Why data is important

[Heckman] Data essentially allow us to discriminate among alternative explanations. [Piketty] I think theory can be useful but I think sometimes economists spend too much time doing very sophisticated theory without knowing what are the facts that they are trying to explain and understand. [Heckman] When Adam Smith was writing The Wealth of Nations many deep ideas that he had came but he didn’t have a body of data. [Piketty] When I started as a graduate student I realized that there was actually very little data collection. [Heckman] We might observe some fact. Then we want to know well, is this fact an anomaly? [Piketty] So we have to measure these things to make proper comparisons to see, you know, when these claims are justified.

[Heckman] And then the second question would be, even if it’s a repeated finding, what are the mechanisms that give rise to that repeated finding? [Piketty] Better data is not going to make the world a peaceful place but at least it can allow us to have a more informed discussion. [Heckman] As you try to collect data and you try to use every source of information available to you. [Piketty] Go back to the historical data. Collect in a much more systematic manner than what was done before. [Heckman] So I’ve looked at data from what are called cross-sections; looking at different individuals, following the same people over time. [Piketty] Taxation is always more than taxation.

It’s also a way to produce information, you know, to society. [Heckman] So you look at multiple sources of data, all of it non experimental. [Piketty] It’s a way to produce legal categories, statistical categories, which can then be used by economists and other social scientists. [Heckman] And also using economic theory to help organize your thinking. [Piketty] The history of income and wealth is not just a pure economic history; it is a political history, or social history. [Heckman] Parts of the story may not yet be fully understood and that’s the challenge. [Piketty] Instead of just proving sophisticated mathematical theorems in order to impress others we should just, you know, try to collect data, establish facts, and try to learn something.