Arkansas Tech University The Culture Wars & Political Polarization in Perspective

Through the 1970’s there was a fairly flat trend in partisan polarization. It was largest in 1976, in the first post-Watergate presidential election (1.46), but for the most part hovered near a group polarization score of 1.0. In the 1980’s we saw partisan polarization spike in 1982, but for the most part the partisan polarization was only slightly higher in the 1980’s relative to the 1970’s. This all changes in 1988, as partisan polarization on ideology established a new and higher plateau (around 1.5). This would continue until 1992. Between 1992 and 1994, proximate to Clinton’s failed attempt to reform health care and leading into the Republican Revolution, partisan polarization on ideology begins to significantly and substantially increase more than it had at any other point of time since the 1970’s. By 1994, partisan polarization is approaching 3, doubling the partisan polarization on ideology that we witness in the 1970’s. The first decade of the 21^st century and the Bush administration is no exception to this trend. Indeed, while there was a small decline in 1998, partisan polarization spiked even more in the 2000’s: peaking near a partisan group polarization score of 4 in 2004. This score is also the maximum polarization score for the time series.

Figure 9.2 shows the relative contributions of Republicans, Democrats, and Independents to the group polarization score. Recall, these are weighted scores, so independents have a small and declining contribution to partisan polarization given their size relative to the parties and the fact that independent identifiers have declined since the 1970’s. That said, there is a slight increase in the contribution of independents to polarization despite its decreasing weight, likely due to the increasing distance of Republicans and Democrats from Independents rather than movement on ideology for the Independent identifiers themselves. Interestingly, Democrats have higher group polarization scores than Republicans for most of the series. The three data points where the Democrats spike above the Republicans demonstrate the face validity of the measure of partisan polarization. In 1982, in response to Reagan’s election, Democrats spike from a polarization score of 0.6 to a score just short of 1.3. Democrats again surge ahead of the Republicans on polarization in 1994, likely in response to the Republican Revolution. And finally, Democrats substantially outpace Republicans in polarization in 2004, coincident to the prosecution of the Iraq war.

The models for the Republican and Democrats proportionately reduce the error over the mean by over 70 percent in the case of the Democrats and nearly 70 percent for the Republicans. Clearly this is strong evidence that the partisan identifiers in the mass public have polarized. Partisan polarization on ideology has increased by 0.082 per for every survey year since 1972, Survey year (I.e. the progression of time) explains 74.5% of the variation in partisan polarization. The trend model for Independent identifiers performs poorly in comparison, explaining less than half of the variation in partisan polarization over the time series. And again, most of that is likely attributable to the Republicans moving to the Right and Democrats moving to the Left.

While the Republican and Democratic percent contributions are clearly affected by their relative group sizes in the weighted measure, it is not nearly as dramatic as the effect on Independent contributions. For the most part there is ordinal consistency between the weighted and unweighted contributions for the identifiers with the two political parties, but there are differences, suggesting the size of the group has an influence on contribution. For example, the largest contribution of the Republican identifiers to weighted partisan polarization on ideology is in 1988 (50.68%), but it is just the second largest contributor to unweighted group polarization (46.56%). The implication is that not only were Republican identifiers more distant from the other party identifiers in 1988, but there were more Republican identifiers in 1988 resulting in a greater contribution to the weighted group polarization measure.

If we look at the weighted group polarization in Table 9.3, we see that there are rather significant oscillations in percent contribution between the Republican and Democratic Parties. For example, in 1982 Republicans account for only 32.84% of the group polarization score, which is 9.66% below the mean contribution for Republicans in the time series. But in 1984, in the next survey year, that contribution climbs to 49.91%, which is 7.41% above the mean contribution to partisan polarization. There is a similar trend with the Democrats. In 1982, the Democrats account for 62.33% of partisan polarization on ideology, coming in at 10.16% above the mean contribution. In 1984, that drops to 44.71%, below the mean contribution. While relative size plays some role in this oscillation, the oscillations are still apparent in the unweighted group polarization contribution table. If you look at Figure 9.3, you can see the up and down nature of the percent contributions for Republican and Democratic identifiers in the unweighted group polarization score. However, there does appear to be some patterns in the contributions. Democrats were well below the mean through most of the 1970’s, and they were well above the mean for most of the 2000’s. Conversely, Republicans were big