Forecasting Blog

Read the Latest News

The rise of Donald Trump, and perhaps also the success of the Brexit campaign, brings into focus the median voter theorem which has informed rational choice analysis in politics since the 1950s when an economics professor, The theorem is predicated on the assumption that people do not vote for policies which harm their own interest. We now witness that the expression of discontent over globalisation around the world is not be guided by the above postulate of rationality but by behavioural patterns reported by psychologists in their experiments on ultimatum games. These experiments suggest that people turn down even those offers that would make them better off if they perceive the distribution of gains as unfair. Many of those that vote for Trump would be worse off, certainly economically worse off, under a Trump presidency because his economic policies are whimsical and erratic. This renders traditional models unhelpful in making forecasts in this election. We shall presently come to the question of the potential outcome of the US presidential election in November, but first let us look at the Black’s median voter theorem, an idea underlying traditional forecasting models.

I must emphasise that Black’s theorem is not disproved by evidence of the rise of Brexit and Trump movements. It is simply that the models of rational choice must be modified to account for the idea of fairness in the design of policies that could deliver prosperity through international cooperation. Insofar as prosperity entails engagement with the wider world, failure to address public concern over distributive fairness in globalisation could, at best, usher in an era of inward-looking nationalism and economic decline or, at worst, give rise to authoritarian and xenophobic movements which pose existential threats to liberal democracies. I have argued this elsewhere [1].

The Nobel Laureate economist Ronald Coase includes Duncan Black amongst roughly only a dozen economists whose ideas he highlights in a set of essays on developments in the history of economic thought. Duncan Black, the economist, collaborated with Dr R. A. Newing in our mathematics department to further examine a problem raised by their colleague, Professor Campbell, who taught Philosophy at Bangor. Campbell was concerned with the problem of how an individual’s actions and her desires might be commensurate. Any particular action has an impact on more than one desire. Black formulated a parallel problem by substituting desires for preferences and choices for actions. The problem was of ascertaining social preferences by aggregating individual preferences manifest in the choices they made. Extant literature in mathematics and philosophy suggested that aggregation of binary preferences of individuals over a domain of alternatives could result in a paradox that the majority choice might not always satisfies a condition of rationality, called transitivity in choice, even if this condition is satisfied by all individuals. Transitivity of expressed preferences – eg if Option A is preferred to Option B and Option B is preferred to Option C, then transitivity entails that Option A must be preferred to Option C – was considered by many to be an essential feature of rational behaviour. An illustration of this feature of social choice was given by Nicolas de Condorcet in the 18th century.

In one of his papers published during his lifetime – many of his other writings were published posthumously by renowned scholars because his ideas were before their time and referees had difficulty engaging with them – he suggested that the above paradox does not arise if a further assumption of rational behaviour is made, called the assumption of single peak in preferences. When a preferred option is not available, then the option closest to it is chosen. Suppose that I should like to spend £100000 on schools rather than spending £50000 or £10000. I would choose £50000 if offered a choice between £50000 and £10000. I would not choose an option, eg £10000, that is less beneficial to me. If plotted on a straight line, eg going from low to high expenditure, my preference would show up as a curve with a single peak.

With the above additional assumption of rational behaviour, Black proved his median voter theorem. In a single dimensional choice, for example casting a ballot amongst right and left wing parties or making a choice amongst various expenditure levels, the majority choice would favour the choice made by the median voter. This theorem has given rise to the Duverger’s Law, a conjecture that political parties must claim the centre ground to win in a first past the post voting system as in the Westminster parliament and in the US elections. Trump’s success in being nominated as the Republican candidate has put a dent in that conjecture.

Writing in The Atlantic, Erick Beinhocker of the Institute of New Economic Thinking at the Oxford Martin School makes the case that the rise of inequality may explain the support for Trump by even those that might become worse off under a Trump presidency. They are alienated from the received political dispensation the distributive outcome of which they perceive as unfair. They form a large swathe of the American population that nurtures a grievance against the political system. Hilary Clinton’s campaign highlighting the dangers of a Trump presidency might be less effective than it might have been if the fruits of economic growth were not so unequally shared. The rise in the real personal disposable income in the US between 2000-2015 was about 21 per cent, but income of the lowest 20 per cent rose by only 1.5 per cent. Even more remarkable is the fact that around 60 per cent of the population witnessed rise in income below the national average. To add salt to injury, the income of the top percentile rose by 41 per cent. This is a trend which began around 1970. In the post-war period between 1947 and 1970, average income in the US rose by 85 per cent, but the income of the poorest 20 per cent almost doubled. Things became different after 1970 when the whole of the bottom 60 per cent of the population began to slip behind the national average increases in prosperity. At the same time, the top percentile suddenly started to witness a phenomenal rise in their share of the national income. The Nobel Laureate economist Robert Schiller in collaboration Maxim Boycko, erstwhile advisor to President Boris Yeltsin in Russia, has conducted surveys to ascertain people’s perception of fairness in the distribution of income. The present state of affairs does not come anywhere close to American people’s idea of a fair distribution of income. An economic system which generates such egregious inequality, as the US system has generated since the 1970s, begins to forfeit the loyalty of even from those that do well under that system.

Under the circumstances, it would be foolish to rule out, polls notwithstanding, a Trump win, especially if alienated voters turn up in numbers to exercise their franchise. Only about 60 per cent of the eligible voters have cast their ballot in recent presidential elections. Higher levels of education and income correlate well with turnout. If some of those that generally do not turn out to vote are energised by the Trump campaign, Clinton would suffer a handicap. For example, Trump disproportionately attracts the less educated white, non-Jewish, males on low income. A sudden increase in turnout by this group could confound predictions. Also, Trump could win even in defeat as explained below.
Trump’s policies are a random juxtaposition of offensive pronouncements erratic policy prescriptions. However, there is a coherent message that the institutions of governance deserve no respect. In this he might even win in defeat exploiting anomalies in the US system. Hilary Clinton could garner a majority in the electoral college by winning only about a dozen coastal states plus Hawai. Many of these are industrial states which have witnessed demographic changes which do not favour Donald Trump. By focussing on congressional and senate races in the interior states, Trump might be able to deliver a majority in the legislature to Trump-supporting candidates. Since the focus of the Trump campaign is to disrupt, a situation could arise whereby Hilary Clinton, having won the presidency, would find a legislature dominated by those whose aim is to disrupt proceedings to underline their contention that the system does not work. Trump would then win in defeat.