December 12, 2017
Elections, referenda and politics have held quite a few surprises in the last 1 ½ years. While there is a general feeling among voters that things are moving in the wrong direction – although there are certainly differences of opinion about the direction itself – the outcomes of recent elections and polls have probably done very little to counter their disenchantment. One reason for this frustration is the increasing complexity of political issues. They just cannot be resolved using the simple answers offered by many populists. Voters crave such simple answers, however, as throughout evolution humans have been very successful in reducing complexity by applying heuristics, simple rules drawn from experience. [more]
Germany Monitor . Elections, referenda and politics have held quite a few surprises in the last 1 ½ years. While there is a general feeling among voters that things are moving in the wrong direction – although there are certainly differences of opinion about the direction itself – the outcomes of recent elections and polls have probably done very little to counter their disenchantment. One reason for this frustration is the increasing complexity of political issues. They just cannot be resolved using the simple answers offered by many populists. Voters crave such simple answers, however, as throughout evolution humans have been very successful in reducing complexity by applying heuristics, simple rules drawn from experience. Many of today’s issues such as AI or climate change are unique and there are few historic parallels to provide guidance. Still, human nature prefers to tackle such topics using the intuitive and associative System 1, rather than the rational but tedious System 2. Moreover, research shows that people – and voters – are subject to a massive “knowledge illusion”. They overestimate their own knowledge of subjects, even with regard to trivial things such as water closets or bicycles, let alone tax legislation, geo-politics or climate change. Research into the actual voting decision provides no comfort either , as it suggests that for many voters the first visual impression plays a very important role, as people subconsciously assess politicians’ competence based on their looks. Many of the key policy positions are ultimately based on value judgements, therefore a rational debate about these positions might be close to impossible. One should of course still discuss the costs and means of achieving them. But people tend to be not open to engaging in such a debate; when confronted with counter-arguments, they resort to motivated reasoning and dig in deeper. Social media provides the required support – quite often of pseudo-scientific substance – to back up almost any position. So should we abandon all hope? Probably not, but voters need to be better educated, especially with regard to statistics. In addition, while the credibility of experts has suffered – not least during the Brexit campaign – the example of the Dutch Bureau of Economic Policy Analysis shows that there is a role for a neutral, bipartisan evaluation of policy proposals. Finally, despite all these shortcomings we are convinced that democracy is still the far superior form of government. Despite all alternative facts and cognition biases people will ultimately figure out if a government is not delivering and then have the opportunity to send it packing. Author Stefan Schneider +49 69 910-31790 Editor Dieter Bräuninger Deutsche Bank AG Deutsche Bank Research Frankfurt am Main Germany E-mail: Fax: +49 69 910-31877 DB Research Management Stefan Schneider * The author would like to thank Franziska Winkler for her contribution to this article. December 12, 2017 Vox populi, vox dei or maybe not? Vox populi, vox dei or maybe not? 2 | December 12, 2017 Germany Monitor Elections and polls during the last 1 ½ years – not to mention political developments after the elections themselves – have held some real surprises in store, especially for forecasters who tried to predict election outcomes sometimes even unconsciously on the basis of an economic cost/benefit analysis. This raises the interesting question of how voters decide who (or what) to vote for. The economists’ model of a homo economicus or, in this case, “voter economicus”, is probably flawed for at least three reasons. First, we know from behavioural research that people are not merely egoistic, microeconomic welfare maximisers, but that they are also driven by altruistic motivations 1 . Second, what you vote for is usually not what you get, so party programmes or candidates’ campaign promises might – even when discounted – not provide a proper basis for the rational expectations hypothesis. In particular, the increasing number of parties entering national parliaments has heightened the risk that a voter’s preferred party ends up in a coalition. As a result, this can force compromise on election promises which might have been key to a voter’s decision. In some cases, the differences perceived by the parties themselves are so large that it might be extremely difficult to find a sufficiently large common denominator to form such coalitions at all. Finally, economies and societies have become so complex that even well-informed voters might have to accept that the measures they voted for will not yield the effects they ultimately desired. In this article, we explore how voters deal with issues of complexity in their decision-making process, how they select the information to base their decisions on and how they process this information – individually and in groups. That it is rather questionable whether voter behaviour is conducive to optimal policy outcomes