Heuristics and Biases Part II: Brexit from a Behavioural Economics Perspective - Risk Aversion, Overconfidence Bias, Present-Bias and David Cameron

by HK Lim (hklim [at] thetroublewitheconomics.com)

Following on from Part I of this series on Brexit from a behavioural economics perspective(where we discussed the role of attribute substitution in framing the Brexit question), we now shift our focus to the most important person in this entire episode, David Cameron, the former leader of the Conservatives and Prime Minister. To understand David Cameron's role in this episode(particularly the claims that he was largely responsible for precipitating Brexit), and ultimately see how a clash of mainstream economic orthodoxy and behavioural biases(and the failure to take into account such biases) could have influenced key decisions in the run up to Brexit, we first need to examine the Conservatives' complicated relationship with Euroscepticism. In doing so, we will see how the political and business establishment operating using one system of reasoning(economic rationality based off rational choice theory) came up against a portion of the electorate that did not necessarily subscribe to that same system's arguments in support of Britain remaining in the European Union.

In short, by examining why David Cameron called a referendum on Britain's membership in the EU even though he was effectively opposed to Brexit we will better understand the system of reasoning that resulted in the Brexit referendum and how a rejection of such a system resulted in a vote to leave. Confused? Well, the history of the Tories' relationship with the EU has been anything but simple.

Tory Euroscepticism in a nutshell

Historically, the Tories have been the more pro-European of the two major British parties and generally stronger advocates of closer European unity as a platform for securing greater British influence in European politics. There was however, a marked about turn in this party position in the 1980s under Margaret Thatcher, who in her 1988 Bruges speech strongly criticised this vision of a united Europe as tantamount to the re-imposition of the frontiers of the state at a European level, "with an European super-state exercising a new dominance from Brussels." This Eurosceptic shift did not sit well with many in the Conservatives and proved instrumental in bringing about both Thatcher's departure from Downing Street and the demise of her political career. These very same internal party tensions regarding Britain's relationship with Europe continued during the time of the succeeding Tory leader and subsequent Prime Minister, John Major, who while having brought Britain into the EU in 1993, was constantly dogged by Euro-feuds within the party which ultimately rendered his premiership ineffectual in articulating its own agenda.

More Euroscepticism this way comes

As the EU continued to grow both in political and economic might, and against the backdrop of sharply increased EU migration into the United Kingdom, Tory Eurosceptics were now able to markedly shift the party's overall tone to one that was increasingly unwelcoming to the very notion of greater EU integration. It was thus an internal party atmosphere that was somewhat hostile to further European integration that Cameron faced when he ran for the Tory leadership in 2005. As part of his platform pledges in running for leadership of the party, Cameron promised to withdraw the Tories from the European People's Party, the Europe-wide organization for centre-right parties, in a gesture that implicitly signalled his commitment to stand with the Eurosceptics if he were to be elected into office. Cameron thus adopted a straightforward modus operandi, in dealing with his party's hostility towards the EU, he chose the strategy of placating the Eurosceptics within the party rather than choosing to lock horns with them. 

Eurosceptic appeasement as a manifestation of risk aversion
 
Cameron's strategy of Eurosceptic appeasement can be understood as a manifestation of risk aversion. Risk aversion is a particular preference characteristic, where an individual is reluctant to engage in a bet with an uncertain payoff(technically there is a chance of either a gain or a loss) and instead prefers a smaller certain payoff. The typical investment related example of risk aversion relates to an investor preferring to place his money in a bank with a small but guaranteed fixed rate of return over the alternative of investing in a stock with a higher possible expected return but also the chance of losses in the initial investment.

In the context of Cameron's available menu of decisions as leader of the Conservatives, this risk aversion corresponded to his reluctance to bet on a confrontation with the Eurosceptics head on(with the possibility of either strengthening his control of the party or running the risk of losing control over the party),  instead preferring to placate the Eurosceptics and retaining the current level of control over the party but making no real progress towards resolving the party's ongoing tensions over Europe.

