Heuristics and policy responsiveness: a research agenda

Abstract

Theories of policy responsiveness assume that political decision-makers can rationally interpret information about voters’ likely reactions, but can we be sure of this? Political decision-makers face considerable time and information constraints, which are the optimal conditions for displaying decision-making biases—deviations from comprehensive rationality. Recent research has shown that when evaluating policies, political decision-makers display biases related to heuristics—cognitive rules of thumb that facilitate judgments and decision-making—when evaluating policies. It is thus likely that they also rely on heuristics in other situations, such as when forming judgments of voters’ likely reactions. But what types of heuristics do political decision-makers use in such judgments, and do these heuristics contribute to misjudgements of voters’ reactions? Existing research does not answer these crucial questions. To address this lacuna, we first present illustrative evidence of how biases related to heuristics contributed to misjudgements about voters’ reactions in two policy decisions by UK governments. Then, we use this evidence to develop a research agenda that aims to further our understanding of when political decision-makers rely on heuristics and the effects thereof. Such an agenda will contribute to the literature on policy responsiveness.

Introduction

In representative democracies, it is important that political decision-makers, such as politicians or party leaders, form accurate judgments about voters’ likely reactions. Firstly, because if elected officials desire re-election (Strøm 1990), then it is in their interests to avoid misjudgements about how voters will react.1 Secondly, it is important because scholars have found a link between voters’ perceptions of policy responsiveness and their support for democratic institutions (Ezrow and Xezonakis 2011; Keman 2014). Therefore, it is critical for citizens’ confidence in democratic institutions that political decision-makers can accurately gauge whether voters will feel well-represented by their decisions.

On occasion, political decision-makers will make misjudgements about voters’ reactions. In this paper, we present illustrative empirical evidence of how biases associated with heuristics—cognitive rules of thumb that may facilitate judgments and decision-making (see Gilovich et al., 2002)—contributed to misjudgements about voters’ reactions in two policy decisions by United Kingdom (UK) governments that proved electorally costly. Our data come from interviews with those involved in the decision-making process. Following our presentation of these findings, we set out an agenda for future research, particularly the need to investigate how the salience of an issue affects decision-makers’ likelihood of drawing on heuristics and whether the UK’s centralisation of powers makes the occurrence of biases associated with heuristics more likely than in less centralised systems of government. Additionally, our findings raise questions around the role of the media in affecting decision-makers’ perceptions and why decision-makers assume that voters will judge them on policy outcomes.

Making accurate judgments about voters’ responses is hard. Political decision-makers must consider multiple factors influencing such reactions, such as (a) the policy decision itself, (b) the framing of the political choices (Chong and Druckman 2007; Slothuus and De Vreese 2010) and (c) the policy outcomes (Achen and Bartels 2016; Green and Jennings 2017). These factors may influence voters’ reactions in different directions. Tetlock (2005) demonstrates that even so-called expert political judges who make their livelihood through political judgments (e.g. by advising on political trends) are not good at making accurate judgments. In fact, more experience could even impair judgment. Further, it is widely established that political decision-makers face considerable time constraints and have to deal with an overload of information (Jones and Baumgartner 2005; Walgrave and Dejaeghere, 2017), which affect their ability to make rational judgments (Jones 2001). Voters’ increasing expectations of politicians, the demands of social media and a 24-h news cycle severely impact decision-makers’ ability to take reflective, informed decisions (Flinders et al. 2020). Behavioural research on judgment and decision-making shows that people typically respond to such constraints by relying on heuristics (e.g. Gilovich et al. 2002). It is an empirical question of whether political elites—who typically operate in an environment in which they have access to information and who may have expertise, factors that may reduce reliance on heuristics—respond this way as well. We argue that it is plausible that they do. Access to information can easily lead to information overload (Baumgartner and Jones 2015), which reduces a decision-makers’ ability to make comprehensively rational decisions.

Little is yet known about whether and how decision-making biases related to the use of heuristics influence political decision-makers’ judgments of voters’ likely reactions. An exception is Miler (2009), who showed that Congressional staffers in the United States appear to rely on the accessibility heuristic when judging which groups were likely to find an issue important. Whilst Miler’s study adds knowledge about how legislative staff form judgments about salience of issues, it does not examine how they form judgments about how voters in general will react to an issue (which is our focus). Several studies have found that politicians are poor at estimating levels of public support for policy positions (e.g. Broockman and Skovron 2018; Pereira 2021). Judgments of voters’ likely reactions typically need to be made either under time and information constraints (e.g., Walgrave and Dejaeghere 2017) or in a context of information overload (Baumgartner and Jones 2015); exactly the conditions under which the use of heuristics is likely. For example, in the early stages of the COVID-19 pandemic when faced with conflicting information signals and considerable time pressures, the Conservative government in the UK used a cognitive shortcut to form an impression that “freedom-loving” Brits would not tolerate a COVID lockdown despite later evidence that British people were highly compliant with, and highly supportive of, lockdown measures (Earle 2020). We thus hold that political decision-makers are likely to rely on heuristics when forming such judgments and expect that this leads to biases affecting decision-making.

