Much of our political debate is dominated by disagreements about economics, disagreements that seem irresolvable: One side believes Keynesian economics is bunk and embraces supply-side economics; the other side believes supply-side economics has been disproven and instead accepts Keynesian economics. (There’s actually more options than just these two, but bear with me.)
The people who say that Keynesian economics is disproven point out instances where the government has spent large amounts of money on stimulus but the economy hasn’t gotten better. Likewise, the people who say that supply-side economics is disproven point out instances where the government has significantly reduced taxes but the economy hasn’t gotten better.
These are both lousy arguments.
Consider, there are cases where people have been provided life vests, but they’ve still drowned. There are people who have gotten vaccinations for a disease but have then gone on to die of that very disease. There are people who’ve been given antibiotics to stop an infection but have died of the infection, and people who’ve undergone surgery to fix a certain problem and the problem didn’t go away.
Does any of that prove that life vests, lifeboats, vaccinations, antibiotics and surgery don’t work? No, of course not. They’re not guaranteed to work, because there are all sorts of other factors that can intervene to undermine or overwhelm their potency. But, by and large, you’re better off with them than without them, right?
So, pointing out the cases where the economy has remained poor despite Keynesian stimulus or supply-side tax cuts isn’t enough to prove that either economic theory is bunk. Unfortunately, this flawed reasoning is regularly used in political debates when it comes to economics. “George W. Bush and the Republicans cut taxes, yet we had the fiscal crisis and downturn!” “Barack Obama and the Democrats gave us all this fiscal stimulus, yet the recovery is the weaker than other ones where there was little or no stimulus!” These aren’t valid arguments. Just because you know a person who was an overweight smoker but didn’t have heart disease doesn’t prove that smoking and being overweight don’t contribute to heart disease. Anecdotal evidence isn’t valid reasoning, nor is it scientific proof. It’s just hasty generalization.
How do you prove whether these economic theories are right or wrong, then?
Well, the same way you prove whether lifejackets, vaccinations, and antibiotics work: controlled experiments. With vaccinations, for instance, you set up take two groups of people, as similar as possible, and then give one of them the vaccination. If the group that got the vaccination has a significantly lower incidence of disease, then you’ve got a viable vaccination; otherwise, not. (This, for example, is how James Lind showed that scurvy is best treated with citrus fruits such as limes or lemons.) Obviously, you want to run the experiment a few different times, just to be sure, but you get the basic idea.
Controlled experiment is basically how we prove the effectiveness of anything, from fire extinguishers and antibiotics to air bags and bug repellants. So that’s the approach to take for economics.
Unfortunately, it’s very difficult to run controlled experiments — large-scale ones, at least — in economics. How on Earth are you going to set up two economies, each valued at, say, a trillion dollars, identical in every respect except that one of them has a lower tax rate? Or that one of them is engaged in Keynesian stimulus spending and the other isn’t? It’s just not possible. There are so many variables in economics — weather, earthquakes, population changes, geography, natural resources, etc. — that you can’t control and make constant. You can look at existing economies that are adopting similar policies and try to see if they get similar results, and look at ones that are adopting different policies and see if they get different results. But, again, there are so many other variables that will play a role in how their economies perform that it’s difficult to separate the relevant data out from the “noise”. Not that you can’t make headway like this, but it’s much tougher.
And that’s why there’s so much disagreement about economics. Without controlled experiments, it’s very tough to definitively rule things in or out. That doesn’t stop people from acting as if they have the definitive answers to our economic questions, of course. I’m baffled as to how people are endlessly confident when it comes to economic predictions but thoughtfully tentative about sports predictions, despite the fact that sporting events have fewer uncontrolled variables.
But acting confident and having conclusive evidence are two different things. Keep that in mind the next time someone says that Keynesian stimulus or supply-side economics is bunk and has been disproven.