In Praise of the Wonk: Dissecting the CEA Letter and Sanders’s Other Proposals

February 19, 2016

There are three points to keep in mind about policy debates: Policy specifics are only one of many considerations voters have when evaluating a candidate, and they’re unlikely to mobilize the base in a polarized age; ideological constraints often determine policy positions; and wonk policy “discussions” can easily become a barrier to exclude people and ideas from the conversation. This has become relevant in the Clinton versus Sanders race, which, as Matt Yglesias notes, is an ideological battle like “Ronald Reagan’s battle with Gerald Ford,” one in which wonks have less of a role to play.


That said, I will defend the wonk. I think the wonk analysis is an essential part of the ideological work currently being done and is capable of advancing the progressive project in crucial ways. Working the numbers and the specifics creates clarity, and it forces people to put their cards on the table.

In this specific moment, the work of the wonk forces one to justify constraints, lets you know if you are looking in the correct places, gives you a sense of whether the scope and scale of your changes is sufficient, and lets you know the obstacles and enemies you’ll face. And it can be fun! Or fun enough.

Let’s go through each of these points with specific examples. We’ll begin with the letter from former CEA chairs attacking economist Gerald Friedman’s estimates of the impact of Sanders’s plan, and then look at some cases where wonk analysis would help the Sanders campaign.

To the CEA: What Happened to the Output Gap?

I would have done Gerald Friedman’s paper backwards. He gives a giant headline number and then you have to work into the text and the footnotes to gather all the details. But a core assumption within the paper is that we are capable of getting back to the 2007 trend GDP through demand. We can get the recovery we should have gotten in 2009. Here’s a graphic from Matthew Klein at the Financial Times to this effect:

I’d recommend reading JW Mason’s excellent analysis about why this is an important and reasonable argument to have: “In other contexts, it’s taken for granted that more expansionary policy could deliver substantially higher growth” when there’s still an output gap, and if the output gap has shrunk understanding why is essential.

Wonks can identify this as an essential disagreement and demand clarity. To reject Friedman’s analysis, as the former CEA chairs do, seems to involve rejecting that component of the analysis. If so, they have an obligation to explain what happened to that potential output trend from 2007. I’m not sure where that trend went, and given the stakes we should look for it. Some potential responses:

1. We could get to that trend, but Sanders’s plan won’t get us there. For every dollar in stimulus there’s a dollar taken away from demand, say from single-payer insurance squeezing health care costs. You need to assume that redistribution will have large demand effects in the income distribution and across genders, and that assumption might not be justifiable. The taxes are too high. And so on. This is a totally reasonable argument the CEA chairs could make.

2. We could get there, but neither the Republicans nor the general public would allow it. Perhaps, but that’s a political analysis that the more numbers-focused wonks should ignore. Maybe political “uncertainty” is a blockage, but beyond that, this has always been a poorly theorized argument and I don’t see why it would be a permanent problem even if true.

3. That trend was unsustainable in 2007, before the Great Recession got going. Perhaps GDP was “too high” in 2007 in some sense. But why is that? Do wealth shocks affect long-term supply? Maybe, but that requires throwing the Solow growth model overboard, which would be no less unmoored from economic theory than Friedman’s work, whatever you think of it. And if that were the case, why was the CEA quoting the CBO extrapolating that trend throughout the first years of the Great Recession (e.g. here)?

4. We could have maintained that trend in 2007, but we didn’t during the financial crisis and the Great Recession and it is now lost to us forever. That’s terrifying. But wonks don’t have the luxury of allowing themselves to feel the emotion of terror, and must instead ask “why?” It’s not intuitive to me, but here are some possible explanations:

a. It wasn’t a Great Recession; it was a Great Vacation all along. Casey Mulligan is right, he was always right, and we’ve lost so much potential output because everyone is taking it easy, hanging out in their hammocks and enjoying their redistribution, Obamacare, Dodd-Frank, and not working. But that doesn’t seem to fit the data.

b. The destruction of the asset-backed securitization process that fueled Wall Street and subprime mortgages was just like the destruction that comes from bombing a city. More specifically, *mumble* *mumble* something financial frictions *mumble*. But what about the current world makes us think this is driven by the financial sector? Were the bailouts a failure, then? As Matthew Klein notes, this would justify financial reform an order of magnitude more repressive than anything in Dodd-Frank.

c. Workers suck now. Hysteresis has atrophied them into a vegetative state, robots took their jobs and refuse to hire them as butlers, instead giving even those jobs to other robots. Workers forgot whatever skills they had and the work requires new skills anyway. More importantly, a lot of generational trends all took effect exactly at the moment we happened to hit the zero lower bound on monetary policy, which is a crazy coincidence. But what about the labor markets would tell us this is the case? And couldn’t high demand function like a type of anti-hysteresis?

d. You could simply say “the CBO says so.” Such appeals to authority are fine for pundits but are anathema to a wonk. At the least it should be clear why the evolution of the CBO’s numbers is convincing, which leads us back to the questions above. It should also be clear that this isn’t simply the numbers portraying our current inaction as a permanent state.