With this focus, our study also contributes to the field of political psychology (see e.g. Huddy et al. 2013); a field in which the study of misjudgements goes at least back to Jervis’ (1976) seminal study on misperceptions in international politics. While there is much literature that focuses on the heuristics that voters use (e.g. Bang Petersen 2015; Bartels 1996; Redlawsk 2004), the literature on the use of heuristics by political decision-makers is still relatively scant (see Vis 2019a, b for an overview and discussion). In this paper, we address this gap by focusing on decisions taken by UK governments, who, unlike the legislative staff that are the focus of Miler’s (2009) work, have considerable agency to set the political agenda. In such cases, it is particularly rational for decision-makers to accurately anticipate voters’ reactions as they are likely to be held accountable by voters for these decisions.

Existing work on heuristics and decision-making biases

Research on heuristics goes back to Simon’s early work (Simon 1955) and has branched out in two directions (see Kelman 2011 for an overview). The first is the heuristics & biases tradition (H&B) pioneered by Kahneman and Tversky (e.g., Tversky and Kahneman 1974; see Gilovich et al. 2002). When using a judgmental heuristic, “(a)n individual assesses a specified target attribute of a judgment object by substituting another property of that object—the heuristic attribute—which comes more readily to mind” (Kahneman and Frederick 2002: 53, emphases in original). Oftentimes, this means that the judgment is based on incomplete or biased information. According to H&B, while facilitating judgment and decision-making, heuristics can also result in decision-making biases, defined as deviations from comprehensive rationality (Gilovich et al. 2002).

The second tradition to heuristics is fast & frugal (F&F), pioneered by Gigerenzer and colleagues (e.g. Gigerenzer and Selten 2001; Gigerenzer 2015). According to F&F, when knowledge, time and computational power is limited—i.e. when rationality is bounded (Simon 1990)—heuristics often result in better decisions than a comprehensively rational approach would. F&F state that heuristics are ecologically rational (Todd et al. 2012) by fitting the specific environment. Since we investigate two cases of electoral misjudgements, we take H&B as our theoretical starting point.

H&B focuses on two so-called general purpose heuristics (Kahneman and Frederick, 2002): availability (Tversky and Kahneman 1973; Schwarz and Vaughn 2002) and representativeness (Kahneman and Frederick 2002; Stolwijk 2020; Stolwijk and Vis 2021; Tversky and Kahneman 1974) which subsume many other heuristics.2 In general, people use the availability heuristic when assessing how likely it is that something occurs by focusing on the ease with which they can think of instances or occurrences of it. They use the representativeness heuristic when answering difficult questions of probability by replacing them with simpler questions of resemblance (Kahneman and Tversky 1972: 431). The availability and representativeness heuristics often facilitate judgment and decision-making but can also lead to decision-making biases. The predictable biases related to the availability heuristic result from the retrievability of examples, with “available events” coming more readily to mind (Tversky and Kahneman 1974).

Working in the H&B tradition, researchers are increasingly finding evidence that political decision-makers display decision-making biases related to heuristics (see Vis 2019). Most work to date comes from the field of foreign policy (e.g., Jervis 1976, 1986; see McDermott 2001), but there are also some more recent studies from the field of comparative politics. For example, Stolwijk and Vis (2021) demonstrated by means of a survey experiment that Dutch local politicians displayed evidence of two biases related to the representativeness heuristic (Kahneman and Tversky 1972): the conjunction error (i.e., assuming an actor has an attribute based on having a characteristic) and scope neglect (i.e., people’s tendency to neglect the scope of a representative event). Moreover, Weyland (2007) showed in his analysis of social security reform in Latin America that hard-pressed policymakers used the availability heuristic (Tversky and Kahneman 1973) by adopting a highly visible policy reform, even though that reform may be unsuitable for their context. The decision-makers also displayed biases related to the representativeness heuristic by drawing “excessively firm conclusions from small samples and short time frames (…)” (Weyland 2008: 292). Böhmelt et al.’s (2016) study on party policy diffusion also finds that parties draw excessively firm conclusions from small samples.3

Case study summaries

Our cases are two UK policy decisions that were salient (hence providing an incentive for political decision-makers to “get their judgments right”), and that ultimately proved electorally costly. Our first case (“immigration”) is the decision by the New Labour government not to introduce transitional controls on migration from the ‘A8’ nations joining the EU in 2004. The decision to encourage immigration furthered two of the New Labour government’s wider goals: (1) increasing economic growth whilst avoiding inflationary pressures (Wright 2010) and (2) carrying favour with Eastern European governments who might support the UK’s resistance to the EU gaining more policy competencies from national governments (Consterdine 2018).

The A8-decision contributed to a steady rise in immigration to the UK during Labour’s tenure in office (which lasted until 2010). This policy outcome damaged Labour’s reputation, particularly following the financial crash in 2008. Analysis of individual data in the British Election Study showed that negative perceptions of Labour’s performance on immigration were a major cause of Labour’s overall decline in support between 1997 and 2010 (Evans and Mellon 2016).