likely comes as no real surprise. We therefore also look at measures which might improve the decision-making process – a process that ultimately determines to a large extent our future as individuals, societies and perhaps the fate of homo sapiens altogether. Some of the rather unexpected results of recent elections and referendums suggest that voters are so dissatisfied with the status quo in their country that they are longing for real change – whatever that might be. There is usually broad agreement that things are going terribly wrong. However, it seems that these frustrated voters are often in for disappointment. Greek Prime Minister Tsipras, who promised a fresh start with regard to the crisis in Greece, ultimately had no other option than to adhere to the Troika’s demands and implement exactly the kind of consolidation he condemned during the election campaign. After one year in office, there is a sense that President Trump’s initial agenda has been bogged down by the Washington swamp he promised to drain. In Germany, the parties of the Grand Coalition saw their vote tallies dwindle by 14 percentage points in federal elections last September, despite the booming economy with close to full employment and very high levels of business and consumer sentiment. As exploratory talks for a so-called Jamaica coalition – consisting of CDU/CSU, the FDP and the Greens – collapsed and the SPD is about to make a volte-face with respect to its election-evening promise to not re- enter another Grand Coalition (GroKo), the probability of such a GroKo has risen strongly, although 56% of the SPD’s voters reject a GroKo 2.0 according to recent polls. Voters likely felt that, given the partly contradictory policy intentions, parties might settle for the status quo. Brexit arguably embodies the most dramatic break with the past, and it is becoming increasingly clear that implementation will be much more complicated than promised by “Brexiters”, making it at least likely that the economic implications will be much more severe than what most of the people who voted for it believed. 1 Schneider, Stefan. Homo economicus – or more like Homer Simpson? Deutsche Bank Research. Current Issues, 2010. Vox populi, vox dei or maybe not? 3 | December 12, 2017 Germany Monitor Complexity fuels the desire for simple answers, but also causes them to fail It seems voters do not realise that one of the major factors behind their disenchantment with the status quo and the widespread feeling of powerlessness – namely, complexity – is exactly the reason why any drastic change in policy is unlikely to take place. Over the last 70 years, voters’ demands for ever more services from the welfare state and the obedient response of politicians who provide them, quite often in anticipation of such demands, have resulted in extremely complex public institutions (both national and international) bogged down by myriad vested interests and interdependencies. In recent decades, the quantum leap in the division of labour driven by globalisation, digitisation and technological progress has increased the complexity of economies and societies, further undermining the cultural identity not just of those left behind economically. The complexity is likely most clearly reflected in the volume of laws and regulations this has brought about, with Brexit serving as an extreme example. More than 20,000 regulations and EU directives have to be dealt with. The UK will have to renegotiate more than 750 international arrangements. In total, there are currently close to 2,000 international agreements related to environmental law alone. But even on a national level, political complexity is daunting. The Affordable Care Act, commonly known as Obamacare, has about 11,000 pages. MiFID, which is a major piece of post-financial crisis legislation going into effect next year, already has more than 1.4 million paragraphs, with regulators still adding to it. There is probably no developed country without complaints about the impenetrable complexity of its national tax code – complexity that results in unwarranted jumps in the tax burden or has altogether unintended side effects which often thwart original policy intentions. Complexity is widespread not only in political and social spheres, but also in technology and production processes. An average car, for example, consists of more than 20,000 single parts, which are provided by approximately 50 different suppliers. The fact that there is no common definition or measure of complexity reveals just how complex complexity actually is. Herbert Simon tried to measure it by the two attributes “hierarchy” and “decomposability”, where the latter refers to the fact that there are stronger interactions within a subsystem itself than between different subsystems 2 . Crutchfield and Young propose measuring the complexity of a system by the minimum amount of information necessary to reproduce the past statistical behaviour of the system 3 . Based on some of the recent unexpected election outcomes, one has to assume that, at least in some countries, the complexity of the political system has moved off the scale. This also implies that at least some – although more likely many – of the cause-and- effect relationships (interactions) in the political system are no longer predictable or, what is worse, have become so unstable that the system is on the borderline of chaotic behaviour. Voters have hardly attributed the – in their view – underwhelming results of politics to the complexity of the political system, and instead have faulted a lack of ability or interest on the part of mainstream politicians. As a result, voters are more likely to give outsiders, whose apparently simple and straightforward “solutions” are nonetheless confronted by exactly the same problems, a chance. Of course, this is reinforced by the denial of most mainstream politicians that complexity has made it increasingly difficult for them to manage the socio-economic system, probably because they fear that such an acknowledgement would limit their (re-)election chances. 