Who becomes a politician? 

Ernesto Dal Bó, Frederico Finan, Olle Folke, Torsten Persson, and Johanna Rickne│ April 15, 2018
Policy brief written with the assistance of Emma Fernandez

The challenge of a representative democracy

The identity of politicians impacts how they govern, including what priorities they set, policies they select, and how they are implemented. In electing politicians, voters care both about the abilities, or competence, of elected officials, and also that they represent their particular interests. In theory, economic models are uncertain about whether a representative democracy can achieve both high-ability politicians and broad representation. According to economic theory, it is easier for less competent people to enter politics because they have less to give up in their previous employment, or a lower opportunity cost. Thus, scholars have argued that democracies face a trade-off between electing competent politicians and achieving representation, and that electoral systems shape how this trade-off affects voters.

This question has been challenging to study because there is very little data available. Data is rarely collected on those who run for political office but do not win, making it difficult to observe the effects of voter selection. It is also challenging to measure the “quality” or ability of a politician. Previous studies have often used education level or prior income, but these measures can reflect social class or luck rather than ability. A democracy’s representativeness is also hard to quantify. Previous research has used the prior occupations of politicians, but this may be a poor proxy for class background. Our research is able to overcome these challenges by using detailed data about Sweden, and finds that Sweden is able to achieve a representative democracy.

What are the qualities of politicians in Sweden?

The political system of Sweden is based on proportional-representation elections at three administrative levels of government: MPs, county-council members, and municipal-council members. Our research focused on MPs and municipal-council members. Using Sweden’s administrative data, we were able to obtain detailed information about the income, education, and family members of residents. We also used the results of a cognitive and leadership test that is administered by the military to Swedish men. The detail and variety of our data sources overcomes concerns about the ability to measure the qualities of elected officials.

Are politicians competent?

To understand whether politicians were elected for their abilities, we compared the leadership scores, cognition scores, previous earnings, and education of the general population, those who were nominated but not elected for office, and all elected officials, mayors, and MPs. All four of these measures show a similar pattern, with the average scores increasing for each population respectively. The average leadership score for MPs is 70% of a standard deviation higher than the average score of the general population, and MPs’ cognition score is 85% of a standard deviation higher. Mayors’ prior earnings are one standard deviation higher than the average and MPs’ are 1.4 standard deviations higher. Elected officials and mayors have roughly an additional year and a half of education compared to the average, while MPs have about three years more than average. We also compared these results to other “elite” occupations in Sweden, and found that politicians have similar cognitive and leadership scores as CEOs, despite earning a lower salary. These patterns show strong evidence of positive selection – Sweden elected politicians that were, on average, smarter and better leaders than the population they represent.

Are politicians representative?

Theoretically, there are three ways this selection of competent politicians could occur. In a system of “elitism,” wealthy families are more likely to be competent and successful, and have access to power. In this system, electing privileged politicians would accidentally lead to choosing competent ones. In an “exclusive meritocracy,” politics selects competent people, who are more likely to be elitist and not representative. In an “inclusive meritocracy,” politics selects competent politicians who are broadly representative of the general population. To determine which system was at work in Sweden, we compared the ability scores of politicians to their siblings. We found that politicians had higher scores than their siblings, indicating that their individual competence, more than their family background, had influenced their election. Though elected officials overall had high previous incomes, we used the incomes and occupations of politicians’ parents to measure their social class. We found that the social class of politicians’ families was very reflective of the general population. When we compared parental income of politicians to that of doctors and CEOs, we found that doctors and CEOs were more likely to have been raised by high-income parents. This shows that Sweden’s political system is an inclusive meritocracy.

Distribution of Fathers of Politicians and Other Occupations Across Income Percentiles