Culture Change and Chasing Tails

EconTalk Extra
by Amy Willis
Arnold Kling on Specialization... Pedro Domingos on Machine Lear...

EconTalk host Russ Roberts welcomed back one of our favorites this week, Arnold Kling. to talk about his forthcoming book, Specialization and Trade: A Reintroduction to Economics. Kling argues that with its reliance on aggregate models, economics today has lost sight of its most important insight, dating back to Adam Smith. Russ ends the conversation with Kling, an MIT trained economist, ruminating on why they have each departed so far from their early training.


1. What does Kling mean when he says that macroeconomics is not an experimental science, but "a bunch of observations?" How does Kling describe his own odyssey as an economist from his days in grad school at MIT? Have you experienced a similar transformation? Tell us about it...

2. Roberts and Kling argue that entrepreneurs can evade government inertia through enterprises such as Uber and Lyft. What other examples of creative entrepreneurial solutions to government failure can you think of?

We may just have some books to offer in exchange for terrific suggestions!

3. How does Kling suggest our culture ought to change to avoid the perverse incentives of politics? More importantly, how might such cultural change be brought about? Explain.

4. What does Kling mean when he suggests economists ought to "stop talking about the optimal?" What should the role of economists be in terms of public policy?

Comments and Sharing

TWITTER: Follow Russ Roberts @EconTalker

COMMENTS (1 to date)
Warren writes:

1. "Aggregate models (macroeconomics) is not an experimental science, but "a bunch of observations?"

Exactly. The models model someone's economic theory. The models do not model reality, therefore they are not experiments in causes and/or effects of actions in reality. In addition, using large data in these models still does not approach reality because the model does not model reality. Large data in currently created models may report the current situation (of the model), and over timed collection of data may report small deviations from the original. That's the best that can result.

On the planet earth reality is bounded in time and space. Time and space never occur in these models. Consider space as, perhaps, the size of a structural event occurring in an economy. The event is perhaps leather tooling competition. There are X number of leather toolers in the U.S. Now along comes China with x to the fourth toolers whose products, selling at half the cost of U.S. product, swamp the U.S. economy in one fell swoop, not over a period of years. All the U.S. toolers are put out of business at once. No Toolers are ever re-hired, no one learns tooling. The supply of China tooling is never fully absorbed. So, the end of U.S. toolers for ever (Mostly). The structure of the U.S. economy can adjust to the absence of toolers over a period of a few years, but it can never substitute a structurally similar additional type of employment at a similar cost/price.

When all the middle class manufacturing jobs are destroyed, the economy replaces those jobs with highly paid financial jobs. To complete the adaption the U.S. must greatly reduce the number of middle class people. (I'm partially joking.)

Comments for this podcast episode have been closed
Return to top