2 Simon, Herbert. The Architecture of Complexity. 1962. 3 Crutchfield, James P. and Karl Young. Inferring Statistical Complexity. 1989. 0 2 4 6 8 10 12 1955 1965 1975 1985 1995 2005 2015 Source: Number of words of the Federal Tax Code/Code of Federal Regulations 1 in million 0 20 40 60 an iPhone App a Space Shuttle Firefox Browser Android OS Windows 7 Facebook in million Number of lines in the code of: 2 Sources: NASA, Quora, Ohloh, Wired Vox populi, vox dei or maybe not? 4 | December 12, 2017 Germany Monitor Voters’ difficulties in dealing with complexity Humans unfortunately do not appreciate complexity. Quite the opposite; they try to avoid it wherever possible. Reducing complexity means less cognitive drain on our limited mental resources 4 . For many day-to-day problems, we have developed certain heuristics, which are mental shortcuts or rules of thumb that work remarkably well 5 . However, as these heuristics by definition cut out a substantial part of the complexity that humans face, they make people prone to systematic errors, especially in more complex situations. Individual perceptions: Where voters fail Deciding who to vote for is, of course, a very complex question. It requires assessing the candidates’ competence, their track records as well as looking at party election platforms. Given the number of candidates and parties – in the recent German federal election, 24 parties ran – this can be an exhausting and time-consuming exercise. Of course, an ideal, completely rational approach of this nature also requires an understanding of the relevant issues of the election campaign. If we consider subjects such as labour market or taxation policies, even experts’ views tend to be miles apart and offer only limited guidance to voters looking for the best policy path in order to benchmark candidates’ and parties’ proposals. Given the minuscule probability that one’s own individual vote will make a difference regarding the election outcome, a completely rational voter would see this as a waste of his or her valuable time and would be inclined not to go to all that effort. One theory that attempts to explain why people still vote is called the “voter’s illusion”, which argues that voters believe that if they vote, similar people favouring the same candidate will vote as well. Other explanations frame voting as a pro-social behaviour, which makes voters feel good. Yet another argument is that voting is seen merely as a habit once learned and not questioned later. The voting decision: Based on an unreflective first impression? However, voting itself might ultimately not be the only irrational exercise. Studies show that the decision about which candidate to vote for is probably not a very rational one either. A massive amount of information about candidates and parties is usually available, and the accuracy of this information can hardly be checked. This was already the case before the notions of “fake news” and “alternative facts” came to the voters’ attention. Furthermore, the information will likely be contradictory in part, thus providing an inconsistent picture and requiring substantial cognitive effort to make up one’s mind. He’s got the looks When people are asked about the single most important factor determining their voting decision, competence always comes first. Unfortunately, competence is very difficult to assess and certainly not the first recognisable feature when confronted with the candidate. The first impression voters have of candidates is usually visual. In an experiment, Princeton psychologists presented pairs of profile pictures of candidates who had actually competed against each other in US House and Senate elections (a fact not known by the participants). They then asked participants to judge the competence of the persons in the pictures 4 Although the brain accounts for only 2% to 3% of the body weight, it consumes 25% of the body’s energy when the body is at rest. 5 Gigerenzer, Gerd and Peter M. Todd. Simple Heuristics That Make Us Smart. 1999. Vox populi, vox dei or maybe not? 5 | December 12, 2017 Germany Monitor based solely on facial appearance. 6 This inference of competence predicted the actual outcome of the election races with a probability of 66.8% and 71.6%. Even limiting the viewing time to a mere second did not change the results. This suggests that voting decisions are strongly influenced by rather unreflective first impressions and are not the result of rational reflection and deliberation. Nobel laureate Daniel Kahneman explains the human decision-making process from a dual-system perspective. A rational, deliberate consideration falls into the realm of System 2, where reflective, structured reasoning takes place. However, this system is slow, highly energy-intensive and tires out quickly. System 1, by contrast, is an intuitive, fast and often unconscious way of thinking which requires little energy or attention. As System 1 is capable of making quick choices based on very little information, it is used when deep reasoning is not required for decision-making. At the same time, however, System 1 is prone to biases and systematic errors 7 . These include the focusing illusion, which means that people tend to undervalue other aspects when they focus on one single issue, and