Cutting Poverty Is Super Easy: A Response to Sumner

I am not really sure what Scott Sumner is all about these days. Many years ago, he was like “monetary policy should utilize an NGDP target” and people were like “that’s an interesting thought.” But now, he’s kind of gone into mission creep mode where he comments on things that he’s not so knowledgeable on.

One of the more glaring versions of this creep is his armchair commenting on wealth inequality. Again and again, he has called wealth inequality data “nonsense on stilts” because it ignores the fact that wealth inequality is just a life-cycle phenomenon. This is straightforwardly false, but to know it’s false, you have to actually be familiar with the wealth data and ambitious enough to run some age-controlled wealth inequality calculations. Sumner is neither of those things.

More recently, Sumner’s mission creep has him opining rather strangely about poverty in America, with a focus on yours truly. The post is such a complete mess that I will utilize a line-by-line approach to explaining where it has gone wrong.

Dynamic Effects
The primary thrust of Sumner’s post is to say that, because I do not have a libertarian-wired brain (read: smart brain), it doesn’t occur to me that poverty-reducing transfer programs can have dynamic effects that also affect the market poverty rate. Although Sumner’s psychoanalysis is creative, it sadly misses the mark. Here I am in 2013, with my progressive-wired brain, somehow getting my head around the idea of dynamic effects that heretofore only libertarians have understood:

Of course, if you eliminate Social Security, some old folks might work longer to avoid living on cat food. Also, some might save more for their old age. So the world without Social Security probably wouldn’t mean exactly 22 million extra impoverished people. But we know, as a historical matter, that life before the more generous Social Security pension we now have was rife with elderly poverty. In 1960, the elderly poverty rate was 35 percent. With the expansion of Social Security however, that rate plunged to 10 percent by 1995, eradicating 71 percent of elderly poverty. For 2012, I calculated that the over-65 poverty rate was 9.1 percent. Subtracting Social Security, it would be 44.4 percent (a difference of more than 15 million seniors).

I note that the existence of Social Security can change the composition of market elderly poverty. But then I also note that historical elderly poverty data show that the pre-robust-SS era had enormous elderly poverty, which suggests (as best as any data on this question can) that the dynamic effects of Social Security on market elderly poverty are not that big. It’s almost as if my inferior mind has not failed to apprehend this point, but that I’ve apprehended it and then worked directly with the data to work out how significant it is in reality and determined that the answer is: not very.

Using market poverty (sometimes called pre-tax, pre-transfer poverty) as a comparator for disposable income poverty (sometimes called post-tax, post-transfer poverty) is extremely common in comparative welfare state literature, which I am sure Sumner has no familiarity with. It’s not used because the economists and economic sociologists who operate in that field have never realized that dynamic effects are possible. It’s used because it’s the best measure available to capture the effect of transfers on poverty and cross-country comparisons do not reveal high-transfer countries to have higher market income poverty rates than low-transfer countries (even though their disposable income poverty rates differ dramatically).

Sumner finishes his psychoanalysis by noting “it’s a mistake to think progressives understand this point and choose to omit it.” Heh….heh.

War On Poverty
Sumner’s next point is to repeat a meme he’s heard without ever bothering to see if it’s correct (this is a staple of Sumner’s mission creep blogging, you’ll come to find):

In fact, we spent trillions on the War on Poverty. Unfortunately, poverty won and we lost. Don’t believe me? Isn’t the blogosphere full of progressives complaining that the poverty rate is just as high as in 1967?

The War on Poverty, as you probably know, was a massive smashing success. From 1967 to 2012, the disposable income poverty rate fell from 26% to 16%, a decline of 38.5%. During that same period, the market income poverty rate increased from 27% to 29%. Over the course of those years, transfer programs reduced poverty by a total of 1.2 billion people-years.

