Bottom-up

On Thursday, the Fed took a dramatic step to try to stabilize the financial system, offering banks trillions of dollars in new short-term loans. The loans would last for up to three months. Banks could only access the money if they turned over super-safe assets—mainly US Treasury bonds—whose value was equal to the loan amount.

The basic idea was as old as central banking itself. In times of financial stress, people want to hold more cash. That puts pressure on banks, who don’t have as much cash as people want, which can lead to a financial panic. To stave off panic, the central bank creates additional cash and lends it to the banks, allowing banks to meet customer demands and stave off a meltdown.

Unfortunately, the large dollar amounts and the convoluted process the Fed uses to distribute the money has led to confusion and misguided criticism. Rep. Alexandria Ocasio-Cortez, for example, griped that the $1.5 trillion the Fed is offering to banks this week could cover “all student loan debt in the US.” A number of people complained that it was a bailout for banks. These folks seem not to know, or care, that the banks will have to pay back ever penny, with interest, within three months. That’s not a bailout.

In this post I’ll try to dispel some of the confusion surrounding the Fed’s actions. First I’ll use a simple example to illustrate exactly what the Fed is doing. Then I’ll explain why actions like this are in everyone’s interest.

The Amazon card loan

Let’s say you’ve got a friend who owes $475 in rent tomorrow, but he’s broke and his $1,000 paycheck doesn’t come until next week. You notice that he has a $500 Amazon gift card, so you make him an offer. You’ll lend him $500, and you’ll hold on to the gift card. Next week, after his paycheck comes, he’ll pay you $500 and you’ll give the card back. If he doesn’t pay, you keep the card.

The risk to you is very low. You’re a regular Amazon shopper, so it wouldn’t be much hassle to load the $500 into your Amazon account and use it for future purchases. One way or another you’ll get your money back.

The Fed’s new loan program works essentially the same way, except that the Fed doesn’t take Amazon gift cards as collateral. To get a loan, banks have to put up US Treasury bonds or a few other forms of US government debt.

To take out a loan, a bank electronically sends the Fed (say) $1 billion in Treasury bonds. The Fed creates $1 billion in new new money and sends it to the bank. Then after a specified time period—which can range from overnight to three months—the bank returns the Fed its $1 billion in cash and the Fed gives the bank its $1 billion in Treasury bonds. (The bank also pays the Fed some interest but I’m gonna be a lazy blogger and not figure out the exact mechanics of that).

In the very unlikely event that a bank fails to repay a loan, the Fed keeps the Treasury bonds, whose value will be very close to the amount originally lent out. So the risk being taken by the Fed is tiny.

To be clear, this isn’t a new type of loan program. The Fed has been offering to make loans like this to banks for years. What’s new is the scale of the program. In the past, this type of loan offering was often capped at $100 billion or less across all banks. In normal times, that was more than enough to meet bank demand—banks typically didn’t even use the full amount offered. But now the Fed has drastically increased its limits, offering three different $500 billion loan programs just this week—with more to come.

OK, so that’s what the Fed announced it was doing on Thursday. The more important thing to understand is why.

“Highly unusual disruptions in Treasury financing markets”

Most people expect the Coronavirus to be economically disruptive. So if you’re a Chief Financial Officer at an American company, one of the first things you’re going to want to do is get more cash on your balance sheet. Suppose your company got a $100 million line of credit from a bank last year and you’ve only used $50 million of it. So you decide to take the extra $50 million now.

That might be a smart move for you, but now the bank has a problem. Every CFO is having the same though at around the same time. The bank set the credit limits on these accounts based on assumption that most customers wouldn’t use the full amount. As more and more companies borrow more and more money, the bank’s cash reserves are dwindling.

This isn’t to say the bank is about to go bankrupt. This particular bank has billions of dollars invested in safe assets like US Treasury securities—more than enough to cover all of its liabilities.

But the bank doesn’t want to disappoint a customer—and possibly set off a panic—by refusing to honor an existing credit line. And you, the business customer, need $50 million in cash. You can’t pay your workers with Treasury bonds.

Can’t the bank just sell some of its Treasuries and use that cash to meet its customers’ cash needs? In normal times, that’s easy. There’s always someone willing to buy US Treasury bonds. But these are not normal times.

In its Thursday statement announcing the new loan program, the Fed wrote that it was aiming to “address highly unusual disruptions in Treasury financing markets associated with the coronavirus outbreak.” That’s cryptic Fed-speak, but a Bloomberg story offers some details on what the Fed might have been seeing.