the halo effect. The halo effect is a way for our minds to reconcile contradictory or incompatible information. Quite often, the interpretation of new evidence is shaped by the first impression. The halo effect can be so strong that information presented later on is more or less neglected. One other quite common – but usually accidental – method to reduce complexity is the substitution of an attribute. When people are confronted with a hard-to-answer question, they – without even realising it – replace it with a similar, although simpler, question that they are able to answer. This leads them to falsely believe that they actually came up with an answer to their original question. One obvious but still rather common example is questioning climate change based on current weather conditions, in which the question “Is the Earth’s temperature rising?” is replaced with “Is today’s temperature high?” It is hard to keep emotions out of politics The question arises as to why System 2, which is supposedly more deliberate and rational, does not make an appeal to avoid such biases, for example if the competence initially inferred from facial appearance is strongly contradicted by later information. The answer is that people tend to be lazy and System 2 tires easily, which is to say, considerable deliberate effort is required to overcome an initial impression. Thus, being aware of how you process information can already help to slow things down and start a deeper analysis by activating System 2. Unfortunately, a state of emotional arousal almost certainly sends System 2 to the back burner, and many of the topics discussed in politics involve strong emotions, such as fear or perceptions of fairness 8 . The aforementioned halo effect is related to the concept of internal anchors. Experiments have shown that these anchors can be set almost arbitrarily, and even unconscious exposure immediately before an experiment can impact them. But these internal anchors can also be strongly influenced by a person’s general beliefs, world view and political ideology. This is especially true with regard to abstract topics that people have little direct personal experience with, exposure to or limited opportunities to learn more about. When ideology comes into play, emotions are not far away. This makes it very difficult for people to approach the subject through rule-based analytical processing that relies on probability calculus and formal logic. 6 Todorov, Alexander et al. Inferences of Competence from Faces Predict Election Outcomes. Science, June 2005. 7 Kahneman, Daniel. Thinking, Fast and Slow. 2011. 8 Ariely, Dan. Predictably Irrational. 2009 System 1 (left) and System 2 (right) 3 fast unconcious automatic everyday decisions error prone slow conscious effortful complex decisions reliable Vox populi, vox dei or maybe not? 6 | December 12, 2017 Germany Monitor Emotions are also the ultimate recourse for politicians who lack rational arguments but seek to drum up support for their proposals. Populists are not the only ones to play to emotions such as fear, anger and revenge; mainstream politicians also quite often appeal to national consciousness or a sense of justice, which can sometimes border on greed. Voters’ ignorance regarding time and non-linearities Things become even worse when people find themselves in situations where they have to cope with long time horizons. People are much better prepared to deal with spatial dimensions, as can be seen in the use of charts where we convert the temporal dimension into a spatial dimension by depicting time on the abscissa. Research has shown that people tend to discount long-term consequences to such an extent that they hardly matter for today’s decision- making. This might explain, among other reasons, why most people, for example, do not consider climate change as a serious problem. In general, people have severe difficulty with the fourth dimension: time. Their perceptions are largely from the present, where the confirmation bias holds great sway, or the very recent past, where the availability bias is also at play. With respect to the more distant past, the brain often plays tricks by filling in growing memory gaps with seemingly compatible, though imagined, pieces. The whole time, the conscious mind is completely unaware that this is happening. On the other hand, time might sometimes be the missing key to finally understanding something that we have experienced in the past. 9 The rather long lag between the voting decision and policy results, which are actually influenced by all sorts of things voters could not even imagine when casting their votes, make it very difficult for voters to learn from past mistakes. In addition, the strong emotional aspects of the voting decision likely also stand in the way of a critical ex post evaluation. Looking ahead to the future, we are similarly lost, as our brains are poorly designed to internalise the concept of time and to imagine the future in general. But even if confined to a small subsystem, we are very quickly at a loss when exponential growth comes into play. We all know the brainteaser about the lilies on a lake doubling their number of leaves within a week, or the anecdote about the inventor of the chess game who asked for just one kernel of rice on the first field and to have the number of kernels on each following field doubled as an actually not-so-humble reward. This kind of extrapolative growth cannot be intuitively grasped. But even more rational approaches using statistical methods quite often fail when trying to predict even more or less linear developments over a longer time horizon. This is due to the fact that we base our reasoning about the future on the current structure, which makes thinking outside the box so difficult. Most quantitative models are optimised by reproducing past data. A change in the underlying structure of the data-generating process therefore quite often makes such models dubious guides for the future. Knowledge illusion – how little we really know (without even knowing it!) Not only do people face severe cognitive limitations they are not really aware of, but they also live in an illusion of knowledge. This insight which, by itself, probably does not surprise any readers, has been examined in a new book by 9 See Dörner, Dietrich. Die Logik des Misslingens – Strategisches Denken in komplexen Situationen. 