Our second case (“tuition fees”) concerns the Liberal Democrats’ support for the trebling of tuition fees in 2010 as part of a coalition government with the Conservatives—despite over a decade of campaigning to abolish tuition fees altogether and a recent history of electoral success in university constituencies linked to precisely this policy position. Polls at the time showed that half of Liberal Democrat voters were reconsidering their support following the decision (Helm and Asthana 2010). Whilst it is hard to untangle the electoral effects of the tuition fees decision from other decisions made by the party in coalition, the U-turn remained a long-standing symbol of how the party abandoned its principles for power (Butler 2021).

Both cases are examples of decisions on salient topics, where the incentive was high for political decision-makers to form accurate judgments of voters’ possible electoral reactions to their decisions. Yet, in both cases, this apparently failed, since the decisions proved electorally costly. By focusing on these cases, we purposefully select on the dependent variable. Selecting on the dependent variable is often seen as problematic (see e.g., Geddes 1990). While we agree that selection of the dependent variable is indeed typically problematic in quantitative studies, it is sometimes precisely what you need in qualitative research. For one, qualitative researchers are interested in explaining specific outcomes (Goertz and Mahoney 2012). This means that some cases from the possible universe of cases (here: all policy decisions by UK governments) are more important than others (our two cases are such examples). It is especially important to know if a specific condition—here, decision-making biases related to heuristics—is present in such important cases. By focusing of these two cases, we also set the bar high for ourselves, since because of their salience, the decision-makers had a strong incentive to “get public opinion right”, pushing them to develop comprehensively rational judgments. Below, we discuss how we examine whether the decision-makers’ misjudgements emanated from decision-making biases related to heuristics.

Materials and methods

Our main data are transcripts from 15 interviews conducted by one of the authors between 2018 and 2020 with elite actors who observed the decision-making process in the two cases; see Appendix I for an overview of their roles. Interview subjects were either Ministers in the relevant departments, such Ministers’ advisers,4 or advisers to the party leader or other MPs with a leadership role in the party. For example, in the tuition fees case, interviews were conducted with MPs involved in the coalition negotiations at which the party chose not to prioritise its opposition to tuition fees. The interviewees were approached as part of a wider project that explores how different factors such as the perceived policy benefits, or the media and pressure groups, influenced the decision-making process. Interviewees were selected due to their known role in the decision-making process, and further approaches were made on the recommendations of initial interviews (i.e., through snowballing). Those interviewed were largely no longer in positions of leadership in their respective parties; only one of the six politicians interviewed was still an MP at the time of the interview, and only one of the nine advisers was still working for the party at the time of the interviews. This reduced the incentive for interviewees to provide inaccurate information regarding the decision-making process. In all, 26 individuals were approached, representing a 58% response rate.

The interviews were designed to gather data for a process-tracing exercise (Rohlfing, 2012) to understand why the decisions were taken and how perceptions about different factors, including the electoral reaction, affected different points of the decision-making process, as well as why alternative policy decisions were ruled out. The initial data collection adopted an interactive strategy (Tashakkori and Teddlie 1998), adapting the questions asked in response to data gathered in earlier interviews. For example, if an early interview identified that a particular event was critical to decision-makers ruling out one course of action, future participants would be asked about their recollections of that event. Further, interviews had a different focus depending on participants’ roles. For example, in the immigration case one interview was conducted with a former Prime Ministerial adviser on public opinion research, while two others were conducted with former Prime Ministerial advisers on policy formation; the interviews therefore focused more on matters directly related to interviewees’ former roles. The questions asked were thus unique to each participant, although the interviews were conducted in a semi-structured manner and followed similar themes. The full template of the interview script is available in Appendix II, along with examples of the questions used. We use the sections of the interviews that focus on anticipations of electoral reactions to assess whether political decision-makers displayed the biases related to heuristics in forming these judgments.

To identify whether political decision-makers display decision-making biases related to the use of heuristics, we have used insights from existing qualitative approaches, especially Weyland (2007), who used qualitative sources to assess the use of heuristics in policy-making. We decided against searching the interview transcripts for incidences of particular words or phrases to identify decision-making biases. Instead, we first examined the judgments relating to electoral reactions made by political actors in our cases. Next, we detailed the decision-making biases that can be identified from these instances. Specifically, the researcher most familiar with the context of the cases examined the data for instances where decision-makers recalled assuming an electoral reaction based on a judgment unrelated to rigorously conducted public opinion research (i.e., the “rational” approach). The researcher then looked for instances of biases associated with the availability or representativeness heuristics. For instances of availability bias, the researcher looked for instances where perceptions of electoral reactions were based on the most easily accessible information rather than the information that would be most appropriate, e.g., research into public opinion. For instances of representativeness bias, the researcher looked for instances where perceptions of electoral reactions were based on the decision in question being similar in one aspect to an area on which public opinion was known, but where there were significant differences between the decision and the comparison being used. The identified instances were then verified by the other researcher. The interview data were supplemented by an analysis of primary and secondary source material including Parliamentary debates, contemporary newspaper reports, policy documents, scholarly and journalistic accounts of the decision-making processes, and decision-makers’ memoirs; Appendix III lists these sources. This expanded the material consulted but uncovered few new revelations about decision-makers’ (potential) misjudgements, perhaps since unlike the interviews, such materials were not designed to highlight decision-makers’ assessments of voters’ likely electoral reactions.

