New Nobelists Duflo and Bannerjee in the NYT:
[E]conomists have somehow managed to hide in plain sight an enormously consequential finding from their research: Financial incentives are nowhere near as powerful as they are usually assumed to be.
David’s critique is excellent. But supposing Duflo and Bannerjee are right, I have a few questions.
If incentives are overrated…
1. Is it OK to reduce penalties for tax evasion?
2. Is it OK to cut regulatory enforcement?
3. Is it OK to curtail the right to sue?
4. Is it OK to cut pay and benefits for government employees?
My underlying point, as I’ve said before: The connection between heterodox economics and left-wing policies is fragile at best.
Yes, “incentives matter” helps the case for some right-wing policies. But “incentives matter” also helps the case for some left-wing policies. If you think textbook economics is misleading – as I often do – you should do a full rethinking of your policy views. Don’t just single out the policies you dislike because they rest on questionable assumptions. Scrupulously investigate the implications for policies you like, too.
What if the heterodox assumptions turn out, on balance, to be politically inconvenient for you?
Perhaps you went heterodox a little too hastily.
Or perhaps reality just isn’t politically convenient.
Update: See Scott Alexander’s superior post on the same theme.
READER COMMENTS
Patrick
Nov 20 2019 at 11:51am
If it’s true that financial incentives matter less than economist assume, I’d answer these questions as follows:
1. Is it OK to reduce penalties for tax evasion?
If other penalties are more effective such as jail time / public embarrassment are more effective then possibly yes
2. Is it OK to cut regulatory enforcement?
IF the financial incentives are less effective than other means yes
3. Is it OK to curtail the right to sue?
It depends on the circumstance. For example, if punitive damages are awarded to punish and are generally ineffective (defendant has already been made whole) – potentially criminal charges / jail time is more effective than punitive damages.
4. Is it OK to cut pay and benefits for government employees?
Problem with government employees is that there’s no market and so market rate… I work in the private sector so I have no idea what the “market rate” would be for a bureaucrat. If a company thought it could save money on labor, it would cut labor cost and see if it hurts the bottom line…. So yes.
Nick
Nov 20 2019 at 12:18pm
I don’t have an account on NYT so can’t see what was written but from the one line summary the authors chose to expose on the top of the page, I’m guessing it’s a lot of resentment of markets.
Fine. Markets are neither efficient nor just. If your complaint is that markets fail, you should be standing on rooftops and announcing at your full voice, “Governments are WORSE!!!”.
Why? Any and all criticisms you have of markets are almost EQUALLY, if not more applicable to governments.
If for instance, you find that, in a particular society, markets are undersupplying a particular good i.e. price is high and people are suffering as a result. What solution does the average social democrat/leftist/big government guy offer that is better than a laissez faire libertarian approach?
Nothing. It’s as “bad” as doing nothing. Calling the libertarian approach “bad” is just for the sake of argumentation, because doing nothing would probably not have caused that shortage.
Why? Governments around the world fail to do basic common sense actions. There is no way a Social Democrat can ‘protest’ these things into existence. And that spells the death of any and all ideas that big government ideologues have. Yes, they are ideologues and not “practical”, because the only reliable human behavior that can be taken for granted is the self interested behavior.
One can therefore safely discard all big Government plans unless there is a straightforward plan of action for implementation, that does NOT rely on people behaving generously and protesting ideas into existence.
Dennis
Nov 20 2019 at 1:47pm
1) NYT: financial incentives are not as powerful for work as economists claim
2) ….
3) This blog: if incentives don’t matter than we should remove penalties for crime as it will have no impact.
There is a missing step 2 in this logic chain.
Only the last question 4) cutting government employee salaries is relevant. But the question is worded to conflate issues. “Is it OK to …” is a moral question. “Would cutting wages of government employees have an impact on employment in the public sector?” would be more relevant. Duflo and Bannerjee would say for that the impact would be small.
Erwin Dekker
Nov 20 2019 at 2:49pm
Yes, this was precisely my feeling when she was giving a public lecture here in Rotterdam recently. On the one hand she was claiming that financial incentives were not that powerful and people were (irrationally) stuck to place and habits.
On the other hand she was saying how smart incentive schemes could move people from rural to urban areas. And how habits were easy to change. Well show me, if incentives are not the way.