A quick aside on the classic principal-agent problem regarding leadership survival and ultimate party(and state) survival is appropriate here. While in light of the Brexit outcome, it might be argued that Cameron's strategy of appeasement of Eurosceptics could be construed as the result of his putting the survival of the self before the party and the state, we must not forget that we are also benefiting from hindsight bias(for hindsight is always 20/20). We should also not inadvertently fall into the trap of the "outcome bias", where there is a tendency to judge a decision by its eventual outcome as opposed to the quality of the decision at the time it was made. And at the time of the initial decisions, all that we could perhaps put a finger on is the inherent tensions introduced by the principal-agent problem that had already been initiated by the conflict between decisions driven by a desire for individual political survival against decisions in the best interests of the survival of the party and state..

[An additional aside: to the extent that Cameron was deemed to have an Eurosceptic bent, it could have been as much motivated by political calculation as much as deeply held personal political ideology. We can never be fully certain, for utterances to Norman Lamont like "Eurosceptic-but not as Eurosceptic as you are," and to Labour's Denis MacShane, "I'm much more Eurosceptic than you imagine," do not sufficiently pinpoint where exactly on the scale of Euroscepticism he stood, even less so if his "official" position changed with time, circumstances and audience.]

Cameron's premiership, intensified Euroscepticism and risk aversion part deux

Following the 2010 British elections, and having formed a coalition with the pro-Europe Liberal Democrats led by Nick Clegg, Cameron tried to steer the Tories away from continued divisive infighting over the party's Europe stance, urging members to stop "banging on about Europe," and instead refocus on more pressing issues of the time. However, amidst falling poll numbers, Cameron's position within the Conservatives continued to be undermined by the failure of his government's austerity policies(he had espoused a platform of fiscal conservatism) which effectively put Britain's recovery on the back burner, the Eurozone crisis that continued to fuel a vocal Tory Eurosceptics criticism of Europe, and pressures from an ascendant UK Independence Party(UKIP) that was also riding a wave of anti-EU and anti-immigration sentiment.

In 2012, Cameron mooted the idea of a referendum, not in its final Brexit form, but on renegotiating Britain's relationship with the EU, which according to sources, was largely a party management issue in response to "unbelievably Eurosceptic" backbenchers, rather than driven by true ideological conviction. So once again, a risk-averse Cameron postponed the possibility of going head to head with the Eurosceptics that he perceived to be potentially challenging his leadership, via another attempt at appeasement, this time by promising a referendum on renegotiating Britain's relationship with the EU(which in the final Brexit form, ultimately evolved into a referendum of whether Britain should stay in the EU). Granted he may not at the time imagined that the referendum would have taken its final form as a vote on Brexit, we nevertheless once again see the principal-agent problem at work.

The 2015 GE surprise, overconfidence bias and present-bias

In an attempt to secure party unity in the run up to the 2015 general election, David Cameron, committed formally to holding a Brexit referendum if the Tories were to win a full majority. At the time (January 2013) of this promise, this may have been perceived by Cameron to be party unity secured at no cost because the majority of political observers and pollsters felt that the Conservatives were most likely to lose, or in the best case scenario, win seats still shy of a majority and hence there would be no actual follow through on this promise. [For the interested, more fascinating discussions of what might have gone wrong in poll forecasts for the 2015 GE can be found here and here.] In the same manner as before, this decision to commit to a future promise at no cost is fully consistent with the risk-aversion tendencies identified earlier and not necessarily inconsistent with existing economic theory and the rational choice school.

The stage was thus set, and when the Conservatives won a shocking majority in the 2015 GE, Cameron was forced to follow through on his promise to hold a referendum on Brexit or face possible mutiny within his party from the Tory Eurosceptic wing and further loss of electoral support to UKIP. When Cameron announced the decision to hold a Brexit referendum, he positioned himself squarely in the corner of the Remain camp and pledged his commitment to oppose the UK departing from the EU.