Yields in the world’s largest debt market have been on a mind-bending, three-week roller-coaster ride. At one point, the entire U.S. yield curve was below 1% for the first time ever. But this week rates have jumped from Monday’s all-time lows even though fear of the virus has intensified… This volatility is happening as trading-platform order books thin out to a degree last seen during the 2008 financial crisis, making it harder to use Treasuries as a gauge of investor anxiety.

I’ll admit I don’t fully understand this. But the basic issue is that what’s supposed to be the world’s most liquid asset—an asset you can always, always sell quickly and for a predictable price—has become less liquid and less stable recently.

Any one person or company can use Treasuries as a ready substitute for cash and everything will work fine. But if a lot of people start to use them that way—and they all start to exhibit herd behavior, with a lot of panic buying and panic selling—it can lead to sudden swings in bond values, or even to moments when it’s hard to find a buyer for every seller, or vice versa.

So one way to look at the Fed’s actions Thursday was that they were just trying to calm down Treasury markets. Now a bank that is trying to sell Treasury bonds for short-term liquidity needs doesn’t actually needs to sell the bond at all—it can use the bond as collateral for a loan. By guaranteeing that this kind of loan will always be available, the Fed hoped to calm down the markets and limits wild swings in the market.

You might think that volatility in bond markets aren’t of any concern to the average American, but I think the experience of 2008 suggests otherwise. Nobody cared very much about mortgage-backed securities in mid-2008, but chaos in that market sparked panic in other markets, ultimately bringing the entire financial system to the brink of collapse. I can’t predict exactly how a dysfunctional market for US Treasuries would create bad consequences elsewhere. But I’m glad the Fed didn’t wait to find out.

Why liquidity is crucial

Here’s a broader way of looking at it that I think makes the stakes for the average American clearer.

Fundamentally, what happened on Thursday was that a lot of people—not just banks, but their millions of individual and business customers—wanted to increase their cash holdings to help them ride out the Coronavirus pandemic and any economic dislocation that comes with it. And fundamentally what the Fed did was create an orderly way for that to happen.

That was important because my hypothetical about businesses stockpiling cash wasn’t just a hypothetical. On Wednesday evening, Bloomberg reported there was a “dash for cash” in corporate America.

“Behind the scenes, some CEOs and their finance chiefs are calling bankers this week to ask for liquidity,” Bloomberg’s Sridhar Natarajan and Yalman Onaran wrote. “And throughout the day Wednesday, word leaked out on company after company pulling from existing facilities.”

“There is fear in the market that liquidity might dry up, so business owners want more cash in hand to weather a host of uncertainties,” the president of an online lending platform told CNBC.

It’s in everyone’s interest to make sure that corporate America can get the cash they need—assuming there’s a bank willing to lend to them. If there’s an economic downturn later this year, extra cash on corporate balance sheets could make the difference between big layoffs and small layoffs—or no layoffs at all.

So that’s what the Fed did: it made sure that banks could access all the cash they needed to serve their customers. It would have been a self-inflicted wound for the Fed to allow a panic to intensify because banks were having cashflow problems and couldn’t make loans to creditworthy borrowers. Even if that hadn’t sparked an immediate crisis—which it very well could have—it would have left corporate America, and the American economy more broadly, less well-prepared for the difficult months ahead.

I enjoyed a recent blog post by Jerry Taylor, president of the formerly libertarian Niskanen Center, on the case against ideology. Unsurprisingly, Taylor has gotten rebuttals from thoughtful libertarians like Ilya Somin.

Taylor says that signing up for an ideology like libertarianism means giving free rein to the human tendency toward motivated cognition. Somin counters that anyone who fancies himself beyond ideology is fooling himself. Everyone has an ideology, he says, and the only real question is whether you choose to make your ideology explicit or not. Self-conscious ideologues have the opportunity to engage in self-reflection and self-criticism about their thinking, Somin argues. In contrast, a self-proclaimed non-ideologue “cannot even begin to curb potential ideological bias on his part, because he believes himself to be above such things.”

I think it’s helpful to distinguish what we might call narrow ideologies from broad ones. For example, I’m personally sympathetic to the “copyleft” ideology of groups like the Electronic Frontier Foundation, to YIMBYism, and to market monetarism. You might question whether these count as ideologies, but I would argue they do. Advocates for these narrow ideologies have shared arguments in favor of these views and common models of how the world works. They often have idiosyncratic vocabulary and a distinctive way of telling the history of their issues.

