Democracy Questioned How Democracy Can Fail Us, and What to do About It

Every society has to answer the core questions of power: With whom does it rest? How is it wrested? How is it expressed? To answer these questions, humanity has experimented, at different times and in different places, with different forms of government; tweaking, syncretizing and discarding systems as we go along. Democracy has come to occupy a sacrosanct place in the global political conversation. A mere mention of its name in said conversation provides the sometimes-specious patina of legitimacy to the political system or process in question, so much so that perhaps the world’s most repressive nation, calls itself the Democratic People’s Republic of Korea; and Russia’s authoritarian leader proclaims Russia to be a sovereign democracy. No system of government has come to have the sort of appeal democracy has today. Considering the many lives sacrificed to topple authoritarian military dictatorships and usher in self-government for the citizen, democracy appears to be a system worth dying for, and, if you take the United States’ putative goals in Nicaragua, Iran and Afghanistan at face value, apparently one worth killing for.

Contrary to common assumption, democracy did not emerge seamlessly and linearly from Athenian experimentation to modern praxis. Instead, its evolution came about in stints and bloody spurts. Initially in fifth-to-fourth-century BCE Athens, only citizens were allowed to vote, but only a culled group of people were considered citizens—resident aliens, women and slaves were not granted access to power. A short-lived experiment, Athenian democracy suffered a crushing defeat in 338 BC at the heels of Macedonia led by Phillip and his son Alexander the Great. It stayed dead for close to 2000 years before being resurrected by political theorists and armed militants in eighteenth-century France as a viable ideal.

Jean-Jacques Rousseau redefined sovereignty, not in the divine right of kings, but, in the power of the people, and inspired militants to reconsider the placement of power. This reconsideration culminated in a realization that an armed revolution was the only viable means to claim power. The ensuing bloody French Revolution, which saw about 250,000 people dead, dismantled the monarchy—literally by executing the King—and introduced democratic ideals to France, bringing the dream of Athenian democracy to life again.

Modern democracy is vastly different from the democratic ideas championed by French revolutionaries and Athenian democrats. In Athenian democracy, power sat with the people, whereas in modern democracy, as one of the signatories of the American Declaration of Independence, Benjamin Rush, put it ‘all power is derived from the people. They possess it only on the days of their elections. After this, it is the property of their rulers, nor can they exercise or resume it, unless it is abused.’ This type of democracy is indirect and takes different forms. For example, the United States of America is a republic while the United Kingdom is a constitutional monarchy. What matters is that representatives are elected by the people.

Indirect democracy is certainly more efficient when dealing with larger groups. A direct democracy worked in Athens because Athens was not only a small city, the city also restricted voting to certain individuals. As modern democracies become more liberal, expanding enfranchisement, it becomes inefficient to have a direct democracy. Robespierre, one of the influential figures of the French Revolution, realized this and made a volte-face on the necessity of direct democracy. He recommended a temporary dictatorship be instituted for the republic to stabilize before instituting a representative democracy. He was later arrested for this and summarily executed. He turned out to be correct. France, today, is a unitary state.


Its exclusionary and gory history notwithstanding, democracy is often claimed to have produced marked benefits. Daron Acemoglu, the American economist, argues democracy causes growth. Julia Ruiz Pozuelo of the Brookings Institute argues it doesn’t. Resolving that dispute would require a careful reading and interpretation of the facts and statistical measures employed, but it is reasonably safe to say that the worst democracies perform better than the worst autocracies.

We must be careful, however. We often conflate liberalism and democracy, but these two, whilst commonly presented as inextricable in the West, can be mutually exclusive. This is supported by the historical (and current) existence of non-liberal democracies and liberal nondemocracies. The Athenians were not concerned about the ‘people’. They limited enfranchisement when they could, at some point limiting it to only people who had Athenian-born parents, a shift from only requiring an Athenian-born father. They did not provide freedom of speech or religion, protection of private property, or constitutional government. Iran is another good example where for most of the Islamic republic’s history, it has employed a functioning democracy, but one rarely marked by any indicator of liberalism. The birthplace of liberalism, modern England, still upheld a monarchy well into the nineteenth century. This is important to note because when we measure the impact of political systems, we often conflate the impact of political and economic philosophies such as liberalism with governing methods such as democracy.

