Schultz: Do democratic institutions constrain or inform

Schultz. 1999. Do democratic institutions constrain or inform: Contrasting two institutional perspectives on democratic peace. International Organization 53 (spring): 233-266.

Two Competing Theories

Schultz identifies two competing arguments that attempt to explain the democratic peace.

1. Institutional: Democratic institutions raise the political risks of war by holding leaders accountable for their actions, so democracies should go to war less often (in general).
2. Informational: The relatively transparent policy-making process used in democracies allows democracies to make more credible threats during crises. Democracies fight one another less because it is easier to avoid the information imbalances that contribute to bargaining breakdowns.

Schultz likes the informational explanation; for an example of the institutional explanation, see Bueno de Mesquita et al 1999.

Empirical Test

These two theories produce two competing hypotheses. Although both arguments explain the same outcome (democratic peace), they yield competing hypotheses about how the other country in a crisis will react. Specifically, if state A is a democracy and it threatens state B, then:

• H1 (Institutional hypothesis): State B will not believe that the threat is credible. After all, A's leader would face political risks if he carried out his threats of war. So State B will resist the threat and escalate the crisis.
• H2 (Informational hypothesis:) State B will believe that the threat is credible. State B will therefore be less likely to resist the threat (and escalate the crisis).

Using Correlates of War (COW) and Polity data, Schultz finds significant and substantial support for the informational hypothesis.

Threats to Validity

Schultz's empirical work may suffer from severe selection bias. If H1 is correct, then states would only pick fights that they know they can win. Recognizing that an attack is imminent, state B might reasonably try to make concessions. If this is true, then the institutional explanation (H1) would yield the same result as the informational explanation (H2): concessions. (Indeed, Bueno de Mesquita et al 1999 argue that democracies act in this manner; they only pick fights that they can win. They also argue for the institutional explanation.)

Schultz acknowledges this problem and even attempts to control for it; nonetheless, he admits that it makes his entire analysis inconclusive.