2007, for an in-depth discussion. Vox populi, vox dei or maybe not? 7 | December 12, 2017 Germany Monitor Sloman and Fernbach, two cognitive scientists in the US. 10 When humans reason, they look at causal relationships, mostly based on a tiny glimpse of the whole system. Unfortunately, this type of forward reasoning – from A follows B – seems so easy and natural that we tend to ignore alternative relationships, namely that something actually completely different might have caused B. We do not, however, make the same mistake when reasoning diagnostically (that is, working backward from an observation to the likely cause), as it becomes obvious that B could be the result of many other factors besides A. This ignorance is not limited to simple causalities. People are – again without even knowing it – very ignorant with regard to how things work. Sloman and Fernbach present a clever method to show this knowledge illusion: People are asked to rate their knowledge about a subject on a scale of 1 to 7. Then they are asked to describe/explain the subject, which is where most test subjects fail miserably. After this disillusioning experience, subjects have to rerate their knowledge and – you guessed it – marks are substantially lower. This sobering pattern has been observed with relatively simple subjects such as the functioning of a toilet. On subjects such as taxation, geopolitical intervention or global warming, the level of knowledge illusion is even more staggering. Unfortunately, there seems to be a strong negative correlation between the interest and effort to acquire information and a readiness to jump to strong conclusions 11 . In 2014, when tensions between Ukraine and Russia escalated, US scientists asked 2,066 registered 2014 US voters to click on Ukraine on an unlabelled map. They then asked the participants how strongly they support or oppose the claim that the US should take military action in Ukraine. Their analysis shows a significant positive relationship between a measure of how far they were off the mark locating Ukraine on the map and their support for military action 12 . In a similar experiment, US citizens were asked whether the US should intervene in North Korea. Once again, support for intervention was highest among those unable to find North Korea on the map. A similar gap opens when it comes to views about inequality and knowledge about actual income distribution. The assessment of income inequality is strongly correlated with the subjective perception of inequality which explains almost 65% of its variation. Unfortunately the perception has little to do with reality. The extent of inequality tends to be overestimated in most European countries including Germany 13 . Wisdom of the crowds? The power of group dynamics The problems due to recognition biases and knowledge illusion are often reinforced when people interact in groups. First of all, the selection bias, which has been greatly enhanced by the advent of social media and the filtering mechanism applied on these sites, ensures that people mostly interact with like- minded folks, sheltering them from counterarguments that might force them to reconsider their positions. In these circles, they often find that the knowledge of their chat partners is even more limited than their own, making them feel like real experts. Research has shown that such groups tend to gravitate to positions which are more extreme than the participants’ individual positions are on average – a consequence of groupthink 14 . Of course, by the time this 10 Sloman, Steven and Philip Fernbach. The Knowledge Illusion – Why We Never Think Alone. 2017. 11 Dörner, page 152. 12 Dropp, Kyle, Joshua Kertzer and Thomas Zeitzoff. Political Knowledge and Foreign Policy Attitudes: A Preliminary Investigation of Support for the Use of Military Force in Ukraine. 2014. 13 Niehues, Judith. Ungleichheit: Wahrnehmung und Wirklichkeit – ein internationaler Vergleich. Wirtschaftsdienst 2016, Sonderheft. 14 Janis, Irving Lester. Groupthink: Psychological Studies of Policy Decisions and Fiascoes. 1983. -60 -40 -20 0 20 40 60 80 Among those who could find North Korea Among those who could not Net support for... 4 Source: New York Times, 2017 Vox populi, vox dei or maybe not? 8 | December 12, 2017 Germany Monitor happens, dialogue with groups that have differing views is close to impossible, since the others “are obviously not getting it”. But even if different-minded people come together, reconsiderations or even a consensus is not guaranteed. Once formed, people tend to stick to their conviction, despite mounting contradictory arguments or evidence. Even worse, it quite often causes people to stubbornly insist and defend their position, as they do not want to acknowledge to others and, most importantly, to themselves that they were wrong. This is called the “backfire effect”. People under the sway of this effect tend to respond with “motivated reasoning”, focusing even more strongly on