Plato’s issue with democracy was that democracy pandered to desire without making sure people wanted the right things. Modern democracy, on the other hand, arrogates to itself a certain protection that often shields it from censure. Many political thinkers refrain from saying anything disparaging against the democratic enterprise, lest they be branded heretics or autocracy apologists.

The imperfections of the many instances of the democratic model in Africa have registered in our collective consciousness. Reports of electoral fraud in Congo, paroxysms of electoral violence in Sierra Leone and barefaced voter intimidation in Nigeria are not unheard of. Some of these failures are attributed to the expected nature of nascent democracies. However, recent crises of governance in the West—Brexit in the UK, America’s Trump—have shaken our confidence in democracy and poked holes in its once-impenetrable armour. This, coupled with observed economic growth in non-democratic countries such as China and Singapore, has undermined our belief in democracy and amplified calls for political alternatives. A survey conducted by the Edelman Global Public Relations shows that citizens in authoritarian China show more satisfaction with, and trust in their government than citizens in democratic USA. ‘Democracy is the worst form of government, except all those other forms that have been tried from time to time,’ many proponents resort to. But this cannot be accepted by fiat. Whether democracy can succeed as a political institution or as an ideal to be aspired to must remain a question to be investigated.

Arguably, democracy has had a good run. Despite its good run, however, there are systemic flaws in democratic models that could prove problematic, and it is necessary to evaluate key changes or consider alternatives. There are chinks in the democratic armour in three significant areas—the voter, the voting and the voted. These three aspects of the democratic undertaking have flaws that might hamper democracy from yielding the best policies.


The entire democratic ideal presumes the individual voter deserves a say in how their life ought to be directed, and since the selected policies would be extended to others, how others’ life ought to be directed. The ideal voter is thus assumed to have clear preferences and ordered priorities about what their government should do or not do, and also a sense of what the common good is. But this idyllic picture of the voter is sharply divorced from reality. It is becoming apparent that we the voters are politically ignorant and irrational.

‘Voter Ignorance’ often stands accused as the Achilles’ heel of democracy as an informed electorate seems a prerequisite for democracy. A survey conducted by University of Pennsylvania’s Annenberg Public Policy Center showed 33 per cent of American voters can’t name any branch of government. 37 per cent could not name a single right protected by the First Amendment. In a 2017 Perils of Perception survey conducted by Ipsos MORI, South African respondents had the most inaccurate grasp of the key features of their population in their country.

It is easy to castigate voters as just being lazy or benighted, but voter ignorance is simply a carefully calculated consequence of human self-interest. The larger the group, the less likely one singular vote would be the deciding factor of an election; the marginal effect for each vote is effectively non-existent. Each vote is counted, but each hardly counts. This realization, coupled with the fact that it can be very costly to obtain accurate information, makes political education a fool’s errand.

This economic fact is anathema to some democracy maximalists, at least the ones who believe the informed voter is key to a successful democracy. They argue massive voter ignorance on pertinent topics such as market economics, the functions of the branches of government and the stances of the voters’ political representatives, combined with the voter amnesia during the next election cycle is a big problem. This ignorance, when juxtaposed with the focused yet extensive insider knowledge that lobbyists, big corporations and special interest groups seem to possess, is a recipe for disaster.

Yet, some democracy maximalists are not convinced this is a big issue. Some argue there is a measurement error—voters know more than indicators show they do, or there are no valid measurable indicators to capture voters’ capacity to vote well. Moreover, all voters need to know is how well their elected candidate performs and if he doesn’t perform well, they punish him by ousting him or his party. Perhaps they use heuristics or mimic the knowledgeable elites they trust. Some argue this is a categorical error—the United States, for instance, is not a direct democracy, and as such the requirement for informed voting is exaggerated.