Identifying the use of biases relating to heuristics in real-life settings is difficult. By relying on transcripts from interviews, we are limited by the assumption that what politicians say in interviews is an accurate reflection of their thinking. However, in another respect our research design is a “hard case” in which to find evidence of biases related to heuristics. The cases were purposefully selected because they caused electoral damage. This means that the likelihood of interviewees falsely mentioning decision-making biases to enhance their reputation is low, given that any heuristics used to anticipate electoral reactions broadly failed. The decision-making biases are examples recalled by interviewees about their own perceptions of how voters would respond, rather than assertions about others’ perceptions. Given that participants were commenting on their own failures of judgments, we are confident that the data from the interviews reveals broadly accurate information about participants’ judgment formation.

Results

Following the approach set out above, we identified 12 instances of misjudgements by political decision-makers about the likely electoral consequences of their decisions. In some instances, these misjudgements were identified across multiple interviews within a single case. Here, we present evidence of three misjudgements from each case that were critical to the outcomes and explore whether these misjudgements can be attributed to H&B heuristics. The remaining instances of misjudgements are included in the overview in Appendix IV, alongside excerpts from the interviews or the source material from which they were taken.

In the immigration case, New Labour combined a punitive approach to asylum applications that responded to public concern—e.g. making it much more difficult for unsuccessful applicants to appeal their cases—with a strategy of encouraging legal economic migration to grow the economy (Consterdine 2018). However, it was not clear that voters distinguished between asylum and immigration issues (Mattinson 2010). A bias related to the availability heuristic contributed to this misperception. Media coverage and public discussion of immigration issues during New Labour’s government concentrated heavily on matters of asylum. In the run up to the A8 decision, the tabloid media set out a fear of benefit migration rather than of economic migration. Policy solutions were sought to respond to both the asylum and benefit migration issues. Conversely, the lack of immediately visible information on the potential electoral implications of encouraging economic immigration led to a lack of responsiveness on this issue. These perceptions were recalled by all interviewees in this case, as per the following excerpts from the interviews:

(…) most of the discussion was not about what became kind of Polish plumbers in popular discourse, it was about Roma gypsies and whether Roma communities would come and claim benefits by virtue of being entitled to do so as European citizens. So that was the issue which became politically salient in 2004. It wasn’t the issue of economic migration per se from Eastern Europe. (Interview with Special Adviser)

We all felt that we were in the middle of a maelstrom led by the right-wing press in relation to asylum. And that coloured everything. It meant that in a sense the focus wasn’t on legal migration because there wasn’t the space. (Interview with Politician)

An example of a bias related to the availability heuristic is visible in the tuition fees case. When undertaking negotiations about a coalition government, Liberal Democrat decision-makers focused on the immediately available information about which issues were of potential electoral significance. As revealed in the excerpt from the interview, what was perceived as critical to the party’s reputation was the need to form a government quickly. The electoral repercussions that might occur if the party enabled an increase in tuition fees by the Parliamentary party abstaining on any vote were not so immediately visible, and therefore were not the focus of deliberation.

Naively and foolishly all of us didn’t give as much attention to the implications of the exact wording of the coalition agreement on tuition fees as we should have done: because it wasn’t an immediate issue and it would come back later in the year but there were lots of things which we were under pressure to resolve now. The pressure of public opinion, or the expected pressure of public opinion, from the day of the election until the Wednesday when the coalition deal was finally done was to get an agreement quickly to show that coalition government wasn’t a recipe for delay and confusion; not risking our economic position in the European markets, not being Greece. So all the pressure then wasn’t on specific policy, it was to see whether we were up to getting a deal. (Interview with politician)

One key misjudgement in the tuition fee case concerns the perception that public anger over the U-turn would dissipate. When forming their judgment as to the longevity of voters’ anger, decision-makers drew on what they considered the representative example of the New Labour government performing a similar U-turn in 2003 but managing to go on to win another general election in 2005. This can be observed in the following excerpt from the interviews.

I think what people didn’t understand enough was how long-term that was; I think partly because of what Blair had done on tuition fees where he had basically said that he wouldn’t increase them and then he had, I think there was some sense that “Well this will all be very hard in the short-term, but we’ll weather through it”. (Interview with Special Adviser)

Whilst Liberal Democrat decision-makers were right that the New Labour tuition fee example was representative in terms of it being an example of a party in government reneging on commitments over tuition fees, there are clear differences that Liberal Democrat decision-makers overlooked. Students, recent graduates and academics were a much greater proportion of the Liberal Democrats’ 2010 electoral coalition than of Labour’s 2001 electoral coalition (Fieldhouse et al. 2006). The tuition fees U-turn also compounded a perception that the Liberal Democrats did not stand up for their principles, a perception formed by the party’s decision to enter coalition with the Conservatives despite being traditionally perceived as a left-of-centre party (Dommett 2013).