the snippets of evidence which corroborate their beliefs. This is basically a more extreme version of the confirmation bias, where people show a strong selection bias towards confirmatory evidence, while brushing contradictory evidence aside as exceptional or not relevant for whatever reason. Thus, a discussion between people with distinct points of view may not result in a common stance but rather in a situation where everybody is even more convinced than before that he or she is right. Actually, people can be quite good at detecting the weak spots in an argument, at least as long as it is not their own. In an experiment conducted by the cognitive scientist Hugo Mercier, participants were asked to answer a series of simple reasoning problems and to justify their responses. When they were later given the possibility to change their answers, less than 15% of all participants did so. The propensity to change jumped when people were confronted with their previous responses but were led to believe that they were actually from someone else. Among those who did not realise this trick, almost 60% suddenly became more critical and rejected the answers that they had earlier been satisfied with. Similarly, people easily see many of the shortcomings and biases discussed here in others, but fail to find them in themselves (blind-spot bias). What could be done? The evidence presented so far should at least challenge the view that democracy and elections lead to optimal solutions for today’s complex issues. This humanistic dilemma has been eloquently stated in a quote attributed to Winston Churchill, saying that “democracy is the worst form of government, except for all the others”. But is it entirely hopeless? For starters, it might help if the general public had a better understanding of statistics, since most of the complex issues discussed in politics are presented in numerical forms, such as measures of inequality, the pensions expected in 2030 or the likely change of the global temperature. Better familiarity could stop at least the most extreme abuses of statistics in political debate, such as fudging on scales (dimensions), opportunistic selection of time periods shown or presenting correlations as causality. It is certainly no coincidence that the debates about the two political topics climate change and inequality – seen by many as the most pressing challenges for our societies – provide some of the most salient examples for (sometimes purposely) abusing statistics 15 . 15 Since 2012, a group of scientists have had a blog on the website of the German RWI research institute titled “Die Unstatistik des Monats”, which provides further ample evidence. Vox populi, vox dei or maybe not? 9 | December 12, 2017 Germany Monitor Basic knowledge of statistics could help people to understand the limits of experts’ views … Some basic knowledge of statistics and probabilities could also help people to understand and discount the assessments and views of experts when it comes to the economic costs/benefits of political decisions. The disenchantment with experts and their advice was epitomised by former UK Justice Secretary Michael Gove’s claim that “people in this country have had enough of experts”. A YouGov survey found that people have some trust in scientists (71%), but very little in civil servants (26%) and economists (25%); among Brexiters, trust was about 10pp lower for each group 16 . The deep distrust with respect to economists is understandable, but at the same time very disturbing, given that most of today’s problems are economic in nature, such as inequality, pensions and global warming, which is likely the most important issue, as the policy response will have tremendous economic implications (which politicians usually do not reveal to their voters). … and reject pseudoscience The debate about the possible economic consequences of Brexit is a good example where, for the sake of supporting one’s own view with “expert advice”, arguments veer off into pseudoscience. After all, even experts have contradicting views on controversial issues. Even if researchers proceed from the same scientific approach, they might still arrive at different results due to different spatial or temporal restrictions or to different data used. Not only does this have the potential to increase the already widespread mistrust in experts, but it could also lead to a situation where everyone chooses the expert who most closely aligns with his or her own opinion. Pseudoscience is false science, however, which is mainly spread through the internet. Confirmation bias and motivated reasoning make people fall for pseudoscience as long as it supports their convictions. In most cases, people do not verify if the information is actually true. Thanks to the internet, it has become a lot easier to “prove” one’s own beliefs by referring to pseudoscience. As a result, pseudoscience can lead people to cling even more stubbornly to their point of view, convinced that it has been scientifically proved as correct. Unfortunately the clear political leaning of many media outlets – especially in the Anglo-Saxon world – has lowered their standing as an unbiased and neutral source of information in the general public. In Germany especially the slow reporting about the events in Cologne on New Year’s Eve 2015 dented the media’s reputation. Nudging – how much of a guiding hand? Since people might be ignorant or might just not bother to make an effort with a System 2 approach, nudging their decisions towards a “preferred” option might seem to be an attractive alternative to explaining how exactly a policy works and why it should be implemented. The notion of nudging, already widespread in marketing, is relatively new in the political context. Its most prominent proponent, Richard H. Thaler 17 , just received the Nobel Prize in economics this year. Supporters of nudging argue that people are not able to make good decisions by themselves because of their limited knowledge and their restricted rationality. Nudging them towards a certain behaviour can lead to socially 16 YouGov survey results, 15 February 2017. 