Democratic systems struggle to thrive among people threatened by disease - or so says a pair of maverick evolutionary psychologists. Jim Giles investigates their controversial claim

Compare these histories. In Britain, democracy evolved steadily over hundreds of years. During the same period, people living in what is now Somalia had many rulers, but almost all deprived them of the chance to vote. It's easy to find other stark contrasts. Citizens of the United States can trace their right to vote back to the end of the 18th century. In Syria, many citizens cannot trace their democratic rights anywhere - they are still waiting for the chance to take part in a meaningful election.

Conventional explanations for the existence of such contrasting political regimes involve factors such as history, geography, and the economic circumstances and culture of the people concerned, to name just a few. But the evolutionary biologist Randy Thornhill has a different idea. He says that the nature of the political system that holds sway in a particular country -- whether it is a repressive dictatorship or a liberal democracy - may be determined in large part by a single factor: the prevalence of infectious disease.

It's an idea that many people will find outrageously simplistic. How can something as complex as political culture be explained by just one environmental factor? Yet nobody has managed to debunk it, and its proponents are coming up with a steady flow of evidence in its favour. "It's rather astonishing, and it could be true," says Carlos Navarrete, a psychologist at the Michigan State University in East Lansing.

Thornhill is no stranger to controversy, having previously co-authored A Natural History of Rape, a book proposing an evolutionary basis for rape. His iconoclastic theory linking disease to politics was inspired in part by observations of how an animal's development and behaviour can respond rapidly to physical dangers in a region, often in unexpected ways. Creatures at high risk of being eaten by predators, for example, often reach sexual maturity at a younger age than genetically similar creatures in a safer environment, and are more likely to breed earlier in their lives. Thornhill wondered whether threats to human lives might have similarly influential consequences to our psychology.

The result is a hypothesis known as the parasite-stress model, which Thornhill developed at the University of New Mexico, Albuquerque, with his colleague Corey Fincher.

Xenophobic instincts

The starting point for Thornhill and Fincher's thinking is a basic human survival instinct: the desire to avoid illness. In a region where disease is rife, they argue, fear of contagion may cause people to avoid outsiders, who may be carrying a strain of infection to which they have no immunity. Such a mindset would tend to make a community as a whole xenophobic, and might also discourage interaction between the various groups within a society - the social classes, for instance - to prevent unnecessary contact that might spread disease. What is more, Thornhill and Fincher argue, it could encourage people to conform to social norms and to respect authority, since adventurous behaviour may flout rules of conduct set in place to prevent contamination.

Taken together, these attitudes would discourage the rich and influential from sharing their wealth and power with those around them, and inhibit the rest of the population from going against the status quo and questioning the authority of those above them. This is clearly not a situation conducive to democracy. When the threat of disease eases, however, these influences no longer hold sway, allowing forces that favour a more democratic social order to come to the fore.

That's the idea, anyway. But where is the evidence?

The team had some initial support from earlier studies that had explored how a fear of disease affects individual attitudes. In 2006, for example, Navarrete found that when people are prompted to think about disgusting objects, such as spoilt food, they become more likely to express nationalistic values and show a greater distrust of foreigners (Evolution and Human Behavior, vol 27, p 270). More recently, a team from Arizona State University in Tempe found that reading about contagious illnesses made people less adventurous and open to new experiences, suggesting that they have become more inward looking and conformist (Psychological Science, vol 21, p 440).

Temporarily shifting individual opinions is one thing, but Thornhill and Fincher needed to show that these same biases could change the social outlook of a whole society. Their starting point for doing so was a description of cultural attitudes called the "collectivist-individualist" scale. At one end of this scale lies the collectivist outlook, in which people place the overall good of society ahead of the freedom of action of the individuals within it. Collectivist societies are often, though not exclusively, characterised by a greater respect for authority - if it's seen as being necessary for the greater good. They also tend to be xenophobic and conformist. At the other end there is the individualist viewpoint, which has more emphasis on openness and freedom for the individual.

