Regularization can improve statistical estimates made with highly correlated data. However, any regularization procedure embeds assumptions about the data generating process that can have counterintuitive consequences when those assumptions are untenable. We show that rather than simply shrinking estimates, regularization can reopen backdoor causal paths, inflating the estimates of some effects, and in the wrong circumstances, even reversing their direction. Recently, Cavari and Freedman (2023), argued that declining cooperation rates in surveys have inflated measures of mass polarization. We show that this finding is driven by large penalty terms in their regularized regressions, which leads to the estimates being confounded with time. Alternative methods do not show a clear positive or negative effect of declining cooperation on estimated levels of mass polarization.

The 2019 UK General Election had seismic consequences for British politics. After three years of political turmoil following the 2016 referendum on Britain’s membership of the European Union (EU), the 2019 election marked a victory for the Leave side of the Brexit debate, putting to rest questions of a second referendum and any chance of Parliament blocking the Withdrawal Bill. The United Kingdom left the EU on January 31, 2020. Although there were clear consequences for Britain’s EU membership, there is debate about whether 2019 was a “Brexit election” (Prosser 2020)—even a critical election (Green 2021)—or the continuation of long-term realignments in British politics (Cutts et al. 2020; Jennings and Stoker 2017). By most accounts, Brexit dominated the 2019 election as a political issue, but whether this represents a key moment in a process of realignment of voters in Britain remains to be seen.

For incumbents to be accountable for their issue stances, voters must sanction incumbents whose positions are “out of step” with their own. We test the electoral accountability of British legislators for their stance on Brexit. We find that there is very limited issue accountability. Individuals who disagreed with their representative’s stance on Brexit were 3 percentage points less likely to vote for them. The aggregate consequences of these individual effects are limited. A one-standard-deviation increase in the proportion of constituents agreeing with their incumbent’s Brexit stance is associated with an increase of 0.53 percentage points in incumbent vote share. These effects are one and a half times larger when the main challenger has a different Brexit stance to the incumbent. A follow-up survey of Members of Parliament (MPs) shows that MPs’ estimates of the effects of congruence are similar in magnitude. Our findings suggest that issue accountability is conditional in nature and limited in magnitude even for an issue such as Brexit, which should be maximally amenable to such effects.

Polls have had a number of high-profile misses in recent elections. We review the current polling environment, the performance of polls in a historical context, the mechanisms of polling error, and the causes of several recent misses in Britain and the US. Contrary to conventional wisdom, polling errors have been constant over time, although the level of error has always been substantially beyond that implied by stated margins of error. Generally, there is little evidence that voters lying about their vote intention (so-called ‘shy’ voters) is a substantial cause of polling error. Instead, polling errors have most commonly resulted from problems with representative samples and weighting, undecided voters breaking in one direction, and to a lesser extent late swings and turnout models. We conclude with a discussion of future directions for polling both in terms of fixing the problems identified and new approaches to understanding public opinion.

Google search data have several major advantages over traditional survey data. First, the high costs of running frequent surveys mean that most survey questions are only asked occasionally making comparisons over time difficult. By contrast, Google Trends provides information on search trends measured weekly. Second, there are many countries where surveys are only conducted sporadically, whereas Google search data are available anywhere in the world where sufficient numbers of people use its search engine. The Google Trends website allows researchers to download data for almost all countries at no cost and to download time series of any search term's popularity over time (provided enough people have searched for it). For these reasons, Google Trends is an attractive data source for social scientists.