Narrow ideologies are narrow because they only say something about a small corner of the policy space. Being a YIMBY doesn’t imply any particular views on abortion, capital gains tax rates, or the war in Syria. At the opposite extreme are broad ideologies like libertarianism that offer official positions on almost every policy issue.

As a young libertarian, I could recite for you the libertarian perspective on almost any issue you can think of (with a few notable exceptions like abortion), and run through the standard libertarian arguments in favor. But I’ve come to believe that it was foolish to think you could derive views on a vast range of policy issues from a single principle.

Libertarianism has compelling things to say about a number of issues—I’m a fan of libertarian work on the importance of secure property rights, the value of school choice, the problems with occupational licensing, etc. But there are a lot of policy issues where libertarianism, as such, doesn’t provide much useful insight. On these issues, trying to apply libertarian insights can have disastrous consequences—see Ron Paul’s views on monetary policy, for example.

Libertarianism isn’t the only broad ideology, of course. Old-school Marxism was a broad ideology that saw every issue through the lens of capitalist exploitation of labor. More recently, it has become fashionable in some parts of the left to view every issue as a struggle of marginalized groups against the privileges of white men. On the right, Steve Bannon has been cultivating the mirror image of this ideology, which sees most policy issues as the struggle of ordinary (read: white) middle-class Americans against foreigners and “globalist” elites.

Marxists have useful things to say about the importance of worker autonomy in the workplace, while intersectionality theorists absolutely have important insights about the subtle ways that structural factors can perpetuate group inequality. But it’s a mistake to think any one ideology can answer every policy question. Libertarianism, as such, doesn’t have much useful to say about monetary policy, just as most self-described socialists don’t have realistic ideas about how to organize capital markets.

Rejecting ideologies entirely is a bad idea because ideologies are powerful engines for social change. To change society you need to formulate a bundle of internally consistent arguments and then convince a lot of people to promote them together—that’s an ideological movement. Somin is right that people never fully transcend ideology, and it can be useful to think about ideology in explicit terms rather than pretending to be entirely non-ideological.

But luckily, we have the option to treat all ideologies as narrow ideologies. We can learn the most important insights from each ideology while resisting the temptation to try to shoehorn every issue into a single over-arching conceptual framework. I like the Institute for Justice, for example, because they’re a libertarian organization that has focused on a handful of issues—school choice, economic freedom for small businesses, property rights—where libertarianism yields a lot of insight.

A wise policy expert has a full ideological toolbox and has a sophisticated understanding of which tools are most usefully applied to which issues. A foolish policy expert has a toolbox full of ideological hammers and approaches every policy problem as if it’s a nail.

Today Congress is either going to pass a last-minute spending bill to keep the government open, or the government is going to shut down. Either way, the situation is an embarrassment. It has been many years since Congress went through anything like a normal appropriations process, where legislators have the time to comb through the federal budget and make thoughtful decisions about how to allocate taxpayer dollars efficiently.

The hour-by-hour style of conventional news coverage tends to obscure the big picture: the perpetual crises the US government has suffered over the last decade are a symptom of America’s deeply flawed constitutional system. This isn’t a new insight on my part. You can read Matt Yglesias’s classic 2015 write-up of the argument, which in turn draws on a large body of political science literature.

The basic issue is that the American system of checks and balances was designed for a nation without ideologically polarized parties. For most of the 20th century, there were genuinely liberal Republicans and conservative Democrats, and that meant there was room for the kind of compromise and horse-trading that’s required to make our system of government—with its many veto points—work. But now that the most conservative Democrat is more liberal than the most liberal Republican in both the House and the Senate, compromise has become more and more difficult.

The problem is compounded by the fact that it’s so hard to remove a bad president from office. If Donald Trump had been elected prime minister in a parliamentary system, Republicans would have recognized their mistake months ago, deposed him, and chosen a new leader. Instead, Republican members of Congress feel immense pressure from their base to rally around an obviously incompetent president.

I explain Matt’s argument to people pretty frequently, and I usually get the polite equivalent of an eye roll. America has had the same basic constitutional structure for so long that it’s hard to even imagine changing it. People who talk about it too much risk being seen as a nut.

But Matt is right, and eventually everyone is going to realize it. The only question is how severe of a crisis we have to have before people start to take it seriously. Unfortunately, I think it’s likely to take a really big crisis—bigger than any we’ve experienced since the 1860s—before Americans seriously consider significant changes to the way the US government works.