A direct democracy might need informed voters, but a representative democracy doesn’t really need voters to be as informed because the ultimate decisions are being made by the representatives. A retort would be that knowledge is still important to judge the representatives. Some take the extreme argument—a democracy does need an uninformed citizenry. The amount of time it takes to be knowledgeable is proportional to passion and interest and this might seem a good thing but it’s actually not. As Bernard Berelson and his colleagues put it, ‘the apathetic segment of America probably has helped to hold the system together and cushioned the shock of disagreement, adjustment, and change’. This argument says expanding enfranchisement is good, not because we make better decisions but because we increase a cushion for inevitable feelings of loss. To what extent this is true is debatable, perhaps participation is sufficient to cause passion and ignorance doesn’t reduce passion.

One last group suggests that while the average voter might be ignorant, perhaps we should cast our gaze away from the average voter and look to the entire group. Large groups of people, after all, are considerably smarter than individuals. One of the oft-repeated examples of the powerful effect of aggregate opinion is guessing the number of jelly beans in a jar. When there is a sufficiently large group, the average answer is highly accurate and better than the best individual guess. Furthermore, in reality, there is a sizeable population of the electorate that is politically engaged and knowledgeable. This might not seem of large consequence, but it actually is. Even if there is a small group of the population, say 1 per cent of the general electorate, that is informed while the rest of the electorate votes at random, the election outcome is going to be very similar to one where 100 per cent of the electorate were fully knowledgeable. This is because the 99 per cent of the completely ignorant electorate vote randomly splitting their votes equally amongst the candidates, the ultimate decision will rest in the hands of the 1 per cent informed voters, who would place informed votes. The cavil of rational ignorance being the death knell of democracy therefore is overextended. Democracy, it seems, has this in-built bulwark against ignorance and can function appropriately as long as voter ignorance is not absolute, they argue.

This bulwark, termed ‘The Miracle of Aggregation’, only holds if the errors are random or, as Bryan Caplan describes in his very powerful book The Myth of the Rational Voter, non ‘systematic errors’. While there is certainly a wisdom in the crowd when we are guessing jelly beans in a jar, the crowd has also been known to show fits of madness—in economic bubbles, stock market crashes and religious mania such as witch burning and the Crusades. These are situations where our cognitive machinery fails us individually, and as a group. The Miracle of Aggregation as a way to cancel out ignorance fails if humans make non-random errors or act irrationally.

Unfortunately, it appears voters are not just rationally ignorant, more perniciously, we are irrational. We have an interest bias. We tend to subscribe to political beliefs that, if selected as governing policies, would benefit us or the group we choose to identify with and not the entire population. We have coherence bias. The real world is complex. We are inundated with so many facts and it is far easier to use mental shortcuts as we engage with the world. To act in this world, we have to denude the world of nuance and reconstruct a simpler model. More importantly, it is important to us that this model aligns with our preconceived notions. Jonathan Haidt in his book The Righteous Mind, demonstrates that our political convictions are primarily determined by intuition and then justified post-hoc. Effectively, we are not reasoning creatures, we are rationalizing creatures. Our brains are not necessarily hard-wired for facts, but they are hard-wired for coherence. We see this confirmation bias in our analysis of facts that disturb our beliefs. Lastly, we have an Identity bias. Our political beliefs, much like religious beliefs, form a core part of our identity. We want our political beliefs to be true and will avoid countervailing evidence that challenges the veracity of our political beliefs and meet any criticism of our cherished political worldview with dogmatic wrath.

