What Patterns of Trade Might Tell us About the Democratic Peace

11 February 2013, 1325 EST

Stop me if you’ve heard this one: it appears that wars between pairs of democracies are relatively rare compared to wars between other pairs of states.  Some people even think this relationship might be causal.

In the decades since this empirical regularity first got everyone’s attention, a number of different theoretical arguments have been developed to account for it.  A little known problem with all such explanations is that their additional observable implications are at odds with empirical observation.  As I mentioned when Dan interviewed me for a Duck podcast, every major explanation of the democratic peace either leads us to expect a disproportionate number of wars between pairs of autocracies (which we don’t see) or, if it can get around that problem, leads us to expect democracies to be disproportionately likely to lose the wars they fight (whereas it seems they actually win more often than other types of regimes).*

One possibility is that some or all of the empirical findings are statistical artifacts.  Our explanations for the democratic peace may not be squaring with observation because they are attempts to explain a non-existent phenomenon.  Another possibility is that the additional empirical regularities our theoretical arguments appear to contradict are statistical artifacts while the pacifying effect of joint democracy is not.  And there are valid criticisms to be made about the empirical findings (see here and here).

But let’s set that aside for now.

Though it hasn’t received quite as much attention, there’s a tendency for democracies to trade with one another more often than other pairs of states.  The dramatic increase in economic cooperation that began in the 1980’s tracks closely with the increase in the number of democracies in the world.

The most common explanations for this mirror the most common explanations for the democratic peace to a considerable degree.  I’m simplifying a little — but not too much — when I say that our two biggest explanations for why democracies trade with one another are:

1.  Trade is good, and democracies do that which is good;

and

2. Cooperation is often inhibited by a lack of information, and democracies are transparent.

Similarly, it is only a slight simplification to say that our two biggest explanations for why democracies do not fight wars with one another are:

1. Fighting (or perhaps losing) wars is bad, and democracies do that which is good;

and

2. War is often caused by a lack of information, and democracies are transparent.

These explanations imply that pairs of democracies will be especially likely to trade with one another (and unlikely to war with one another).  That is, they provide an explanation for that which we already believed to be true.  So, you know, good job there.  They also lead us to expect pairs of autocracies to be especially unlikely to trade with one another (and likely to war with one another).  We already know that autocracies are not particularly likely to war with one another.  Are they unlikely to trade with one another?

To the best of my (limited) knowledge, no one has even checked to see.  There have been many studies of trade openness that include linear measures of democracy and have found democracy to be positively associated with trade.  But that’s not the right way to do it.  If autocracies trade with one another more than they trade with democracies, but not quite as much as democracies trade with democracies, then we’d find exactly what we find.  If autocracies trade with democracies more often than autocracies, but do not trade with anyone as much as democracies trade with democracies, we’d also expect to find exactly what we find.  That simply isn’t the right specification.  What I’d love to see is what happens when you use the research design from this article to study trade rather than militarized conflict (while also including the standard “gravity” measure).

If we find that autocracies trade with one another more often than they do with democracies, that tells us that our understanding of why democracies trade with one another is as problematic as our understanding of why democracies don’t fight one another.  Moving forward, we’d want to focus our attention on developing a unified theory of why countries of similar regime types get along with one another so well.  It wouldn’t be so important to differentiate joint democracy from joint autocracy, nor to differentiate conflict from cooperation.  What would be important would be to generate some additional observable implications — to make sure we developed something more than a “just so” story.

If we find that autocracies trade with democracies more than they do autocracies, even after accounting for the effects of “gravity”, that would tell us that economic cooperation is not merely the flip side of conflict.  It would therefore be problematic that we essentially tell the same stories when theorizing about the effects of democracy on these two different processes.  It might, however, mean that existing theoretical accounts of the relationship between democracy and trade are on the right track.  Moving forward, we’d want to focus our attention on figuring out why autocracies have peaceful ties with one another but do not trade with one another, recognizing that our current attempts to explain the democratic peace are probably in greater need of revision than our current attempts to explain why democracies trade with one another.

Anyone feel like looking into this? :)

 

*Eventually, I’m going to write a paper proving this, but I have to clear a few other items off my to-do list first.