Talk to Me Like I’m Stupid: “The sky is (not) falling!” Edition

Mark Thompson in reply to Glyph says:

Like Glyph, I don’t have much of a dog in the specific fight that is the subject of the post. I will, however, say that the study of economics, in and of itself, tells us nothing – and purports to tell us nothing – about normative questions. In other words, whether the divergence between wages and average productivity (as opposed to marginal productivity) is a “problem” is entirely a normative question that economics does not purport to answer. Individual people with economics backgrounds may certainly claim that it is not a “problem” but they can only do so by referencing other normative values that others may weigh less than they. Indeed, there are surely plenty of economists that would view this divergence as a “problem” worthy of a solution (lest we forget, Paul Krugman is an economist, as was Milton Keynes, etc.).

So with respect to your original post, while economics teaches us that, all other things being equal, wages will track marginal productivity rather than average productivity, it does not teach us that they should do so. As importantly, it does not teach us that all other things are equal – in fact, it recognizes that all other things are not in fact equal, and thus that it is possible for even marginal productivity to diverge from wages.

Economics can then give us an understanding of why average and marginal productivity are diverging from each other and why one or both of them are diverging from wages. It can also give us a possible solution to that problem. It can also give us a range of possibilities for second order effects of that solution.

It is for policymakers and voters to determine how to weigh the likelihood of a proposed policy solving the problem against the likelihood of the possible second order effects, not only in terms of how likely those effects are, but most importantly in terms of how important those effects are as a normative matter.

But the key thing about economics is that it typically begins with the assumption of “all other things being equal” except perhaps for a small number of known variables. This means that its predictive value is very limited. It’s probably an overstatement to say that economists don’t do predictions; however, it’s not an overstatement to say that they don’t do predictions outside a limited range.

In other words, they can say that “policy type X” will tend to have effects “Y and Z,” but they can’t do much in the way of predicting the full extent of those effects and the specific time frame of those effects. About the most they can say is that a policy will have certain general effects in a vaguely defined “short run” and certain other effects in the vaguely defined “long run.” One of the debates between economists of different ideological stripes is often between how much to value the “short run” and how much to value the “long run” as a normative matter.

At this point, as I say in my comment above, we have a very limited amount of knowledge, but that is a far cry from saying that we have no knowledge. It is far from the point that physics or astronomy has reached, in which there is quite a bit that can be predicted (but also still quite a bit that cannot be predicted – otherwise we’d have no further need for those areas of study).

In economics, there are some things that can be predicted. But we can’t predict that one policy change will have a specific, measurable effect outside of a particular range of activities. There are just too many damn unknown and unknowable variables – the so-called “calculation problem.”

There are also a good number of things that all or virtually all economists agree upon – for instance, that price ceilings (e.g., rent control) will create significant and artificial supply shortages of those products as compared to the demand for those products. Or that allowing competition between suppliers in a given market will increase consumer surplus in that market: http://www.igmchicago.org/igm-economic-experts-panel/poll-results?SurveyID=SV_eyDrhnya7vAPrX7

Etc., etc.

But none of that tells us whether the other effects of those policies are good, nor does it much tell us about what the unintended consequences (or benefits) of those policies might be. Hell, it doesn’t even tell us whether the primary objects of those policies is a normative good, although it is certainly safe to assume that most people would think that in most situations and all other things being equal, a greater supply of affordable housing is normatively good as is a greater consumer surplus (since everyone is a consumer).

We can predict at least some vaguely defined unintended consequences as well – for instance, we can predict that price ceilings and floors will result in black markets or will result in people trying to find ways of otherwise evading those ceilings and floors – for instance, it should have been entirely predictable that the result of a generalized wage freeze would be that at least some workers would be paid under the table for wages above or below the ceiling (black market) and that firms would try to find other forms of compensation and benefits to compete with each other for labor. What would have perhaps been comparatively unpredictable would have been the full extent of those effects, particularly on the black market side.

Similarly, it is almost impossible, based on what is currently known and knowable, to predict with any kind of specificity and reliability how a broad-based measure such as GDP or unemployment will look a year from now, much less a decade from now.* For the most part, we cannot generally forecast that a particular policy will cause a widespread recession within a short period of time – there are just too many other variables at play. We can say that the policy makes a recession somewhat more or less likely, and we can aggregate policies as a whole to say that they are likely to cause a recession – or even economic ruin – at some unknown point down the line.

For instance, we can predict that economy-wide wage and price ceilings (floors are usually less problematic) will eventually cause economic ruin, but we can’t predict exactly when that will happen – the country, for instance, may experience a natural resources boom (I’m looking at you, Venezuela) that creates enough wealth to mitigate these effects for an extended period of time through black markets, subsidized imports of controlled goods, etc. Or commodities prices in the global market may remain unexpectedly stable for an extended period of time such that the price controls are relatively close to the “invisible hand” price. But eventually you’re going to wind up with a drastic undersupply of needed goods and workers, an oversupply of shit no one wants or needs and, without economic reforms, you’re going to need a pretty totalitarian government to keep things afloat.

So it would be foolish to predict that, say, “within five years of instituting these price controls/regulations, the economy will collapse and GDP will contract by at least X%.” It would equally be foolish to predict that “within five years of instituting these tax cuts, the economy will boom and GDP will increase by at least X%.” Now there may or may not be some predictive value in aggregating such predictions, but no one person’s or ideological group’s formula is likely to be consistently reliable – there’s just too many variables for any formula or group of like-minded people to account for. And that’s before we even get to the issue of true “Black Swan” events.

So we’re never going to have perfect knowledge in the field of economics, nor would we want to, frankly – such perfect knowledge would mean that people had become automotons whose needs and desires never changed or were at least entirely predictable. Indeed, in just about every instance where a true centrally planned economy was attempted (ie, an economy in which it was assumed that economists could at least approach perfect knowledge), there has been an attempt to turn people into exactly that through pretty totalitarian measures.

But that hardly makes economics useless – that we can make vague but narrow predictions is certainly valuable to informing policy debates, and as time goes on and knowledge increases, those predictions can eventually become more specific and less narrow. As importantly, economics can give us an idea of how, if at all possible, to achieve specific policy goals without undermining other specific policy goals. Economists just shouldn’t be expected to define what those goals should be.

*One thing I find interesting in this thread, by the way, is that it’s the more liberal commenters who have tended to take the harshest stand against economics even though it is liberal economists who are vastly more likely to view the economy as predictable and manageable.Report

Mark Thompson in reply to Glyph says:

@michael-drew No, I do not.Report