Joe Biden and the care economy

Joe Biden and the care economy

By Matthew Yglesias

Slow Boring.com

April 8, 2021

Inside the $400 billion black box

When I first heard that the Biden infrastructure bill would include $400 billion for the care economy — possibly characterized as “care infrastructure” — I was imagining, I dunno, a very complicated, multifaceted program to construct child care centers and create subsidized summer programming for low-income kids.

Or something. I don’t know exactly what I thought. But I have a six-year-old, so to me care means kids, infrastructure means construction, and “the care economy” seemed like something elaborate.

The actual proposal, however, seems pretty simple — it’s just really large.

Basically the way it works is that right now, Medicaid pays for home care services for elderly and disabled people. But it doesn’t pay the caregivers very well. And there isn’t enough money in the program for everyone who needs help to get it. And Biden’s proposal, which offers very few details beyond the large topline figure, is to put more federal money into that Medicaid funding pipeline that should make it possible to help more people and raise pay in the sector simultaneously.

Baumol’s cost disease strikes again

The basic issue here — which you see in a variety of care-related fields — is that there is very little productivity growth in care work. Helping the disabled or elderly with needs they can no longer physically meet themselves is a very labor-intensive process. It’s similar to taking care of little kids — it’s eyes-on, hands-on work where technology doesn’t help you very much.

Costs in this kind of field tend to be very burdensome on the consumer. And yet the actual earnings of the people doing the work are often very low — were they not so low, the cost burdens would be even higher.

The intractability of this kind of problem is one of the reasons I am always so impatient with people who run around and fret about automation eliminating jobs.

It just does not seem likely that we will have robot nannies or eldercare workers any time particularly soon.

What’s more, if we did invent a decent robot nanny, that would be incredibly good news for humanity!

The thing that’s troubling in reality is that not only are there no robots who can do this work, but technology is also doing basically nothing to improve productivity in this sector. Some kind of robot-assisted caregiving could lower costs and raise pay, but we’ve got nothing.

In fact, because care work is such a zero-productivity sinkhole of costs, there’s every reason to believe this sector could absorb a huge influx of new workers if automation makes cashiers obsolete or something.

Now what would be cool is if Biden had identified some ingenious plan to solve the productivity problem in the care sector. Or at least if he had an ingenious plan to solve the productivity problem in some subset of the care sector. What he actually has is a solution to a narrower problem, which is the distributive consequences of unbalanced productivity growth.

The dual-track economy

In an interesting post on Bidenomics that I think is probably more theoretically ambitious than Joe Biden, Noah Smith pitches the concept of a two-track economy by analogy with Japan:

I basically get this notion from Japan. In the 1970s and 1980s, Japan cultivated a world-beating export sector, based around all the companies you’ve heard of (Toyota, Panasonic, etc.). But this was only perhaps 20% of its economy, and the rest was a domestic-focused sector. Although some domestic-focused industries were highly productive (health care!), much of the domestic-focused sector — retail, finance, agriculture, utilities, and a few non-competitive manufacturing industries — was not very productive compared to the U.S. But those sectors did manage to employ a huge number of people; Japan has traditionally had very low unemployment, and that has not changed with the mass entry of women into the workforce since 2012. Japan in many ways built the most effective corporate welfare state in the world. 

How does this apply to America? Well, basically one worry you might have about the United States is that we have some highly productive sectors that are very competitive in the international arena (mostly stuff related to high-tech services), but these sectors just don’t employ that many people.

Note that this is not inherently true of productive economic sectors. One story you might tell about midcentury manufacturing is that we got more productive at making lots of kinds of stuff (cars, houses, appliances), which made it more affordable, which increased demand for the stuff so much that employment in making-stuff kept rising even though productivity was also rising.

graph showing thousands of employees in manufacturing jobs from 1945 to 1976

But let’s say we are pessimistic about that. We think the software sector will keep growing, but will fundamentally remain small because the whole point of software is that once you write it anyone can use it. And so over time, software will keep “eating the world,” as Marc Andreesen says. Back office functions will be replaced by software. Cashiers will be replaced by software. Truck drivers will be replaced by software.