But it should be possible to widen the Overton window on this, which might improve the chances of fixing the problem before a catastrophe occurs. There are lots of ideas that are a little out there—Social Security privatization, single-payer health care, nominal GDP targeting, a universal basic income—but nevertheless have networks of think tank scholars, academics, and activists working to make them more mainstream.

People in these networks work on fleshing out the details of how these reforms would work in practice. Reporters can call these scholars up and get quotes saying that the real problem behind this week’s mini-crisis is that the nation hasn’t adopted this or that more fundamental reform.

Every ambitious reform starts out sounding a little crazy, but if the arguments for idea are persuasive then it will gradually pick up more followers and become an idea that people feel they have to take seriously when covering the topic. Then, sometimes, the ideas get attached to a winning political coalition and become policy.

Constitutional reform is a little different from the other ideas I mentioned because the hurdles to fundamentally changing the Constitution are much higher than the hurdles to overhauling the health care system or Social Security. Probably the only way we’ll get reform is with a significant political crisis—a war, worse-than-2008 economic crisis, military coup, etc.

In these kinds of extreme situations, people will be looking for more ambitious reforms. But if the necessary groundwork hasn’t been laid, there’s a danger that people will draw the wrong conclusions and make the wrong reforms—or no significant reforms at all. Having a network of activists and scholars ready with a cogent explanation of what went wrong and a menu of reform options greatly increases the chances that we’ll only have to go through one major crisis, rather than suffering an increasingly wrenching series of them.

So if you’re a think tank president, a foundation grantmaker, or just a rich guy looking for a way to make the world better, this should be on your radar screen. In the grand scheme of things, it doesn’t take very much money to turn a good but obscure idea into an idea that normal people feel compelled to take seriously.

This past week Anne-Marie Slaughter, President of the New America Foundation, fired Barry Lynn. Lynn was the head of New America’s Open Markets project, which had become Washington’s leading center making the case for stricter enforcement of antitrust laws.

New America is heavily funded by Google, and Lynn says he was fired for his anti-Google work. Slaughter says Lynn is wrong, that he was fired for his “repeated refusal to adhere to New America’s standards of openness and institutional collegiality.” She elaborated on this in a Thursday interview with the New York Times’s Ken Vogel, arguing that Lynn had violated “strong implicit norms about providing a heads-up when you are doing something that could have an impact on the funding for your fellow directors.”

Slaughter seems to regard this as a defense of her actions. But it looks to me like an admission that New America does not, in fact, defend the intellectual independence of its scholars, despite promising on its website that New America’s research is not “influenced in any way by financial supporters.”

New America’s norms are at odds with scholarly independence

This promise of independence is crucial to the credibility of a think tank like New America. But think tank leaders will always be tempted to cheat, encouraging scholars to publish donor-friendly research—or at least avoid publishing openly donor-hostile research—to help with fundraising.

And the problem for think tank leaders who are actually committed to scholarly independence is that it’s difficult to make that promise credible. Every think tank leader tells scholars that they’re free to pursue their research wherever it leads. Some of them mean it. Others are simultaneously dropping hints that there are certain lines that scholars shouldn’t cross. But because they’re only hints, scholars can’t be sure which kind of organization they’re working for–at least before they get fired for crossing one of those invisible lines.

And this is why a think tank president who really is committed to defending scholarly independence needs to be super careful not to send mixed signals. That means never saying things that could be misinterpreted as pressure to avoid antagonizing donors. It also helps to establish processes and norms that re-assure scholars that they don’t need to worry about how funders might react to their work.

Slaughter’s comments about Lynn’s departure make clear that Slaughter has done the opposite of this at New America. She told the New York Times that Lynn had violated “strong implicit norms about providing a heads-up when you are doing something that could have an impact on the funding for your fellow directors.”

In other words, every time Lynn’s group did something Google might not like, Lynn was expected to go around to other groups that receive funding from Google (or Google chairman Eric Schmidt) and give them a “heads up” that their fundraising might take a hit. Nobody likes being the bearer of bad news, so this creates an obvious incentive for Lynn to avoid criticizing Google too often. Forcing program directors to do this, then, is a clear signal from management that scholars should be constantly thinking about which donors their work might offend.

Even more blatant was the email Slaughter sent to Lynn last year, a few days before Lynn sponsored a speech by Sen. Elizabeth Warren (D-MA) attacking monopolies like Google. “Just THINK about how you are imperiling funding for others,” she told Lynn. “We are in the process of trying to expand our relationship with Google on some absolutely key points.”

