM refers to the ability of researchers to adapt the definition of sub-state regions according to the needs of policy analysis; this iteration of the model presents ten sub-state regions defined by contiguous counties with shared economic characteristics. The relationships between energy and economic variables are captured by 300 unique equations that capture historic relationships within each region; thirty additional identities aggregate the regional endogenous variables to represent state-wide effects. The structure of the model allows estimates at the state level to feed back into the regional equations, an important feature given the significant relationship state-level activity plays in the economic performance of certain sectors in certain regions.

This paper mainly serves as a “proof-of-concept” of a sub-state energy and economic modeling, though we also illustrate its usefulness for policy analysis by simulating an increase in motor gasoline prices driven by a national policy of allowing crude oil exports. The policy analysis produces a set of results describing the effect of such a policy at the state and sub-state level on energy consumption and broader economic indicators. From this analysis, we can draw several lessons that inform policy design and anticipate both intended and unintended consequences.