Pathogen peril

In 2008, the duo teamed up with Damian Murray and Mark Schaller of the University of British Columbia in Vancouver, Canada, to test the idea that societies with more pathogens would be more collectivist. They rated people in 98 different nations and regions, from Estonia to Ecuador, on the collectivistindividualist scale, using data from questionnaires and studies of linguistic cues that can betray a social outlook. Sure enough, they saw a correlation: the greater the threat of disease in a region, the more collectivist people's attitudes were (Proceedings of the Royal Society B, vol 275, p 1279). The correlation remained even when they controlled for potential confounding factors, such as wealth and urbanisation.

A study soon followed showing similar patterns when comparing US states. In another paper, Murray and Schaller examined a different set of data and showed that cultural differences in one collectivist trait - conformity - correlate strongly with disease prevalence (Personality and Social Psychology Bulletin, vol 37, p 318).

Thornhill and Fincher's next challenge was to find evidence linking disease prevalence not just with cultural attitudes but with the political systems they expected would go with them. To do so, they used a 66-point scale of pathogen prevalence, based on data assembled by the Global Infectious Diseases and Epidemiology Online Network. They then compared their data set with indicators that assess the politics of a country. Democracy is a tough concept to quantify, so the team looked at a few different measures, including the Freedom House Survey, which is based on the subjective judgements of a team of political scientists working for an independent American think tank, and the Index of Democratization, which is based on estimates of voter participation (measured by how much of a population cast their votes and the number of referendums offered to a population) and the amount of competition between political parties.

The team's results, published in 2009, showed that each measure varied strongly with pathogen prevalence, just as their model predicted (Biological Reviews, vol 84, p 113). For example, when considering disease prevalence, Somalia is 22nd on the list of nations, while the UK comes in 177th. The two countries come out at opposite ends of the democratic scale (see "An infectious idea", below).

Importantly, the relationship still holds when you look at historical records of pathogen prevalence. This, together with those early psychological studies of immediate reactions to disease, suggests it is a nation's health driving its political landscape, and not the other way around, according to the team.

Last year, they published a second paper that used more detailed data of the diseases prevalent in each region. They again found that measures of collectivism and democracy correlate with the presence of diseases that are passed from human to human - though not with the prevalence of diseases transmitted directly from animals to humans, like rabies (Evolutionary Psychology, vol 8, p 151). Since collectivist behaviours would be less important for preventing such infections, this finding fits with Thornhill and Fincher's hypothesis.

"Thornhill's work strikes me as interesting and promising," says Ronald Inglehart, a political scientist at the University of Michigan in Ann Arbor who was unaware of it before we contacted him. He notes that it is consistent with his own finding that a society needs to have a degree of economic security before democracy can develop. Perhaps this goes hand in hand with a reduction in disease prevalence to signal the move away from collectivism, he suggests.

Inglehart's comments nevertheless highlights a weakness in the evidence so far assembled in support of the parasite-stress model. An association between disease prevalence and democracy does not necessarily mean that one drives the other. Some other factor may drive both the prevalence of disease in an area and its political system. In their 2009 paper, Thornhill and Fincher managed to eliminate some of the possible "confounders". For example, they showed that neither a country's overall wealth nor the way it is distributed can adequately explain the link between the prevalence of disease there and how democratic it is.

But many other possibilities remain. For example, pathogens tend to be more prevalent in the tropics, so perhaps warmer climates encourage collectivism. Also, many of the nations that score high for disease and low for democracy are in sub-Saharan Africa, and have a history of having been colonised, and of frequent conflict and foreign exploitation since independence. Might the near-constant threat of war better explain that region's autocratic governments? There's also the possibility that education and literacy would have an impact, since better educated people may be more likely to question authority and demand their rights to a democracy. Epidemiologist Valerie Curtis of the London School of Hygiene and Tropical Medicine thinks such factors might be the ones that count, and says the evidence so far does not make the parasite-stress theory any more persuasive than these explanations.