While initially searching only for biases related to the representativeness and availability heuristics, when analysing the empirical material, we also found instances of another decision-making bias: narrow bracketing. Narrow bracketing is a variant of the phenomenon choice bracketing, a term coined by Read et al. (1999), and also qualifies as a deviation from comprehensive rational choice:

Choice bracketing “designates the grouping of individual choices together into sets. A set of choices are bracketed together when they are made by taking into account the effect of each choice on all other choices in the set, but not on choices outside of the set. When the sets are small, containing one or very few choices, we say that bracketing is narrow, while when the sets are large, we say that it is broad. Broad bracketing allows people to consider all the hedonic consequences of their actions, and hence promotes utility maximization. Narrow bracketing, on the other hand, is like fighting a war one battle at a time with no overall guiding strategy, and it can have similar consequences”.

An example of narrow bracketing can be found in the immigration case. Since some voters expressed their dislike of immigrants based on immigrants being from different cultures and speaking different languages, some New Labour decision-makers advocated promoting British citizenship as a way of encouraging migrant communities to demonstrate their commitment to the UK. A broad-bracketed approach might have considered that policy solutions needed to consider voters’ economic, as well as cultural, concerns. In the words of a politician who had advocated for the citizenship solution:

So there’s a two-way street, people can see that there’s a commitment, they get that people have put themselves out to learn more and to speak better, and therefore we can ask people to provide a warmer welcome. It didn’t wholly work unfortunately but that was the message. And underpinning it was citizenship. We’d rather people seek naturalisation and go through that route and be citizens than we would to just see them as temporary workers because they would have a commitment to community and communities would perhaps understand that they had that commitment. (Interview with politician)

We do not claim that the biases related to heuristics detailed above solely explain these misjudgements. In both cases, the anticipated electoral reactions to the decisions considered other aspects of public opinion (see Wright 2010; Butler 2021 for more detail), for example Liberal Democrat decision-makers’ judgment that voters would reward the party for successful stewardship of the economy and that this would supersede voters’ disappointment of the tuition fees U-turn. However, the data we present here shows evidence that biases related to heuristics can affect political decision-makers’ judgments about how voters will react. These biases have previously been overlooked in the literature on policy responsiveness.

Toward a research agenda for studying the effect of heuristics in policy responsiveness

What do the immigration and tuition fee cases teach us about the conditions under which political decision-makers rely on heuristics and how it influences their ability to form accurate judgments about voters’ reactions?

Our findings reveal several patterns concerning how decision-makers calculate electoral incentives in real-life scenarios. These raise important questions for scholars of policy responsiveness and for understanding how different actors — e.g., the media — can influence the policy-making process. The prevalence of decision-making biases associated with the availability heuristic—such as the one presented in the main text about the immigration case — reveals that signals about the likely salience of issues, particularly their media coverage, can disproportionately affect the decision-making process. This contributes to our understanding of the media’s effect on policy-making (Van Aelst et al. 2014), especially regarding how individual political decision-makers react to media coverage; which is still a rather unexplored terrain (for an exception, see Helfer and Van Aelst 2020). It would be a relevant avenue for future research to examine if political decision-makers also display the biases related to the availability heuristic when making judgments about voters’ likely electoral reactions on low salience issues. For such cases, decision-makers’ incentives to “get it right” and follow a comprehensively rational approach is lower, because of which it is likely that they will display decision-making biases when making such judgments. However, low salience also means that the issue probably receives less media coverage, which lessens the opportunities for decision-makers to rely on the media to form their judgments. An experimental approach could be useful to examine the occurrence of decision-making biases related to the availability heuristic for high vs. low salience cases.

Furthermore, the biases related to the representativeness heuristics that we identified — such as the one presented relating to the tuition fee case — show that political decision-makers regularly make analogies that turn out badly in terms of anticipating voters’ reactions. Analogies have been a central topic in International Relations for decades, especially in case studies (e.g., Khong 1992). Our findings on biases related to the representativeness heuristic call for more systematic empirical research on the role of analogies in political decision-making more broadly, and policy responsiveness in particular. A relevant question could again be whether this differs between issues of high salience such as our two cases here and issues of low salience. In line with research in International Relations, such future work could use a case study approach. To enable insights into whether and how political institutions shape the use of analogies and their effects, comparisons of high vs. low salience issues both within a country and between countries would be a fruitful comparative case study approach.

A key condition affecting the likelihood of decision-makers relying on heuristics concerns the time pressures they face. Interviewees in the immigration case presented a portrayal of a government constantly on the back foot with the immigration issue; e.g., responding to media pressure rather than undertaking deliberative exercises to develop initiatives to seize control of this policy agenda. In the tuition fees case, decision-makers felt rushed into a decision to provide reassurance to universities about their future funding (see also Clegg 2016). This confirms our expectations that the time and information pressures faced by political decision-makers contribute to displaying biases related to heuristics. Such pressures are particularly acute for British political actors. The centralisation of powers in the British system has long been identified as a potential cause of poor governance as Ministers struggle with their workloads (Dunleavy 1995). Future comparative research should consider whether political actors in less centralised systems are less likely to display the biases related to heuristics when forming judgments of voters’ likely reactions. One route would be to follow the qualitative, inductive approach we have presented in a less centralised system, such as Germany or the United States.