17 Thaler, Richard H. Nudge: Improving Decisions about Health, Wealth and Happiness. 2009. 0% 50% 100% Nurses Doctors Scientists Historians Sports commentators Civil Servants Economists Politicians Source: YouGov, 2017 General trust in 5 37% 57% 6% has dropped didn't change that much has risen Trust in the German Media 7 Source: infratest dimap, 2015 Vox populi, vox dei or maybe not? 10 | December 12, 2017 Germany Monitor desirable results, such as an increase in organ donation if donating is set as the standard (this means that people have to register if they are not willing to donate, and not vice versa). At the same time, nudging might be just another word for psychological manipulation, prone to abuse by those who are in a position to decide which preferred options people should be nudged towards. This means that nudges can only be used to a limited extent, taking into consideration the principles of transparency. The humanistic view that people should decide for themselves what to do and who they want to be governed by would suggest that people should be provided with neutral and unbiased information, which is impossible . Yuval Harari, for example, doubts that science is in search of truth and notes that in modern societies, science is not used to challenge (quasi-religious) dogmas, but is utilised instead to implement them 18 . This might explain the general public’s increasing scepticism with regard to scientific research and advice. Monetary policy could serve as an example where one might ask whether the adherence to the neo-Keynesian dogma (Phillips curve, output gaps and inflation expectations) has dominated analysis and policy response to such an extent that alternative interpretations of the situation (Austrian, supply side) have been considered almost heresy. Former ECB chief economist Otmar Issing described central banks, which have based their policy measures on only a few models, as being on the wrong track 19 . Around 53% of all Germans state that they “rather distrust” the European Central Bank (ECB) 20 , despite its formal independence from government – which, at face value, should improve the general public’s trust in the institution. Indeed, the ECB has had little success in “nudging” Germans towards borrowing more by enticing them with record-low interest rates. On the contrary, the ECB’s extreme measures, which have led to negative interest in markets, have added to the concerns – widely heard at least in Germany – that something must be fundamentally wrong. This suggests that formal independence by itself does not guarantee being perceived as an independent and unbiased institution. The Dutch experience with the CPB’s role, the national statistics office, on the other hand, provides an encouraging example. Ahead of an election, the CPB evaluates the economic and fiscal effects of the various election programmes. This prevents parties from promising measures where the financing is just pie in the sky. Although the bureau’s analysis is used rather opportunistically by politicians during the campaign, it still anchors the debate and shows that it is possible – at least to some extent – to perceive experts as non-partisan and independent. Can we influence how people approach political questions? Research shows that when opinions are driven by emotions and values, they tend to be immune to reasoning. Sloman and Fernbach observe that politicians quite often cast their policies in such value-based terms in order to avoid a consequentialist analysis of their likely implications. They suggest that political discussions should focus on the actual consequences of a policy proposal, rather than on defending values and beliefs. Being forced to think about likely consequences of a proposed political measure for others could soften dogmatic views. It might also help to make people aware of their own very limited knowledge, and it might even lead to a more open and less self-opinionated debate. Sloman and Fernbach demonstrate that much like the overestimation of 18 Harari, Yuval Noah. Homo Deus – A Brief History of Tomorrow. 2016. 19 Frankfurter Allgemeine Zeitung, 12 May 2016. 20 europaeische-zentralbank/ 0% 20% 40% 60% 80% 100% Nurses Doctors Scientists Civil Servants Lawyers Economists Bankers Politicians Trust to tell the truth 6 Source: Ipsos MORI survey, 2016 0 10 20 30 40 50 60 70 80 Trust and confidence in the US press 8 Source: Galup data May 1972 - Sept 2017 % Vox populi, vox dei or maybe not? 11 | December 12, 2017 Germany Monitor their knowledge, people also initially overestimate their agreement with a certain policy. If people realize that they are unable to adequately explain how a certain political issue works, they tend to reduce the extremity of their position. Hence, discussing the implications of political proposals can help to make people aware of how little they know and how unfounded their extreme position is. In other cases, it might also help to think outside the box. Thinking outside the box describes a thought process which aims at the implementation of an unusual approach to the logical thinking structure. More precisely, it is a metaphor for thinking in a different and unconventional way or from a different perspective. The “box” might be the implicit assumption that one’s own point of view is “right”, whereas all the others are “wrong”. Leaving the box by changing one’s point of view can therefore help to overcome tunnel vision and to reduce the extremeness of political positions. Final thoughts These normative considerations about how to improve the political decision- making process make us somewhat optimistic that people and society can improve, which is probably also evident in the changes we went through during the last one or two centuries. Still, the evidence presented here, which shows that people’s voting decisions are not solely the outcome of a rational process confined to System 2, holds a stern warning for forecasters, in particular economic forecasters, whose calls quite often crucially depend on their assumptions about political developments. Assuming that people or governments will not do stupid things because they do not want to face the costs might be a futile approach for making forecasts about politics. People and governments do senseless things perhaps because they do not anticipate the costs, which might not even be known at the time of the decision, or because the aforementioned individual biases and group dynamics lead them to ignore the costs altogether. Still, when it comes to democracy, we firmly side with Winston Churchill. If a democratic government consistently fails to deliver, it can and will be ousted in a structured, non-violent process called elections. In our view, this element makes democracy far superior to all other forms of government, even if the voter cannot be sure that the new government will really do a better job. Stefan Schneider (+49 69 910-31790, In the "Germany Monitor" series we address political and structural issues which have great significance for Germany. These include commentaries on elections and political decisions, as well as technology and industry issues, and macro- economic topics which go beyond the business cycle matters addressed in "Focus Germany". Germany Monitor Our publications can be accessed, free of charge, on our website You can also register there to receive our publications regularly by E-mail. Ordering address for the print version: Deutsche Bank Research Marketing 60262 Frankfurt am Main Fax: +49 69 910-31877 E-mail: Available faster by E-mail: * Vox populi, vox dei or maybe not? ......................... December 12, 2017 * German stock market rises to record highs: All is well – or is it? ................................................ November 28, 2017 * German labour market: Much remains to be done! ..................................... November 13, 2017 * Bundestagswahl 2017: The final countdown ......... September 19, 2017 * German industrial policy: General restraint commendable, but key areas that need to be addressed are being neglected .............................................. September 12, 2017 * Housing policy in Germany – Changing direction ........ August 31, 2017 * State or private, redistribution or growth – Who will be the winners according to the parties’ tax proposals? ........ August 28, 2017 * Time to fully recognise the potential of digital start-ups .......................................................... August 22, 2017 * German defense policy: Towards a more integrated security framework ...................................... August 8, 2017 * Packed European agenda for the next government: Numerous challenges, no (easy) answers ................... August 3, 2017 * Germany’s fiscal situation: Full employment and zero interest rates result in budget surpluses – but demographic development might become a problem! ..... July 19, 2017 © Copyright 2017. Deutsche Bank AG, Deutsche Bank Research, 60262 Frankfurt am Main, Germany. All rights reserved. When quoting please cite “Deutsche Bank Research”. The above information does not constitute the provision of investment, legal or tax advice. Any views expressed reflect the current views of the author, which do not necessarily correspond to the opinions of Deutsche Bank AG or its affiliates. Opinions expressed may change with out notice. Opinions expressed may differ from views set out in other documents, including research, published by Deutsche Bank. The above information is provided for informational purposes only and without any obligation, whether contractual or otherwise. No warranty or representation is made as to the correctness, completeness and accuracy of the information given or the assessments made. In Germany this information is approved and/or communicated by Deutsche Bank AG Frankfurt, licensed to carry on banking business and to provide financial services under the supervision of the European Central Bank (ECB) and the German Federal Financial Supervisory Authority (BaFin). In the United Kingdom this information is approved and/or communicated by Deutsche Bank AG, London Branch, a member of the London Stock Exchange, authorized by UK’s Prudential Regulation Authority (PRA) and subject to limited regulation by the UK’s Financial Conduct Authority (FCA) (under number 150018) and by the PRA. This information is distributed in Hong Kong by Deutsche Bank AG, Hong Kong Branch, in Korea by Deutsche Securities Korea Co. and in Singapore by Deutsche Bank AG, Singapore Branch. In Japan this information is approved and/or distributed by Deutsche Securities Inc. In Australia, retail clients should obtain a copy of a Product Disclosure Statement (PDS) relating t o any financial product referred to in this report and consider the PDS before making any decision about whether to acquire the product. Printed by: HST Offsetdruck Schadt & Tetzlaff GbR, Dieburg ISSN (Print): 2511-8552