Removal of the dividend tax would: Lift stock prices – since the “yield” is now higher. This could be classified as a tax-cut for the rich since it increases the company owners net-worth. More companies increase dividend or start paying a dividend – this effect is probably easy to check. Stable, large companies paying a dividend were known as “widow-orphan” stocks. Maybe, the grandma part is a reference to pocketing dividend from these companies.

but does more math reveal the disconnect? sometimes.

I agree. And I like the cautious “sometimes”.

they aren’t married to the fantasy of efficient market theories or gaussian distribution of market returns.

Again, I agree. However, I must mention that it was rather depressing to look at the literature from the early days of quantitative finance (lot of it of non-math-dept origin); especially its accent on gaussian distribution of market returns! I mean – something as mutually-non-independent as a slew of sell orders in a small span of time was approached from a law-of-large-numbers perspective back in those days!!

Yeah, I’m with ups @36 on this one. This seems like a combination of a strawman and some general misdirection. As far as I remember, no one argued that the dividend tax cuts would result in increased executive compensation (except insofar as they have a lot of stock and thus receive more in dividends) [there's the straw man]. Instead, my understanding was that it would do exactly what Chetty discovered: return more money to people in the form of dividends. The question is: who was getting the dividends? If most of the dividend earners were rich, this still amounts to a tax cut for the rich, regardless of how many of them are also grandmas [there's the misdirection].

it’s pleasantly ironic that Kahneman and Tversky, who won the most deserving of Nobel’s, were brilliant in unraveling some of the disconnect, making economics more accountable and creating space for more empirical approaches; but, in doing so, their results were simultaneously and decidedly falsifiable. that is the peculiar essence of economics- more math on top of a disconnect is rubbish; more math on top of a ‘connect’ is good. but does more math reveal the disconnect? sometimes.

in my experience, the quants at hedge funds are the best practitioners in highly mathy, yet reality-based outlook on markets. they aren’t married to the fantasy of efficient market theories or gaussian distribution of market returns.

I have been humbled by many outcomes that did not follow the proven empirical theories of consumer behavior.

Thanks, Floridian, for the above observation and the further contextualisation in your comment. The sense in which many economists, and most sales-and-marketing “theorists” use the word “empirical” is what provoked my parenthetical comment in #38. The disconnect between theory and real-life outcomes that you’ve witnessed are, of course, largely a result of the sheer complexity that economics tries to model and/or explain, and of the fact that social experiments are difficult. But these disconnects occasionally make me wonder if economists need to acquire more sophistication in how they use their data sets (after all, empiricism is a lot more than the mere acquisition of data points a la the “research” of the gentlemen-naturalists). When one looks at the severe hypothesis-testing that biologists typically perform nowadays, economics sometimes feels like it is stuck in the gentlemen-naturalists’ era of empirical science.

I think Chetty is very aware of all this, but I don’t think his “. . . better match empirical evidence” is an immodest declaration that he is an empiricist and others are not (as you seemed to interpret it), OR that he has found a way of turning economics into a hard-empirical discipline such as, say, experimental chemistry. I think those words are merely a gloss of his methodological outlook.

“Fields of science are commonly classified along two major lines:

* Natural sciences (e.g. physics, chemistry), which study natural phenomena (including biological life), and
* Social sciences (e.g. economics), which study human behavior and societies.

(e.g. inserted by Floridian)

These groupings are empirical sciences, which means the knowledge must be based on observable phenomena and capable of being experimented for its validity by other researchers working under the same conditions.[4]

Mathematics, which is sometimes classified within a third group of science called formal science, has both similarities and differences with the natural and social sciences.[3] It is similar to empirical sciences in that it involves an objective, careful and systematic study of an area of knowledge; it is different because of its method of verifying its knowledge, using a priori rather than empirical methods.”

Sorry I lost the link in my effort to copy-and-paste. In a nutshell, I see empirical and a priori methods as the two extremes with economics using the former. Sometimes I even question if economics is indeed a science, social or not. Having spent most of my life in marketing, I have been humbled by many outcomes that did not follow the proven empirical theories of consumer behavior. Since economists eat from the same plate, I would think that marketing’s frequent non-replicability, a horribly unscientific condition, applies to them as it does to us lowly marketers.

Again, neither my remark on empiricism nor non-replicability is meant to be a criticism of economics.

But for someone to make a claim of empiricism in a discipline that is nothing if not empirical. . .

I will have to go with comment #37 by random, Floridian. (Going off on a tangent: there are people in the hard sciences who will justifiably take issue with your assertion that economics is “nothing if not empirical”; a hard-core experimental chemist, for instance, might get quite a chuckle out of it.)

So what could Chetty’s “. . . better match empirical evidence” mean. Let’s look at how economic models are crafted. A model mathematically encodes certain assumptions in such a way that that the empirical data is predicted by it. However, a model is no good if it is so complex that one cannot determine how it evolves in time. Therefore, models are kept faily simple, and aspects of the empirical data are often: a) declared as lying outside the regime of the model; OR b) shown to be a “small perturbation” to the model. Not infrequently, a model turns out to be a bad fit with the data because a shoddy job was done in establishing (b), or — more commonly — because, in the absence of the analytical tools that would make a complex model grow legs and run, the economist opted for a simpler model.

From the few economists I’ve been talking to, I infer that the use of modern mathematics is seeing a minor renaissance in economics. This has conferred on economists greater technical power to eschew option (a) and grapple with more complex models. The complexity confers a better match with empirical data . . . and I think this is the sense in which the phrase “better match” is meant.

The 7-page appendix to the paper Vinod linked to illustrates a part of my point. It is still not very hard, but it is a sea change from the cognitive style of economics papers from the 1980′s.

Neither do I. Floridian’s interpretation seems like reading a bit too much into it.

Floridian, are you saying that economists do not study abstract theory? (If your answer is no, then his point with that sentence in his research statemtn might be that he is a bit more on the applied side of the spectrum…)

What are shareholder grandmas? That’s really arguing at an emotional level. These are probably rich people to begin with. Income inequality is higher now than any time since the gilded age.