And this was where things got really interesting. While previously risk averse with respect to confronting the Europe issue within his party, when forced to play an unwilling hand, Cameron appeared overconfident that winning the referendum would be easy with most of the British establishment and big business in the pro-Remain camp. This is a classic case of the overconfidence bias, where an individual's subjective confidence in their judgments is greater than the objective accuracy of those judgments. Notwithstanding the evident problems with polls themselves not getting things quite right(again, discussions of what might have gone wrong in poll forecasts for the 2015 GE can be found here and here), more evidence of this overconfidence bias could also be seen in Cameron's apparent inadequate consideration of the risks to his Remain campaign's credibility from his previous Eurosceptic stance within the Tories. It was perhaps not best-advised to think that one could convincingly campaign on behalf of the Remain camp suddenly when the track record of one's previous stance on Europe didn't align properly with that expected of a Europhile.

The economic rationale was that once voters were presented with the economic arguments against leaving the EU, individual economic self-interest(as dictated by mainstream economics) would prevail and engender a preservation of the status quo(with a vote to remain) over any concerns regarding the EU's encroaching of Britain's sovereignty. As Alex Massie noted, "They might not do so with any great measure of enthusiasm but a reluctant vote Remain was all Cameron, and his government, needed." As detailed in Part I of this Brexit series, we all know how that turned out. And as Charles Moore wrote rather prophetically, "The fear factor which may well win the referendum for Mr Cameron actually operates even more strongly on the elites than on the mass of the population."

Furthermore, polls in the run up to the referendum also seemed to give the Remain camp the edge which likely further reinforced the overconfidence bias at play(although the margin of Remain's edge fluctuated and even turned in favour of the Leave camp). And the disconnect from the ground probably was also due to an over-reliance on representative statistics and not nearly enough working and listening on the ground. As John Harris wrote in the Guardian with regards to polling, "understanding of the country at large has for too long been framed in percentages and leading questions: it is time people went into the country, and simply listened."

This overconfidence bias and a political establishment seemingly out of touch with the electorate was also underscored by the Remain campaign's constant reliance on parading their stable of "experts" to persuade voters of the soundness of their position, warning of the severe economic costs if Britain were to leave the EU. Michael Gove, who campaigned for Leave, noted, perhaps somewhat ironically, “I think people in this country have had enough of experts.”

Before we end this section, we should also note that Cameron's Eurosceptic appeasement strategy is also consistent with what behaviouralists term present-bias, where one tends to over-value immediate rewards at the expense of one's long-term intentions. As the next section elaborates, this often arises from the problemmatic way in which payoffs are discounted over time. Thus, Cameron's choice to appease Tory Eurosceptics by promising to a referendum can be interpreted as an over-valuation of the immediate reward of retaining party control at the expense of a (hopefully) once and for all settling of party divisions via an outright confrontation on the issue of Europe. Not surprising from a behavioural perspective but nonetheless clashes with what traditional economic reasoning would expect.

Present-bias, time inconsistency and a digression on conventional economic modelling of intertemporal choice

We now briefly discuss Cameron's Brexit decision tree in light of the problemmatic manner in which conventional economic theory models preferences and choices over time. A fundamental notion central to models of decision making at different points in time(that is, intertemporal choice) is discounting, whereby a payoff, say $100, is worth more today than it is sometime in the future(say a year from today). One implicit assumption in conventional models of intertemporal choice is that there is a constant exponential discount rate that applies as the time horizon to the future payoff lengthens. For example, suppose $100 received a year from today is worth $90 today, $100 received two years from today is worth $81 today, $100 received three years from today is worth $72.9 and so on(discount factor 0.9). This stability in this discount factor is crucial in ensuring that preferences are stable over time. That is to say, if you had to choose between two possible payoffs A or B to be received at some point(s) in the future, as time progresses, your preferences or choices should not change. That is what it means to have time-consistent preferences.