But it is not enough that we are irrational, the errors that come from our irrationality have to be relevantly harmful to democracy. Caplan asks ‘Are voter errors systematic on questions of direct political relevance?’ Economic issues and social issues are perhaps the most politically relevant issues and as such we ought to be concerned about whether our errors in those magisteria are indeed systematic. With regard to economic issues, Caplan answers affirmatively ‘People do not understand the ‘invisible hand’ of the market, its ability to harmonize private greed and the public interest. I call this antimarket bias. People underestimate the benefits of interaction with foreigners. I call this antiforeign bias. People equate prosperity not with production, but with employment. I call this make-work bias. Lastly, people are overly prone to think that economic conditions are bad and getting worse. I call this pessimistic bias.’

The problem of ignorance is indeed massive. If you don’t know how your government works you cannot keep your representative accountable when the time comes to re-elect or oust. The problem of irrationality is worse, because it is intractable and because it underpins every attempt to solve other important problems.


In general, aggregating individual preferences into group preferences is a perennial problem. One of the central issues in the voting theory is described by Arrow’s Impossibility Theorem, which implies that no fair voting system can result in a sensible result. This is not to say they are all equally senseless. Some, I believe, are more senseless than the others. The very popular modern democracies—The United States, The United Kingdom, and other countries where the west has exported its version of democracy, such as Nigeria—make political decisions via the most senseless of systems.

Nigeria, like the United States, uses plurality voting (or First-Past-The-Post) where a voter is allowed to vote for only one candidate and the candidate who polls the most is elected. This seems logical at first glance. But to what extent is it free? This method consigns you to voting for one and only one candidate even though there are many other candidates whom you think could be equally qualified. It is stifling and inexpressive. But this is not the only problem.

In 2016, Mayor Bloomberg decided to abstain from the US presidential election race because ‘there was a good chance [his] candidacy would could lead to the election of Donald Trump’. He chose not to run, reducing the number of potential options, because he did not want to cause adverse results. This is a consequence of the voting method. As a voting system, First-Past-The-Post contains spoiler effects wherein a candidate entering a race changes the outcome of the race. Since plurality only lets you choose one candidate, voters are forced to divide their support among similar candidates. This has two effects. The first is that when a ‘spoiler’ candidate takes away some support from a similar candidate, they could change the outcome of the election. The second effect here is that a smart and competent politician self-selects themselves out of the race reducing the number of options for the voters. How is this fair to us or to the excluded candidates?

Plurality voting also makes it easier for voters to vote against their conscience. Let’s say you have a favourite candidate, but because you are convinced he or she won’t win, you decide to vote for someone else so you don’t ‘waste your vote’. This has longer run-off effects in that it tends to favour two-party systems in a phenomenon called Duverger’s Law. Where plurality voting exists, we see weaker political parties subsumed into larger parties—in Nigeria, for instance, All Nigeria People’s Party (ANPP), Action Congress of Nigeria (ACN), All Progressive Grand Alliance (APGA), and Congress for Progressive Change (CPC) all merged to form the All Progressive Congress (APC). And we also see the weak parties leave the political arena as voters gradually desert them because they have no chance of winning and influencing governance even if these weak parties have great policy ideas that could benefit the population.

Recently, Nigeria conducted its presidential election. Muhammadu Buhari, the incumbent, who, from social media, appeared to be considered ineffective because the economy was spiralling down the drain, won with 15 million votes. His main opposing candidate, former vice-president Atiku Abubakar, whom many consider to have hands that frequently wander into the national coffers, came close with 11 million votes. The remaining challenging votes from Omoyele Sowore, Kingsley Moghalu and Fela Durotoye—did not sum up to 100,000 votes. This is curious as many voters often complain about the duopolistic hegemony of Nigeria’s political elite, the All Progressives Congress (APC) and the People’s Democratic Party (PDP). Many desire a third option, yet this desire doesn’t seem to materialize. It is possible that these independent candidates need to spend time building a political base, but the First-Past-The-Post system makes their job significantly harder and more expensive.

Restricting candidate choice is problematic, at least if democracy’s intent is to be believed. It calcifies partisanship and reduces the range of potential solutions to be debated and explored.


It is not enough to consider the demand-side of the political economy, we ought to consider, as Thomas Sowell mentioned, the ‘supply side; that is, to the kinds of people who choose politics as a career and the kinds of policies such people are likely to prefer.’

