Much ink has been spilt in the pursuit of a theory of politics. Some of the earliest reflections on political power can be found in ancient Greece. In Plato’s dialogues, characters such as Callicles and Thrasymachus argue that justice is ultimately a matter of power: the strong rule, and the weak obey. At the same time, the Greeks developed some of the first democratic institutions, allowing matters of policy to be decided by public vote. As Pericles put it, “we do not all take part in making policy, but we are all capable of judging it.” Plato’s writings amount to a sustained and often hostile critique of these new and unprecedented forms of government, which he resisted vehemently, and whose criticisms still echo to this day. One such critique, expressed through his portrayal of Socrates, describes democracy as a system in which the people rule, yet are fickle and susceptible to demagogues and despots, rendering the system inherently unstable.
Popper’s Criterion
According to the philosopher Karl Popper, this rests on a profound misunderstanding of what democracy is. Plato frames political theory as an answer to the question, “Who should rule?”, and argues that in a democracy it is the people who rule, even though they are, in his view, unfit for this role. Popper rejects this framing altogether. The central question is not who should rule, but how rulers can be replaced. He therefore proposes that democracy is best understood as a system that allows for the peaceful transfer of power. In such a system, one ruler can be replaced by another without violence, allowing for peaceful error correction. This is sometimes called Popper’s criterion.
Popper’s theory of politics is distinct from that of other thinkers in that it does not describe in detail the workings of actual democratic countries. It does not mention parties, prime ministers, or presidents. All it mentions are rulers and peaceful transitions of power. Yet it is precisely because of its simplicity that Popper’s criterion acquires a generality that would not otherwise be achievable, as different political systems can realise peaceful transitions of power in very different ways.
In the United Kingdom, Parliament is the highest law-making body in the country, and it is often said there that “Parliament is sovereign”. That is, it cannot be bound by other law-making bodies, nor even by its own past decisions. Given this degree of authority, how does the British system ensure that bad rulers can be replaced without violence? The British system does this by concentrating responsibility in the office of the Prime Minister. If the country is not being run well, responsibility is clear. A Prime Minister who loses the confidence of Parliament can be removed without violence, either through internal party mechanisms or a vote of no confidence, triggering a change in leadership or an election.
Other countries, such as the United States, adopt a different approach. There, power is divided between separate branches of government. The President can be checked by Congress, Congress can be checked by the President through veto powers, and both are subject to judicial review by the Supreme Court. These checks and balances are intended to prevent the concentration of power and to allow for correction without violence.
An Extension: The Cost of Replacement
Popper’s criterion prescribes that systems of government should ensure that bad rulers can be replaced without violence. Violence, however, is only one constraint. Systems that require violence to correct mistakes are clearly inefficient, but even non-violent decision-making may be open to improvement: decision-making bodies can be slow, wasteful, or poorly informed even when no violence is involved. I therefore propose the following extension of Popper’s idea: good political systems minimise the cost of replacing bad rulers and decision-makers (this reformulation also makes the criterion easier to incorporate into modern social choice theory, where efficiency has a more precise meaning). This extended version of Popper’s criterion points more broadly towards decentralisation of decision-making, both in political life as well as in the economy and even within families.
For instance, politicians are decision-makers who decide on policies and laws governing a country. Sometimes, that can include nationalised industries, which then become part of the political system of that country. Whenever this is the case, we may question whether the nationalised industry satisfies Popper’s criterion and whether there are alternative arrangements that would better satisfy it. Whenever a nationalised industry is being mismanaged, there are few incentives to change course: even when voters agree that there is mismanagement, they might not agree on what policy to adopt to solve it; moreover, the average voter will lack the information necessary to determine which potential leader of the industry could best replace the current one. In addition, nationalisation often results in a single provider dominating the market, making it difficult for customers to switch to alternative services, while employees have fewer employers available to them if they are dissatisfied.
This naturally raises the question of whether alternative arrangements would perform better. In particular, which systems allow bad decision-makers to be replaced more efficiently? In a competitive, privately run industry, employees can change employers and customers can switch suppliers. If a CEO governs poorly, a board of directors, equipped with detailed, local knowledge, can remove them. In a publicly run company, by contrast, replacing leadership requires voters to acquire extensive information about managerial decisions, at considerable cost and with little individual benefit, since any gains from that knowledge are shared across the electorate.
Between these models of organising industry, the private one is therefore more efficient at enabling the replacement of bad decision-makers. This is for two reasons. First, privatisation allows multiple organisations to exist side by side, competing for workers and customers. When one organisation is dysfunctional, individuals can exit and choose another. Second, decision-making is placed in the hands of those with specific, relevant information and strong incentives to use it well.
Private companies themselves have some inefficiencies. Anyone who has had a hard time cancelling a subscription to an overly pricey, bad service knows that there is substantial friction in changing services and that some bad actors in the economy can exploit this. Yet, when comparing private alternatives to public ones, the inefficiencies of the latter are generally much more substantial than those of the former.
Politics, too, allows for decentralisation of decision-making. In countries like Switzerland, local government plays a large role in shaping policy in those areas. This effectively makes those polities compete with one another for taxpayers. Hence, there are large differences within Switzerland regarding the taxes one pays to the canton, and cantons have even been known to negotiate tax rates with businesses willing to relocate their headquarters there. Similar competition existed in the Netherlands during the Dutch Golden Age, as the Dutch provinces were free to make laws in much the same way that the cantons are today. Thus, even within democratic governments, which satisfy Popper’s criterion of non-violent change of power, there are significant differences in terms of the efficiency of government, and generally those with more decentralisation are better able to leverage competition and local knowledge to make decisions more effectively.
Limits of Exit: The Case of the Family
Another domain in which relations of rule often go unrecognised is parenting. Parents are traditionally the ones with decision-making power within a family, choosing where the family lives, which schools children attend, which activities are permitted, and what punishments are imposed for rule violations.
In the case of parenting, it is extremely difficult to replace one’s parents. Children could, in principle, put themselves up for adoption, but this is usually undesirable for three reasons. Children typically love their parents; adoption offers no guarantee of improvement; and the process itself is slow and emotionally taxing. Other options, such as running away, are scarcely better.
Because exit is so costly, parents can behave badly without their children being able to respond by leaving. From the perspective of Popper’s criterion, this suggests that parental authority should be exercised with great restraint. Where rulers cannot realistically be replaced, rule-making and enforcement are especially likely to generate conflict, and at times even violence, that is difficult to resolve peacefully. That is one argument in favour of involving children in decision-making as much as possible. Such liberalisations in family dynamics have already happened to a large extent. It used to be commonplace for parents to decide whom their children would marry, what they would study, or what career path they would follow, showing that significant life decisions were left to parents with little to no involvement from those affected. Today, many of those choices are instead made by children, although they are still obliged to attend school by force, often by law, showing that not all decision-making is shared.
Decentralisation
What all of the above examples have in common is that decision-making can be greatly improved by two mechanisms: competition, which allows people to try out different ideas simultaneously to see what works best, in such a way that the best plans and choices generate the most value and are therefore the most likely to survive; and second, incentives to use local knowledge, since much decision-making depends on having the requisite information, which is costly to obtain and therefore requires incentives. Broadly speaking, these two aspects are reasons why decentralisation is more effective. In this sense, the central problem of politics is not who should rule, but how easily bad rulers can be replaced, and at what cost.