Furthermore, our analysis reveals that decision-makers tend to anticipate that voters would rationally respond to policy outcomes. This intriguing finding warrants further qualitative investigation, for instance by means of case studies. Although some studies have found that voters reward governments for delivering competent policy outcomes (Green and Jennings 2017), others have found that voters are unaware of policy outcomes such as the actual performance of the economy (Achen and Bartels 2016). We speculate that since decision-makers may consider more information about a decision’s potential policy outcomes than about its potential electoral consequences, this may lead them to overlook that voters will be unlikely to consider the same outcome-focused information when forming their judgment as to the policy’s merit.

One policy area where comparative analysis may be particularly illuminating concerns governments’ responses to the COVID-19 pandemic. There are echoes of Weyland’s (2007) findings that policymakers are drawn to policy solutions made by nearby countries in the Australian and New Zealand governments’ decisions to adopt ‘zero Covid’ policies more in line with East Asian, than European countries. Further, governments have had to make timely and significant policy decisions, often when faced with conflicting signals about the policy and electoral implications, particularly towards the start of the pandemic. Analysis of the use of the biases related to heuristics may reveal why, for example, countries such as Sweden at the start of the pandemic or the UK more recently, adopted different policies to their neighbours (Hale et al. 2021).

Overall, our findings indicate that, like “the rest of us” (Gilovich et al. 2002), political decision-makers are also liable to decision-making biases in ways that influence their responsiveness to voters. There is merit in future research exploring how prevalent the use of biases related to heuristics is when decision-makers are forming judgments of voters’ reactions.

Appendix I: Summary of Interviewees’ Roles

See to Table 1.

Table 1.

Summary of interviewees’ roles

	Case: Immigration	Case: Tuition Fees	Total
Role: Minister in relevant department	2	1	3
Role: Minister’s adviser	1	1	2
Role: Party leader’s adviser	3	4	7
Role: MP in other leadership role	0	3	3
Total	6	9	15

Appendix II: Questions used in Interviews

Template Interview Script

Section 1–Individual information

 Confirm the participant’s role at the time of the decision.

Section 2–General

What was the process behind how the party in government chose to pursue particular policies over others?

Section 3 – Decision-making process

What were the main objectives behind the party’s decision to….?

Did the process behind the decision differ to other decisions made in government?

Section 4 – Anticipated electoral reactions

Did the party ever discuss the likely electoral consequences of its decision?

Did the party ever test or model the effect of its decision on public opinion?

Did the party try to convince the public of the benefits of this decision/ its preferred policy?

Section 5–Reflections

Do you think the government could have made a different decision?

Did you ever see the party having to trade-off between its desired policies and the need to retain the support of the electorate?

Examples of questions asked to participants.

		Tuition fees case	Immigration case
Section 2: General	What was the process behind how the party in government chose to pursue particular policies over others?	What were the guiding principles behind the policies the Lib Dems pursued in government?	A lot has been written about how New Labour conducted a lot of public opinion research, focus groups etc. When you were in government as opposed to just at election time, was that information often fed back to senior Ministers?
Section 3: Decision-making process	What were the main objectives behind the party’s decision to….?	Was it that there was an intention that the party was likely to abstain on it, then Vince ends up at BIS and someone spots a banana skin and the party starts to strategise for it? Or was it always a kind of “We’re going to engage in this policy process and see what we can get out of it”?	Was that a genuine policy response or more a reaction to tabloid/ opposition attention?
	Did the process behind the decision differ to other decisions made in government?	Having gone through the tuition fees experience, did that then affect how the party approached other issues in government that were perceived to be electorally tricky?	During your time as a Home Affairs Policy adviser, how much attention was given to economic migration as a potential political issue separately from asylum?
Section 4: Anticipated electoral reactions	Did the party ever discuss the likely electoral consequences of its decisions?	Was there ever a strategic discussion, knowing that the party was going to take an electoral hit, of what kind of policy might offset it, might offset these voters or bring in other voters; or was it never done in that kind of calculus way?	When do you think New Labour began to see immigration as opposed to asylum becoming a politically salient issue?
	Did the party ever test or model the effect of its decision on public opinion?	Were you aware at any point either when the policy was trying to be changed in 2009 or in the run-up to the Browne report that the party ever sort of modelled or tested the potential response to it?	When you were discussing policies in that sense, were they routinely assessed against any public opinion research that the party had done?
	Did the party try to convince the public of the benefits of this decision/ its preferred policy?	Once the decision was made that we’re going to end up raising the cap, what attempt was there to try and shape public opinion, carry people?	Do you think there was much of a strategy to try and shift public opinion towards a more liberal approach on economic migration?
Section 5 – Reflections	Do you think the government could have made a different decision?	How did other alternatives such as delaying the decision, or trying to repackage the rise as a graduate tax or graduate contribution get ruled out?	Tony Blair wanted this system of counting ppl in and out. Was that scuppered by the difficulties of getting the policy to work?
	Did you ever see the party having to trade-off between its de-sired policies and the need to retain the support of the elec-torate?	Did you find that there were any factors that meant the party was more likely to follow pub-lic opinion over its de-sired policies at times in government?	How do you perceive that the government sought to balance the desire for an immigra-tion policy which re-sponded to the needs of employers against public concern about levels of immigration or at least the admin-istration of immigra-tion, asylum policy?