It could be argued that politicians, more than voters, have a bigger incentive to be knowledgeable and rational. They have to understand how their policies affect their constituents and how that affects their re-election or their party’s stance. For them, the cost of error is high. But this is not sufficient. Even if we granted that politicians were fully knowledgeable and rational, by virtue of their seeking votes from irrational voters, their rationality and clear-headedness become compromised. Caught in a tango where the aspirant follows the lead of the voter, a politician is incentivized to pander to the voter’s desires. This causes a problem of selection. Since voters don’t know how to select competent representatives, the information asymmetry will cause an adverse selection problem where the high-quality candidate options are driven out candidate options are low-quality, effectively resulting in a democracy for lemons. Honest politicians are driven out of the game and charlatans who are willing to dissemble with flattery and lies are left behind.

The bar to run for office is already low. In the USA, to serve as president, one must simply be a natural-born citizen of the United States, at least 35 years old, and a resident of the United States for at least 14 years. American political leaders often come from great educational pedigree, but it is often by incidence. There isn’t a concerted effort to choose leaders based on intellectual skills. As Paul Krugman said in The Great Unraveling, ‘Voters have a visceral dislike for candidates who seem intellectual, let alone try to make the electorate do arithmetic.’ Many of the leaders that we elect do not necessarily have a solid grasp on economics despite that being a significant issue for the voters. They are selected based on their skill to persuade or appeal to the emotions of the voters. Bryan Caplan points out that it is telling that many of the leaders in American politics have a law degree. The current president of the United States had never served in public office, and has had a string of failed businesses, but to his credit, exhibits remarkable curious charm. In Nigeria, there was, and still is, a serious debate as to whether the current President actually completed secondary school, but Nigerians, in 2015, and again in 2019, traumatized by corruption and the possibility of its resurgence, chose to elect a president who appeared honest, if not competent.


Many argue that more democracy is good in at least three ways. Firstly, it is an intrinsic good in and of itself as it fosters autonomy, dignity, inclusiveness, equality, and solidarity which are good in themselves whether or not they yield tangible consequences and as such we must encourage everyone to participate. Secondly, more voting and political engagement by the citizens increases the likelihood of getting effective, stable, representative and fair policies. Lastly, democratic participation serves as a way to enlighten the public and create civic culture.

But it is not immediately apparent that democracy is an intrinsic good, or that if it were, we must deploy it the way we currently do. A lot of Nigeria’s leaders – are appointed not elected. Nigeria’s laws have excluded from voting, the resident alien, the non-resident citizen, and those younger than 18. We seem to think there can be conditions when we offer suffrage even if we think of these individuals as ‘equals’ in humanity or citizenship. Neither is it obvious or necessary that democracies and democracies alone always give the best policies. Lastly, it is an appealing idea that politics can help ennoble the average voter and forge civic bonds, but a glance on ‘political’ Twitter would prove that this is not turning out so well.

So, what to do?

Reducing voter turnout

Perhaps the cure for democracy is not more democracy.

Some political theorists have pushed for properly administered voting tests. They argue drivers and pilots have to take exams before they get their license to drive or fly themselves and others, why can’t this be extended to voting, seeing as voters steer the direction of the country?

I understand the appeal of that suggestion, but it is unfeasible. Not only are the logistical issues intractable, the ethical valence is suspect. We have seen how literacy tests have been used to exclude marginal populations in the past. It is also problematic to make a test mandatory when the political education machinery is defunct. It would also be difficult to carry out in a country like Nigeria, where the literacy rate is about 60 per cent, and there are about 520 languages. Questions such as ‘what exactly is to be tested?’, ‘what means should be used?’, ‘who will monitor?’ amongst others necessarily arise as we try to maintain objectivity and prevent incumbent government capture.

