AP Comparative Government and Politics

What comparative political scientists do

Comparative politics sets the political systems of different countries side by side so that patterns can be seen that no single country would show. Studying one country tells you how it works; comparing several tells you which features are shared, which are unusual, and which conditions seem to go with which outcomes. Six countries are compared throughout: the United Kingdom, Mexico, Nigeria, Russia, Iran and China. They were chosen because they differ in regime type, in how power is divided, in economic position and in history, so that almost any concept can be seen working in more than one setting.

Political scientists build their conclusions from evidence. Quantitative evidence is numerical: tables, charts, graphs, maps and indices. Qualitative evidence is everything else that can be analyzed: speeches, foundational documents such as constitutions, political cartoons, commentary and news reports. Both kinds are used to make comparisons between countries and to draw inferences about how institutions and people behave.

Empirical and normative statements

An empirical statement is a claim about what is the case, which can in principle be checked against facts: "Russia's president is elected for a six-year term" or "Mexico's Chamber of Deputies has 500 members". A normative statement is a claim about what ought to be the case, and it rests on values: "executives should be limited to two terms" or "a free press is more important than public order". Comparative research needs the two kept apart. A normative claim can be argued well or badly, but it cannot be proved by data, whereas an empirical claim can be tested.

Most useful analysis combines them with care. A writer who argues that term limits improve democracy is making a normative claim, and the evidence offered for it (for example, how often executives who were free to stay on changed the rules to keep power) is empirical. Reading an author's argument means finding the claim, the evidence offered for it, and whether the evidence really supports the claim.

Correlation and causation

A correlation exists when two or more variables are associated: as one changes, the other tends to change as well. Higher national income and higher levels of democracy, for example, tend to appear together across many countries. A correlation does not show that one thing causes the other. The link could run the other way, or a third factor could produce both, or the relationship could be chance in a small group of cases.

Causation is hard to establish in comparative politics because outcomes such as regime stability or the content of a policy usually have many possible influences at once, and there is no way to hold all but one of them fixed and see what changes. Political scientists therefore speak carefully: a pattern is consistent with an explanation, or suggests it, and a strong claim of cause needs evidence beyond the association itself, such as a plausible mechanism and cases that fit it.

Empirical information is used for three main purposes: applying a concept to a case (showing that a country fits the idea of a rentier state), supporting a generalization (that proportional representation tends to produce more parties), and making an argument. Comparing different systems lets researchers draw conclusions about politics in general and not only about one state.

Describing and explaining data

Working with data moves through steps of increasing difficulty.

1. Describe the data presented: say exactly what the table or chart shows, with the units and the countries or years it covers.
2. Describe a pattern or trend: whether a value is rising, falling or steady, and how countries compare with each other.
3. Explain the pattern and draw a conclusion about politics: for instance, that fast growth alongside rising inequality may help explain protest.
4. Say what the data implies about a political system, institution or behavior, and what it cannot show.
5. State limitations: the data may be old, may cover only some years, may be self-reported by governments with a reason to flatter themselves, may measure something only roughly, or may leave out the cause that matters most.

A figure that is accurate can still mislead if it is read without its limits. An average income, for example, says nothing about how unevenly that income is shared.

Standard data sources

Several standard sources are used to investigate relationships among the six countries.

Each has limits worth knowing. GDP counts output, not wellbeing, and ignores unpaid work and environmental damage. The HDI leaves out inequality within a country. The Gini index shows how unequal a distribution is but not how high the average is, so a poor country and a rich one can share a Gini value. Freedom House and Transparency International rest on expert and survey judgments, which are open to dispute. A corruption index based on perceptions can lag behind real change.

Worked illustration of reading a comparison

Suppose a table gives the GDP per capita and Gini value for two countries, and the first has the higher GDP per capita and the higher Gini value. A careful description says the first country is richer on average and also more unequal. A conclusion about politics could be that income is more concentrated there, which may raise pressure on government for redistribution, though the table alone does not prove that. A limitation is that a single year was shown, so nothing is known about the trend.

The same discipline applies to qualitative sources. For a speech or an editorial, identify the claim being made, the evidence given, the reasoning that connects them, and the author's position. Then ask how the argument relates to an institution or behavior in a country: for example, an author who argues that a legislature is merely a rubber stamp is making a claim about how independent the legislature is, and it can be tested against what the country's institutions are known to do.