While Sumner’s cite to “the blogosphere full of progressives” is creative in the realm of statistical argumentation, whatever it references has it wrong. I suspect he is vaguely conjuring up readings that focus on the Official Poverty Rate, which has bounced up and down and more recently settled back to its 1967 levels, but this rate does not take into account non-cash or tax credit transfer programs like SNAP, school lunch, WIC, housing subsidies (e.g. Section 8), home energy subsidies (e.g. LIHEAP), EITC, and CTC. When you take those into consideration, you get the 38% drop mentioned above. I made an entire website dedicated in part to trying to explain this to people.

As a closing note, it deserves pointing out that adding up the total amount of money “spent” on the War On Poverty and then remarking there is still poverty is a really confused way to understand poverty reduction. Poverty is a flow, not a stock. You don’t ever spend enough total dollars on poverty to make it end because poverty is the state of having an inadequate level of income over a specific time period. On some base level, people must realize this because the paychecks they get from their jobs don’t manage to solve their individual quest to stay out of poverty. You can tell because those paychecks run out and then they have to go get another one.

If I were trying to pull a Sumner here, I would say that last year we spent over $14 trillion dollars (the ballpark of our net adjusted national income) fighting poverty but failed to solve it. Another $14 trillion will be spent this year doing the same thing. If I were trying to be more level-headed, I would say that since poverty is a flow and not a stock, the best way to talk about its solutions is not in terms of dollars, but in terms of institutions. The operative question is what kind of institutions better ensure that the annual flow of poverty is lower? The comparative welfare state literature strongly points in the direction of Nordic social democratic institutions, which heavily utilize transfers and universal welfare benefits.

Disability and Old Age
After trying to argue that transfer programs are merely crowding out market incomes (a conclusion he derives from his prior wrong claim that poverty has not fallen since 1967 even though it has fallen by 38%), Sumner turns around and says that actually you can fight some poverty pretty easily:

Obviously it’s easier to reduce poverty for the elderly, since they are mostly retired. Nixon did that. That would also be true of the disabled, if we could accurately measure disability.

It is indeed easy to do these things. Using the 2013 Official Poverty Metric, we see that transfer programs cut the disabled poverty rate by 31 percentage points and the elderly poverty rate by 35.8 percentage points. What’s interesting here is that at the top of his post, Sumner wants to quibble with my saying transfers reduce poverty a lot using his dynamic effects point. Then he says disability and elderly are two areas where transfers can do big work. But get this: disabled and elderly people make up around 20 million of the 27 million people who were pulled out by transfers last year! Sumner wouldn’t know this because he doesn’t know anything about the data. But his exception for elderly and disabled people actually comprises the vast majority of poor people pulled out by transfers, meaning my point about the market income poverty rate is able to succeed purely by using his exceptional populations.

His Solution
Here is what he wants to do:

That’s not to say we can’t do anything. I’d eliminate all occupational licensing laws and all minimum wage laws and all “welfare.” Legalize employment contracts between consenting adults. No more chronic involuntary unemployment. Then I’d institute a large wage subsidy for low wage jobs. But it wouldn’t be “easy,” there’d be lots of fraud in my preferred program.

As I mentioned above, this is not what the comparative poverty literature for rich, developed countries tells us the best methods for cutting poverty are. But Sumner’s mission creep blogging does not rely upon any expertise or interrogation of the available information, just kneejerk intuitions.

Finally, he ends with this:

In fairness, the way libertarian brains are wired also leads them to overlook lots of obvious points. Which ones? How the hell would I know, I’m a libertarian! Ask Bruenig.

I’ll indulge of course. First, libertarian brains misconceive how superior they are to other brains. One example I could give is how sometimes libertarians think I don’t understand the idea of dynamic program effects when I do. Second, libertarian brains desperately crave the feeling that they are super-logical and super-rational and that other brains don’t get it. Which is funny because libertarian political philosophy is the most logically incoherent political philosophy, perhaps ever. Lastly, libertarian brains generally (Nozick excepted) fail to realize that property is theft, which it is.