Sebastian Ille, Lecturer in Economics, New College of the Humanities

The public vote on Brexit has left many flabbergasted and rekindled the media discussion about a more direct and decentralised democratic process. Those against referenda essentially claim that a representative democracy is better placed to make decisions on complex matters of state, while being less susceptible to demagogues or “fake news”. Supporters of referenda suggest that elections are subject to gerrymandering, while representatives are only bound by incomplete contracts, thus face the same demagogic forces and their own personal ambition for power. Furthermore, according to them referenda strengthen voters’ interest, increase turnout, and acceptance of results. In the end, isn’t an individual more apt to express their intentions and preferences than a representative?

This discourse, so it seems, mainly concentrates on the cognitive and emotional limitations of voters. Consequently, neither advocates nor opponents seem to question the underlying premise that once members of a collective are sufficiently informed, sensible and consistent, a majority decision based on an aggregation of the individual direct votes will provide a proper representation of the collective’s will and interest. However, we should ask: Is this generally the case or does our intuition fool us?

Economic theory may shed some light on this conundrum. Indeed, the current controversy regarding referenda bears a remarkable resemblance to the age-old economic discussion on state intervention. On the one hand, we find proponents of demand intervention who believe that the state can mitigate market imperfections caused by coordination failures. On the other hand, supporters of laissez-faire economies rely on the self-correction and long-run efficiency of markets which should best be left to alone.

It is therefore not surprising that economists have been pondering a similar question in the context of good provisioning. Most publicly provided goods are some sort of mix between a public good and a common good (or common property resource). When creation and use of these goods is based on individual decisions, public goods are frequently underprovided (which explains low vaccination rates, for example), whereas common goods are oftentimes over-exploited (such as fish stocks). The reason is that these goods are defined by two critical characteristics – firstly, they are non-excludable i.e. access to these goods are granted to all members of a community, and secondly, individual actions cause (frequently unintended) consequences that benefit or harm a third party, i.e., they create positive or negative externalities. Although individuals optimise their own wellbeing, the collection of these actions does not translate into a welfare maximising outcome for the community. This inconsistency between individual decision and aggregate poses a typical aggregation problem – an aggregate does not inevitably possess the same properties as its individual constituents. In other words, a choice or action which an individual considers beneficial when left to itself might not be equally beneficial for the individual as part of a collective.

An example will clarify the issue. Assume that a mayor of a small town plans to create a recreational area. Once constructed, the area will be publicly available to all residents. Each of the 10,000 inhabitants is asked to voluntarily provide a financial contribution of £500 or an equivalent time of community work. If every member of the community contributes to the project, a recreational area can be created that generates a life time benefit of £1,500 for every member, translating into a net benefit of £1,000 per capita. Each individual contribution conveys a small external benefit to every member, since access to the area is open to any inhabitant, yet also independent of one’s contribution.

Under these conditions and especially in the context of larger communities, an individual member probably decides not to contribute. In the end, the effect of his or her individual contribution is very small and thus has barely an effect – at least to a much lesser degree than one’s own costs of £500 (a budget reduction from £5 million to £4,999,500 will not change the shape and quality of the recreational area). Each member can reason in the same way. If left to individual choice, a majority of inhabitants will choose to free-ride and therefore not contribute, leaving our mayor without the means to construct the recreational area, which would have generated a net benefit for each member.

Our knowledge of the aggregation problem in the presence of externalities and free-riding can help us understand the shortcomings of referenda which concern questions of distribution and contribution. The same reasoning applies, although in a slightly modified form. Take the recent Brexit vote. Much like the recreational area, the Eurozone possesses public and common good characteristics. Admittedly, the EU suffers from weaknesses, such as over-exploitation, but it also provides benefits from which none of its members is excluded – like access and trade in a joint market, as well as better consumer and labour rights. Yet, voters ignored the EU’s public good nature, mistaking the question of remain or leave as one of redistributing a fixed pie of benefits, while neglecting that the outcome will alter the amount to be distributed.

Similar to the recreational area, the EU comes at a monetary (and probably social) cost. Indeed, most Brexit supporters solely focused on the cost of membership. In reality, British citizens contribute less than £100 per capita to the EU; an amount at the lower end of Western European countries (at about one fourth of the contribution of a Dutch and roughly equivalent to the contribution of a French or an Italian) and identical to what a Norwegian pays for accessing the single market, yet without obtaining voting rights. However, UK citizens were wrongfully made to believe that the UK would still harness the benefits of the EU, such as the single market, without having to pay a contribution – in other words to free-ride.

As illustrated before, the aggregation problem thus raises the question of whether referenda are inadequate to solve political questions concerning distribution and contribution that bear a public good nature, since they may incorrectly represent the citizens’ interests. This holds especially if referenda are considered irreversible and individuals are not allowed to abrogate their decision. Clearly, a similar argument applies to a fully representative democracy, but it allows for parliamentary debate, a constructive opposition,  and eventually requires consensus of the ruling coalition, all of which render decisions revocable or should at least mitigate the aggregation problem.