In contrast, if during early periods the discount factor falls rapidly but then falls more gradually over longer periods, this gives rise to what is termed hyperbolic discounting and present bias(technically, hyperbolic discounting is one way present-bias can arise). Individuals with present-bias(or employing hyperbolic discounting) have a tendency to make choices that are not consistent over time, that is the choices are inconsistent with the choices their future selves would have made. This is because future rewards are much more heavily discounted under hyperbolic discounting compared to that of conventional economic theory's exponential discounting. The bigger question here is how to real humans (not just the "homo economicus" of conventional economic models) actually apply a discount factor to future payoffs? That is something to think about, as I am certain the majority do not think in such a consistently mechanistic fashion.

We can also similarly analyse David Cameron's decision tree in the run up to the Brexit referendum from this perspective. For simplicity, let us focus on two dates, January 2013 when Cameron made the promise to the Tories that an EU referendum would be held if they won a majority in the 2015 GE, and February 2016 when the date of the Brexit referendum was formally announced. [Of course, the would be uncertainty in the actual referendum date and the actual referendum wording initially at the January 2013 point which will complicate matters.]

In January 2013, in appeasing the Eurosceptics within his party, Cameron's stance can be classified as leaning towards Euroscepticism(even if solely a result of political calculation). In February 2016, Cameron instead threw his weight behind the Remain campaign, which can be construed as a shift away from Euroscepticism to a pro-Europe stance(or a shift to benign Euroscepticism if you might). At face value, this might be construed as exhibiting time-inconsistency in preferences. But let us also note that it is not unreasonable to expect that the payoffs in choices(pro-EU or anti-EU) would have shifted with time, and consequently, political expediency would typically dictate a switch in position if necessary for survival, so the simple analysis should be considered with the usual caveats. [The astute reader would also by now have had a hint at how Cameron's strategy of Eurosceptic appeasement could be framed in the perspective of being present-biased.]

The objective of this brief discussion of time inconsistency is mainly for the purpose of illustrating the limitations of conventional economic notions of intertemporal choice. To ensure preferences over time are stable, models implicitly require a rather specific discounting factor(exponential). In actual situations, there is evidence that people tend to exhibit present-bias. While actions taken in the real world are not quite so surprising to observers, from the perspective of conventional economic theory they can sometimes be outright baffling.

Concluding thoughts - "don't hate the player, hate the game"

 And therein lies the paradox of David Cameron's failed political gamble. While Cameron's role in the series of events leading to Brexit resulted primarily from a combination of conventional risk aversion and the principal-agent problem, the final domino in the Brexit referendum toppled due to a combination of Cameron's overconfidence that Remain would eke out a win, combined with a disconnect from voter preferences and an incorrect understanding of electorate human self-interest. One cannot really get away from the conclusion that the principal-agent problem played a central role in getting us to this point where there was a clear conflict in interest between what might be good for the state and what might be necessary for a leader's political survival. As they say, "don't hate the player, hate the game." We really need to be re-examining our institutions, be they political, public service or business and re-design them so that they properly serve the stakeholders they should be serving. Recognising how easy it was(outcome and hindsight bias anyone?) to be critical of decisions made prior to the referendum, it was probably the case that with regards to the referendum(and particularly Cameron's political future), it was a damned if you do, damned if you don't type of situation.

And so this is how the cookie biscuit crumbles. David Cameron, as a product of the neoliberal establishment that espoused conventional economic rationality was ultimately taken down by a clash between the establishment view of how the majority should behave according to mainstream neoclassical economic rationality and the pushback from an electorate that(at least in part) chose to reject the current economic orthodoxy by choosing NOT to vote along the lines of pure economic self-interest. [In perhaps a slight tinge of irony, David Cameron was also instrumental in the founding of the UK Behavioural Insights Team(aka the Nudge Unit) that was originally placed within the Cabinet Office.]

Furthermore, Cameron's present-bias in appeasing Tory Eurosceptics and overconfidence bias in winning the Brexit referendum is consistent with the general perspective of how behavioral considerations play into our daily lives, not so much overriding conventional economic notions of rational choice completely but rather by modifying various aspects of our behaviour. Nevertheless, by omitting the human element in economics, one ignores the consequences at one's own peril, as David Cameron so unfortunately discovered.

In part III of this series, we will examine the other major player in this European tragedy, the Leave campaign. Stay tuned.