Furthermore, some nations buck the trend altogether. Take the US and Syria, for example: they have sharply contrasting political systems but an almost identical prevalence of disease. Though even the harshest critic of the theory would not expect a perfect correlation, such anomalies require some good explanations.

Also lacking so far in their analysis is a coherent account of how historical changes in the state of public health are linked to political change. If Thornhill's theory is correct, improvements in a nation's health should lead to noticeable changes in social outlook. Evidence consistent with this idea comes from the social revolution of the 1960s in much of western Europe and North America, which involved a shift from collectivist towards individualist thinking. This was preceded by improvements in public health in the years following the second world war - notably the introduction of penicillin, mass vaccination and better malaria control.

There are counter-examples, too. It is not clear whether the opening up of European society during the 18th century was preceded by any major improvements in people's health, for example. Nor is there yet any clear evidence linking the current pro-democracy movements in the Middle East and north Africa to changes in disease prevalence. The theory also predicts that episodes such as the recent worldwide swine-flu epidemic should cause a shift away from democracy and towards authoritarian, collectivist attitudes. Yet as Holly Arrow, a psychologist at the University of Oregon in Eugene, points out, no effect has been recorded.

Mysterious mechanisms

To make the theory stick, Thornhill and his collaborators will also need to provide a mechanism for their proposed link between pathogens and politics. If cultural changes are responsible, young people might learn to avoid disease - and outsiders - from the behaviour of those around them. Alternatively, the reaction could be genetically hard-wired. So far, it has not proved possible to eliminate any of the possible mechanisms. "It's an enormous set of unanswered questions. I expect it will take many years to explore," Schaller says.

One possible genetic explanation involves 5-HTTLPR, a gene that regulates levels of the neurotransmitter serotonin. People carrying the short form of the gene are more likely to be anxious and to be fearful of health risks, relative to those with the long version. These behaviours could be a life-saver if they help people avoid situations that would put them at risk of infection, so it might be expected that the short version of the gene is favoured in parts of the world where disease risk is high. People with the longer version of 5-HTTLPR, on the other hand, tend to have higher levels of serotonin and are therefore more extrovert and more prone to risk-taking. This could bring advantages such as an increased capacity to innovate, so the long form of the gene should be more common in regions relatively free from illness.

That pattern is exactly what neuroscientists Joan Chiao and Katherine Blizinsky at Northwestern University in Evanston, Illinois, have reported in a paper published last year. Significantly, nations where the short version of the gene is more common also tend to have more collectivist attitudes (Proceedings of the Royal Society B, vol 277, p 529).

It is only tentative evidence, and some doubt that Chiao and Blizinsky's findings are robust enough to support their conclusions (Proceedings of the Royal Society B, vol 278, p 329). But if the result pans out with further research, it suggests the behaviours involved in the parasite-stress model may be deeply ingrained in our genetic make-up, providing a hurdle to more rapid political change in certain areas. While no one is saying that groups with a higher proportion of short versions of the gene will never develop a democracy, the possibility that some societies are more genetically predisposed to it than others is nevertheless an uncomfortable idea to contemplate.

Should the biases turn out to be more temporary - if flexible psychological reactions to threat, or cultural learning, are the more important mechanisms - the debate might turn to potential implications of the theory. Projects aiming to improve medical care in poor countries might also lead a move to more democratic and open governments, for example, giving western governments another incentive to fund these schemes. "The way to develop a region is to emancipate it from parasites," says Thornhill.

Remarks like that seem certain to attract flak. Curtis, for instance, bristled a little when New Scientist put the idea to her, pointing out that the immediate threat to human life is a pressing enough reason to be concerned about infectious disease.

Thornhill still has a huge amount of work ahead of him if he is to provide a convincing case that will assuage all of these doubts. In the meantime, his experience following publication of A Natural History of Rape has left him prepared for a hostile reception. "I had threats by email and phone," he recalls. "You're sometimes going to hurt people's feelings. I consider it all in a day's work."

An infectious idea

Jim Giles

Jim Giles is a New Scientist correspondent based in San Francisco

Healthy Democracy