Random Sample election offers a way out. A mix between sortition and democracy, it suggests that we only need a smaller subset of people to vote to represent the majority. Condorcet’s Jury Theorem (CJT) states that if each voter has an independent probability of greater than half of voting for the good choice, then as the size of electorate increases, the probability that the electorate will elect the good choice over the bad choice approaches 1 really quickly. When the conditions of CJT hold, as democracy maximalists argue they do, having more than 100,000 citizens vote bears no impact in actualizing positive political outcomes. Additional voters increase the accuracy, but only marginally, and after a certain number of votes, additional votes don’t add accuracy but increase cost. Once the appropriate sample size is determined, we can randomly select the voters from the general population through a lottery, similar to jury duty selection.

Knowing that your vote counts offers an incentive to be more knowledgeable. We can educate these people with the relevant information—bias sensitivity training, policy statements of the candidates and experts’ opinions on the policies—to better exercise the ability to vote.

Approval voting

Plurality voting is perhaps the worst form of voting. It stifles preferences and excludes potentially great candidates. If we are not going to be defenestrating democracy, the least we could do is make it freer and fairer. One interesting approach is Approval Voting. With approval voting, the voter selects as many candidates as they want, not just one. If they want to choose just one, they can, but they also have the option to select as many as they want. They don’t even have to rank their votes in order of preference. This approach avoids vote-splitting or the choosing of the lesser of two evils. It avoids exclusion of great candidates and gives third party candidates a chance. Approval voting tends to elect the Condorcet winner—a candidate who would win if placed head-to-head with other candidates individually. It is perhaps not a catch all; the Gibbard-Satterthwaite Theorem says that every voting method can have some kind of tactical voting vulnerability, but it seems to be better than the current option in places like Nigeria where the winner is the first past the post.

Limiting political candidates

I am partial to the Singaporean and Chinese approach of political meritocracy. As Daniel Bell put it in The China Model, ‘Cadres are put through a gruelling process of talent selection, and only those with an excellent record of past performance are likely to make it to the highest levels of government. The training process includes programs for public administration at various levels of government, consulting experts, learning from best practices abroad, rotational career postings though different sectors, as well as the cultivation of virtues such as compassion for the disadvantaged by means such as limited periods of work in poor rural areas.’ But this would hardly work in a multi-party state.

Perhaps we could make it a requirement for anyone who wants to run for office to have passed certain exams. In China, public service examinations are effectively intelligence tests that select for those with analytical skills. It also ensures that candidates are cognizant of policy and can analyse complex matters from a variegated perspective. This seems to be at least a feasible way to de-risk the democratic process by making sure the people who run are competent. This means that the choices that the electorate can choose from are reduced to almost fool-proof choices. It is more rational for a person who wants to compete in the elections to meet the competency restrictions than it is for every voter to become competent, because the chances of the candidate winning is far greater than the chances of the voter swaying the election.


It is often assumed that anyone who suggests that democracy needs to be checked or significantly altered is either elitist, overly cynical and pedantic or an autocracy sympathizer. But that is not necessarily true. One can criticize democracy’s penchant for choosing populist leaders who spread nationalism, leaders who discourage free trade or leaders who do not invest in the benefits of citizens, yet unborn, without longing for autocracy. One could even agitate to overhaul democracy as it exists today without being a dictatorship apologist.

I make no pretence to the converse—I certainly lean towards democratic pessimism, at least considering its current form. I have no unqualified reverence for the system per se. I have a predilection for a consequentialist approach to leadership selection—any means is only as good as how it improves the quality of decisions that we make and the lives of the governed. Democracy is just a tool, and a tool is only good if it gets the job done. And if it doesn’t, I have no issues seeking other alternatives.

There is significant opportunity to limit the influence of a biased and ignorant electorate who, through inefficient voting methods, select the worst candidates. It would require limiting the number of people who vote, setting a rigorous standard for those who want to lead and creating effective means of selecting the leaders.

As African countries contemplate their destinies, overthrow autocratic rulers, fumble through the dark on their leader selection process, it is worth considering whether democracy, as currently practised, is really the only feasible option