Does that mean mass unemployment?

I’ve always felt that the answer is no. It means more people around to do the kind of care work that is currently very hard to afford. But that still leaves the question: Who is going to pay for it? Biden’s answer is that we will tax the profits accruing to capital and use the money to increase subsidies to the care sector. Google is never going to expand to become a mass employer of working-class labor, but Google’s profits can finance an army of care workers to make sure everyone who needs help gets it and that the people doing the work make a decent living.

But why this, in particular?

The logic of this all seems reasonably sound to me. Years ago, I had an idea for a book that I was going to call “The Era of Big Government,” and it was going to be all about the application of Baumol’s Cost Disease to the future of employment and taxation — basically saying we needed to brace ourselves for a large increase in taxes to rebalance income away from the productive sector to the caring sector.

But then in 2013, Baumol and five co-authors published a book called “The Cost Disease” which advocates exactly that policy agenda, and it didn’t make much of a splash. I figured it would be one thing for me to hijack a famous economist’s idea and turn it into a popular book, but since he himself had already just done that and nobody seemed very interested, there was no point in me doing it.

Biden, however, isn’t quite doing what my hypothetical book would suggest.

Instead, he’s focused on subsidizing one very particular sub-set of this whole subject, namely Medicaid-subsidized home-based long-term care. Not nursing homes. Not preschools. Very specifically home-based long-term care for the elderly and disabled.

There’s an official reason for this, but the actual reason is the Service Employees International Union, SEIU, which is one of the biggest and most important unions. SEIU has found a lot of success in recent years by basically working with politically sympathetic governors to organize home care workers in Medicaid programs. If you just subtract the $400 billion care economy stuff, then you have a big multifaceted program that has lots of union-friendly features. But nothing for SEIU. So this $400 billion proposal — and in particular its promotion into the Jobs Plan while other ideas like child care are left for a hypothetical third bill — is their big win.

Is this a good idea?

As a diagnosis of the problem, Biden’s seems perfectly plausible.

I think everyone agrees that it’s desirable for non-institutional care options to exist for people who have some needs they can’t meet by themselves. And I haven’t heard anyone with a plausible-sounding answer to the cost disease problem.

More broadly, the basic idea of making successful multinational companies pay higher taxes and using those taxes to finance the creation of union jobs that are accessible to less-educated people seems very sound compared to some other notions floating around these days. A lot of people seem to have a slightly half-baked idea that if we do a big antitrust crackdown on tech companies, somehow an industrial economy in the Midwest will flourish.

So I think Biden is onto something.

But the specific idea of putting all the money into this one particular bucket — home-based care that’s financed by Medicaid — rather than spreading it around across a half dozen different modalities of “care” seems so odd that I spent days thinking I misunderstood. Like, surely the nursing home lobby or someone else will pop up and insist that they get their fair share of the cash.

It’s also not really clear from the White House fact sheet how the policy works exactly. Medicaid is structured as a joint state-federal thing where the feds put up a lot of money and the states make most of the decisions about program design. So presumably what we’d be talking about here is a change to the federal matching rate for home-based care. Then you could also tack on an incentive (or a requirement?) to raise pay or facilitate unionization or something else that would meet the administration’s goals on job quality. But there are genuinely no details on how they envision that working.

On one level, I appreciate a White House which acknowledges that Congress writes laws and it doesn’t really matter what details the administration prefers. But normally the way this works is that the White House canvasses the congressional caucuses’ views and then recommends something it knows is likely to be adopted. The fact that this is so much of a $400 billion black box suggests to me that there isn’t any strong consensus here, and this fund could end up getting sliced down either to reduce the need for tax offsets or else to finance other pet projects that key members of Congress are interested in.

At the end of the day, it strikes me as a fairly odd construction. You have a high-level observation about the desirability of increased subsidy for low-productivity care work. Then you have a very specific vision of subsidizing a particular sub-category of that work. And then it gets totally vague again as to how the subsidy should be designed. So far congressional Democrats haven’t said much of anything about this slice of the plan, and it’ll be interesting to see what they do with it.