As Jared Bernstein notes in the graphic above, the line is moving to us, not the other way around. Did we ever really have an economic recovery? This is really the question that should animate us at the end of the Obama presidency, because it animates all the politics of the moment. But only a wonk looking at the constraints can see this and demand that the CEA tell us what happened to that potential output.

To Sanders

A decent wonk read can also bring clarity to Sanders’s proposals. Let’s not go to health care right away, but start with:

Mass incarceration policy needs a bottom-up approach Sanders has repeatedly said, “[H]ere’s a promise that I make to you: Number one, at the end of my first term, we will not have more people in jail than any other country.” Leon Neyfakh at Slate argues that this is “ridiculous.” So have many others. Even if Sanders pardoned all the federal prisoners, which are only about 10 percent of those in a cell, we’d still be the world’s leading jailer.

So what? Wanting to end mass incarceration is an excellent aspirational goal. But here’s the problem: You can’t tackle this goal solely at the federal level, and you can’t do it by focusing on nonviolent offenders and for-profit prisons as Sanders does. Those are incredibly important goals, of course, but those numbers and that focus aren’t where you need them to be in order to achieve this goal. To achieve this end, you need a different plan that also involves state and local incarceration, as well as reduce the sentencing of violent offenders.

Identifying where the problem exists is an advantage of working the numbers, which is something a lot of researchers have been doing. (See this interview with Marie Gottschalk for an example.) Running the high-level numbers forces discipline on the actual problem at hand and can allow people to innovate in response. It would be easy and great for Sanders to say “as part of my revolution I want to see everyone in the audience run for local office, and then run for state office, and then pass common-sense measures designed to reduce mass incarceration.” Or to say “county prosecutors and upscaling of felony charges are the problem, and my Department of Justice will investigate this.” But that’s a much broader agenda than what he proposes.

You can’t break up the banks easily through executive policy Another concrete promise from Sanders is that “[w]ithin one year, my administration will break [financial] institutions up so that they no longer pose a grave threat to the economy as authorized under Section 121 of the Dodd-Frank Act.” Rob Blackwell of American Banker describes this as “cuckoo.” Crucially, it is impossible to pull this off within one year. You’d need a supermajority of regulators to do it, and you’d need a Federal Reserve chair on board to guide it. Yellen’s term wouldn’t be up until the end of the first year.

Showing the scale and scope of the problem is an advantage of working the numbers and policy. Again, don’t we want to reduce the power of the largest banks and the risks they pose to the economy? Yes. But the problem here is the idea that executive will alone is sufficient to break up the banks in a short timeframe. You instead need a deeper bench taking over the regulatory environment over many years, or you need to break the banks up in more roundabout ways (like through living wills), or you need to be ready for the next crisis, or you have to go back to Congress. The idea that it can be accomplished through executive orders and executive will distracts us from the severity of the problem and doesn’t allow us to focus on a proper long-term strategy.

Moving to single-payer involves more enemies than the normal health care reform pitch I am not a health care policy person, but on my surface read, the fundamental disagreement between estimates from the Sanders campaign and economists like Kenneth Thorpe is how many rents and excess costs you can squeeze out of the system. All sides agree the plan squeezes private providers of health care, and if Thorpe’s more pessimistic estimates are true, you would need to squeeze private providers harder in order to avoid increasing taxes as much as you’d expect.

This is a different group of enemies than those normally discussed. This isn’t private insurance shareholders, drug companies, and administrative bloat. It’s nurses and doctors. You would need a different campaign of delegitimization and organizing if they were the targets rather than hedge fund profiteers like Martin Shkreli. It also means the GOP campaign—“Sanders wants to put everyone on Medicaid and is lying to you about jacking up your taxes”—would write itself.

Why the Right Can Slack on This and the Left Can’t

My personal thought is that path dependency matters, and building through and on what’s already been done is better than starting over. The reasons why are worth their own post.

But it’s important to remember that the left can’t simply be symmetrical to the right on the analysis question. The right gets away with the asterisks in its plans because it wants tax cuts to bankrupt the government, or it thinks that deficits don’t matter, or it just creates another excuse to cut programs for the poor. More generally, as Thomas Frank noted, conservatives benefit when the government doesn’t work well. They have incentives to do bad analysis in a way the left doesn’t.

Democrats can never duplicate this dynamic, no matter how ideological they become. Government failure fuels a right-wing agenda in a way it can never do for liberals. Left-liberals come at the situation with a different economy of power and they need to be strategic about what they propose. The correct analysis, done well, can help point out which directions to take and the likely obstacles you’ll find along the way.

Also wonk is know spelled backwards.