Thomas Hutcheson
Nov 20 2019 at 8:15pm
I think it works the other way, too, that economic orthodoxy leads to quite non-libertarian views that admittedly not all Left Wingers have adopted yet: like progressive consumption taxes, carbon and other Pigou taxes, higher EITC instead of minimum wages, increasing immigration up to the point that the marginal immigrant no longer adds to he income of existing residents, greater freedom of trade and investment, regulation and law enforcement guided by cost benefit considerations.
Thomas Sewell
Nov 20 2019 at 10:50pm
I would count the end of your list:
… as typical libertarian views. Someone might quibble that they aren’t perfect compared to an ideal vs. compared to the status quo, but if proposed compared to current U.S. government policies/laws, I’d put them on the libertarian end of the spectrum. What makes them “quite non-libertarian” in your experience?
Thomas Hutcheson
Nov 21 2019 at 8:42am
Perhaps a mis-perception on my part. I “hear” regulation discussed pretty often around here as less is better with an implication that zero is best which is guided by cost-benefit analysis only if benefits are zero.
robc
Nov 21 2019 at 1:32pm
My utilitarian (I am not one) theory on regulation is based on an interview in the 90s with Ronald Coase (Reason mag), in that when he looked at regulations trying to sort the good from the bad ones, he couldnt find any with a positive cost/benefit analysis (amongst the new regulations). My theory is we have passed a point of diminishing returns where every new regulation, no matter how good it is in principle, is net negative due to the total number of regulations.
If we would roll back enough regulations (and lets do it with a chainsaw, as no need to pick and choose if I am correct) far enough, we would then be able to select the good regulations to reapply. I think start with eliminating 90% at random, then work from there.
Or, sunset laws might work too. No group reapproving, they need to be carefully discussed one by one and slowly reapplied.
Thomas Sewell
Nov 21 2019 at 10:40pm
I guess it’s a direction rather than an end goal difference, perhaps?
In other words, I think many libertarians would see those policies as in the right direction from the current less justified norm, even if they believe it would be better to go even farther toward leaving decisions to individuals.
Mark Z
Nov 21 2019 at 12:14am
They argue specifically that social incentives matter much more than financial ones. If that’s so, maybe one of the best social weapons we could deploy against poverty and criminality is the intense stigmatization of having children out of wedlock; and stigmatizing deviation from productivity-enhancing social norms in general. Far from warranting a general rolling back of ‘shaming’ and stigmatizing behaviors and choices, we ought to be far more judgmental of each other’s choices in this case, for the good of society. So yeah, I can see how financial incentives not mattering leading to unlibertarian conclusions, but I think it largely leads back to traditional moralism and paternalism.
Phil H
Nov 21 2019 at 12:10pm
I basically agree with what Caplan says, but I will offer one small quibble, and that’s the nature of economic organizations, e.g. companies.
The B&J argument is that social factors affect people much more than financial incentives. But an awful lot of the economic behaviour of the wealthy is not actually the behaviour of people; it’s the behaviour of companies. Companies, I would suggest, are relatively insulated from social pressures, but respond, as a group, very sharply to financial incentives.
Here’s an anecdotal example from the UK. A comedian named Jimmy Carr was outed in the newspaper for having carried out legal tax avoidance. He’d set up a company on a small tax haven, and had his income registered to that company, even though he’s not a resident. The newspapers gave him some grief, and he publicly apologised for doing it (and I assume, stopped doing it).
But of course, *he* didn’t do any of the things I mentioned. His accountant did. He’s very successful, so he has professionals to manage many parts of his life, including his finances. And as specialists in their own fields, each professional just does what they are supposed to do: his accountant found a way to save him some money, without thinking about the moral hazard, because that’s not what accountants are paid to think about.
So to the extent that professional accountants and hedge fund managers and lawyers get involved, financial incentives may make very large differences; but for those people who don’t generally use use professional services, the impact of social incentives may be much greater.
Mark S
Nov 22 2019 at 2:20pm
Esther Duflo seems to measure school attendance (not literacy or numeracy) after the deworming intervention. I want to know Bryan’s take on this measurement as someone who studied education and poverty extensively.
Comments are closed.