Appendix III: List of Supplementary Primary and Secondary Sources

Parliamentary debates

Tuition fees debate, Higher Education and Student Finance, Volume 516: debated on Tuesday 12 October 2010.5

Tuition fees debate,6 Higher Education Funding, Volume 517: debated on Wednesday 3 November 2010.

Tuition fees debate, 30 November 2010.7

Tuition fees debate, 9 December 2010.8

Secondary Sources

Immigration Case

Balls, E. (2016). Speaking Out. Arrow Books.

Blair, T. (2004). Speech to CBI 27^th April. Available from: https://www.theguardian.com/politics/2004/apr/27/immigrationpolicy.speeches, accessed 17 April 2020.

Blunkett, D, (2006), The Blunkett Tapes: My Life in the Bear Pit. Bloomsbury.

Consterdine, E, (2018), Labour’s Immigration Policy : The Making of the Migration State. Palgrave Macmillan.

Gould, P, (2011), The Unfinished Revolution: How New Labour changed British Politics for ever. Abacus.

Home Office (2002). Secure Borders, Safe Haven – Integration with Diversity in Modern Britain, Cm. 5387, https://www.gov.uk/government/publications/secure-borders-safe-haven-integration-with-diversity-in-modern-britain, accessed 15th April 2020

Mattinson, D, (2010). Talking To A Brick Wall: How New Labour stopped listening to the voter and why we need a new politics. Biteback Publishing.

Pollard, S. (2005). David Blunkett. Hodder and Stoughton.

Straw, J. (2012) Last Man Standing. Pan Books.

Watt, N., & Wintour, P, (2015) How immigration came to haunt Labour: the inside story, The Guardian 24 March, https://www.theguardian.com/news/2015/mar/24/how-immigration-came-to-haunt-labour-inside-story, accessed 19 October 2019.

Wright, C,F, (2012), Policy Legacies, Visa Reform and the Resilience of Immigration Politics, West European Politics, 35(4), 726–755.

Tuition Fees Case

Bowers, C. (2011). Nick Clegg. The Biography. Biteback Publishing.

Clegg, N. (2016). Politics: Between the Extremes. Penguin Random House.

Gerard, J. (2011). The Clegg Coup: Britain’s First Coalition Government since Lloyd George. Gibson Square.

Laws, D. (2016). Coalition: The Inside Story of the Conservative-Liberal Democrat Coalition Government. Biteback Publishing.

Scott, P. (2014). The Coalition Government’s reform of higher education: policy formation and political process. In: C. Callender and P. Scott (Eds.) Browne and Beyond: Modernizing English Higher Education (pp. 32-56). Institute of Education Press.

Williams, S. (2010). The long route to fair funding of Higher Education, http://www.stephenwilliams.org.uk/the_long_route_to_fair_funding_of_higher_education, accessed 4 April 2018.

Appendix IV: Overview of Instances of Misjudgements by Political Decision-makers in the Immigration and Tuition Fee Cases

See to Table 2.

Table 2.

Overview of instances of misjudgements by political decision-makers about the likely electoral consequences of their decisions

