Interpretation: the final spatial frontier

Abstract

The use of spatial econometric models in political science has steadily risen in recent years. However, the interpretation of these models has generally ignored the important substantive, and even spatial, nature of the estimated effects. This leaves many papers with a (non-spatial) interpretation of coefficients on the covariates and a brief discussion of the sign and strength of the spatial parameter. We introduce a general approach to interpreting spatial models and provide several avenues for an exposition of substantive spatial effects. Our approach can be generalized to most models in the spatial econometric taxonomy. Building on the example of the diffusion of democracy, we elucidate how our approach can be applied to modern political science problems.

Voting in a global pandemic: Assessing dueling influences of Covid‐19 on turnout

“…To visualize the spatial effect, we follow Whitten, Williams and Wimpy's (2021) suggestion to use a map to simulate how a shock in the explanatory variable in a given city affects the outcome in neighboring municipalities s. Our simulation increases the number of deaths per 1000 inhabitants in the month leading up to the election in the city of Bauru in the state of São Paulo from 0.01 to 2.…”

How transnational party alliances influence national parties' policies

“…Since our model includes a temporally lagged variable (Lagged RILE Party Position), the coefficient estimates of the spatial lags only reflect short-term effects in a current year (Whitten et al , 2019). Therefore, we estimate asymptotic long-term effects for our spatial lag variables by considering the coefficient of our temporally lagged dependent variable (Plümper et al , 2005; Plümper and Neumayer, 2010).…”

Should we talk about the weather? How party competition and coalition participation influence parties’ attention to economic issues

“…the “focal party”) along the left–right dimension affects all the other parties, first, by incentivizing an action by the contiguous parties 3 (first-order neighbors) and then a subsequent action by the neighbors of the contiguous parties (second-order neighbors) as they find the strategy that maximizes their vote share. All of the parties’ responses following the initial shift by the focal party are known as indirect effects and can be divided into both local effects —those shifts that occur among first-order neighbors—and global effects —those shifts that occur among higher order neighbors, including the originator of the shift (known as feedback effects ) (Whitten et al, forthcoming). Unless the model is specified in a manner that connects the parties across time and space (see below), the OLS model eliminates this crucial stage of strategy adjustment from formal models of issue competition.…”

“… 13. As Whitten et al (forthcoming) illustrate, directly comparing the coefficients from the ordinary least squares (OLS) and spatial Durbin model (SDM) is not an appropriate way to compare effect sizes. Unlike OLS models (but similar to other models estimated via maximum likelihood, such as logit and probit), the substantive effects in an SDM are not easy to discern from the coefficients.…”