Code	Case	Decision-making bias	Summary	Supporting quote(s) from interviews
TF1	Tuition Fees	Related to availability heuristic	The Liberal Democrats did not foresee the political consequences of negotiating an abstention on higher tuition fees since their attention was focused on more urgent matters.	Naively and foolishly all of us didn’t give as much attention to the implications of the exact wording of the coalition agreement on tuition fees as we should have done: because it wasn’t an immediate issue and it would come back later in the year but there were lots of things which we were under pressure to resolve now. The pressure of public opinion, or the expected pressure of public opinion, from the day of the election until the Wednesday when the coalition deal was finally done was to get an agreement quickly to show that coalition government wasn’t a recipe for delay and confusion; not risking our economic position in the European markets, not being Greece. So all the pressure then wasn’t on specific policy, it was to see whether we were up to getting a deal. (Interview with politician)
IMM1	Immigration	Related to availability heuristic	Media focus being on Roma potentially claiming benefits, rather than on ‘Polish plumbers’ taking jobs, influenced what decision-makers perceived as potential political problems.	Actually as it came towards the main 2004 accession date, most of the discussion was not about what became kind of Polish plumbers in popular discourse, it was about Roma gypsies and whether Roma communities would come and claim benefits by virtue of being entitled to do so as European citizens. So that was the issue which became politically salient in 2004. It wasn’t the issue of economic migration per se from Eastern Europe. (Interview with Special Adviser)
IMM2	Immigration	Related to availability heuristic	Immigration was not regarded as a potential political problem given the extensive media focus on asylum.	The real absolute focus from the Prime Minister downwards was on being seen to reduce the number of illegal asylum seekers coming into the UK. In a sense there wasn’t much of a focus on immigration as a problem, and therefore I think across government there was an explicit acknowledgement that immigration was important in terms of the needs of employers and we had to strike a balance, but it wasn’t regarded as the problem because the problem was asylum at that point. (Interview with politician)
TF2	Tuition Fees	Related to representativeness heuristic	Since the Liberal Democrats’ election campaign focused on four policy pledges and they performed well, there was a belief that by concentrating on those four policy pledges in office, they would be electorally rewarded. This did not account for the possibility that voters might still have associated the party more with tuition fees than with some of its 2010 election pledges.	There was a belief that what had worked in the campaign would carry through. (Interview with politician)
TF3	Tuition Fees	Related to representativeness heuristic	Tony Blair had performed a U-turn on increasing tuition fees in 2003, but still won a 3rd general election in 2005. This led to a belief that the Lib Dems could survive a similar U-turn.	I think what people didn’t understand enough was how long-term that was; I think partly because of what Blair had done on tuition fees where he had basically said that he wouldn’t increase them and then he had, I think there was some sense that “Well this will all be very hard in the short-term, but we’ll weather through it”. (Interview with Special Adviser)
TF4	Tuition Fees	Related to representativeness heuristic	The Browne report into Higher Education funding had strongly recommended against a Graduate Tax, therefore advocating such a position would be unpopular.	The Browne report had been very damning about graduate tax. It had just produced a report saying that graduate tax is a terrible idea mainly because of the money going to the Treasury rather than going to individual universities and because they believed strongly in a market principle that of course they should vary in cost and price so it was very difficult to say “we support a graduate tax” because Browne had just published a report saying graduate taxes are a terrible idea. (Interview with politician)
TF5	Tuition Fees	Related to representativeness heuristic	Since the new tuition fees system was more redistributive than the previous system for higher education funding, voters would support it.	It was a good scheme, it was better than the existing scheme in fact for anybody perceptive enough to work out the arithmetic. I was persuaded, I persuaded myself, that that would be good enough. (Interview with MP)
IMM3	Immigration	Narrow Bracketing	By focusing on addressing voters’ concerns about asylum, the issue of immigration will no longer be salient.	I think the substantive political mistake was to assume that if you could sort out the asylum system, if you could show the public that you had dealt with any kind of abuse of it and that it was working properly that the public would accept that and then they would accept economic migration. And in reality as soon as the asylum numbers did start to come down, focus just shifted straight to economic migration. (Interview with Special Adviser)
IMM4	Immigration	Narrow Bracketing	Encouraging migrants to become naturalised citizens can lead to existing citizens behaving in a more welcoming way.	So there’s a two-way street, people can see that there’s a commitment, they get that people have put themselves out to learn more and to speak better, and therefore we can ask people to provide a warmer welcome. It didn’t wholly work unfortunately but that was the message. And underpinning it was citizenship. We’d rather people seek naturalisation and go through that route and be citizens than we would to just see them as temporary workers because they would have a commitment to community and communities would perhaps understand that they had that commitment. (Interview with politician)
“A simple examination for citizenship applicants similar to that which exists in many other countries. This will strengthen the ability of new citizens to participate in society and to engage actively in our democracy. This will help people understand both their rights and their obligations as citizens of the UK, and strengthen the bonds of mutual understanding between people of diverse cultural backgrounds.”
Source: Secure Borders, Safe Haven. UK Home Office document from 2002 setting out managed migration strategy. This paper formed the basis for the 2002 Nationality, Asylum and Immigration Act.
TF6	Tuition Fees	Narrow Bracketing	The Liberal Democrats’ performance at the 2005 election was compromised by their unbelievable spending pledges. Following the financial crash, the party needed to appear fiscally credible.	The objective was to get votes, but by looking fiscally credible. It came back to 2005 where we had this 50p tax rate which was going to raise £3 billion and was going to fund 93 billion things so the shorthand “orange book strategy” view was that if we continued to promise lots of money when everybody knew there wasn’t lots of money we would look like idiots and not win any votes. (Interview with Special Adviser)
IMM5	Immigration	Related to representativeness heuristic	Gordon Brown advocating Riker’s theory of heresthetics – it is rational for a party to avoid issues it has a poor reputation on.	Brown believed very strongly that both immigration and law and order issues were issues that Labour could never win on. If you wanted to have a successful election campaign, you had to keep immigration off the agenda as far as possible (Interview with former Special Adviser from Consterdine, 2018:.140).
TF7	Tuition Fees	Related to representativeness heuristic	Tuition fees was not really the main reason why voters withdrew their support from the Liberal Democrats. They were other policies which they perceived as a ‘betrayal’ and would have led to the same outcome in terms of Liberal Democrat support.	Had there not been the U-turn on tuition fees, many of the voters who cited this as a reason or no longer voting Liberal Democrat would still have found some other ‘betrayal’ to point to, in order to justify withdrawing their support. (Laws, 2016: 50).

Funding

Chris Butler’s research has been funded by a grant from University of Manchester's Faculty of Humanities Internationalisation Fund.

Declarations

Conflicts of interest

The authors have no conflicts of interest to declare.