Slaughter’s explanation for this is that she wasn’t objecting to Lynn’s substantive work—holding a conference with Warren as a keynote speaker—but about his failure to provide her and Google with adequate notice that the speech was coming. But again, if you think about the incentives this requirement creates, there isn’t much practical difference between the two.

Nobody wants to bring bad news to their boss, and obviously telling Slaughter that the Open Markets team is about to sponsor a conference with Google-bashing speakers counts as bad news. When Slaughter found out about the Warren keynote, she told Lynn to drop everything and write some talking points she could use in a forthcoming meeting with a Google representative. The clear subtext was that Lynn had created a problem for Slaughter, and Lynn needed to help Slaughter clean up the mess.

If your think tank’s scholars spend a lot of time worrying that their work could be bad for donors, then your think tank isn’t doing a good job of defending scholarly independence. And New America’s policies, as described by Slaughter, force program directors to think about this kind of thing constantly if their teams do work that criticizes major New America donors.

Media companies set a higher bar

I think it’s instructive to compare New America’s policies to those of media companies, which face a similar dilemma. Like think tanks, media companies get their credibility from a public perception that their reporting is independent of advertiser and investor pressures. But credible media organizations have much stronger protections for editorial independence than New America does.

Take Vox Media, where I worked until July, as an example. Vox Media counts Comcast as a major investor. Writers for Vox.com (and sister sites like the Verge) are sometimes positively gleeful about criticizing Comcast. I always enjoyed writing articles criticizing Comcast and then putting a disclosure at the bottom of the article that Comcast was a major Vox Media shareholder. It was the strongest possible signal that we could send that we were not, in fact, beholden to our financial backers.

Maybe those articles caused headaches for Vox Media’s CEO, Jim Bankoff, when he met with Comcast representatives. But if so I never heard about it. To the contrary, my editors encouraged me to write more articles like this, because they tended to be popular with readers. Vox’s norms of editorial independence were strong enough that we never worried that criticizing our financial backers could put our jobs at risk.

It would have been considered a flagrant violation of journalistic ethics if Vox’s CEO had asked me or my editor to give him a “heads up” when we ran anti-Comcast articles, to say nothing of preparing talking points for meetings with Comcast or other financial backers. And the problem with this isn’t just that it would have forced us to spend time thinking about how to suck up to Comcast. The larger problem is that Bankoff had the power to fire any one of us if he wanted to. So even if he claimed he was only asking for a “heads up,” we couldn’t help but wonder whether this was really a subtle signal that we should knock it off—or else.

Slaughter seems to consider this kind of thing overkill. “I recognize that the best journalists operate on a different principle — notice seems to imply interference,” Slaughter wrote in a Friday blog post. “But we are not a newspaper, yet we try to uphold the best journalistic standards in our writing.”

I think she’s fooling herself. The news business has the norms it does because the industry has learned, over decades of hard-won experience, that that’s what it takes to safeguard their credibility. News organizations that stray too close to the line tend to blunder into scandals that damage their credibility with the public.

Of course, it’s possible that raising New America’s ethical standards would alienate some of its corporate donors, costing some New America employees their jobs. So having built an empire based largely on corporate money, Slaughter is in an awkward position.

But New America also gets a lot of money from genuinely philanthropic sources like the Ford Foundation, the William and Flora Hewlett Foundation, and the Open Society Foundation. The New York Times quotes one of New America’s foundation funders saying that “you want to let the grantees do their work without worrying about how it impacts the funders.” So New America could raise its ethical standards and still raise millions of dollars from these kinds of donors. In the long run I think this would make New America a better, more influential institution.

Throughout the 19th and 20th Centuries, the world had far more investment opportunities than we had capital available to invest. With the population growing, there was always demand to borrow money to build more homes, offices, and stores. And with a steady stream of new inventions — railroads, telegraphs, automobiles, refrigerators, televisions, cell phones, and so forth — there were always plenty of opportunities to build new factories, to make productivity-enhancing upgrades to existing factories, and to do research and development to advance the state of the art.

But what if that came to an end? Obviously, we’re never going to reach a point where there’s no investment opportunities at all. There will always be some people wanting to build a new house or start a new business. But the pool of available capital keeps getting bigger and bigger, while population growth is slowing and many recent inventions are not as capital intensive as those in previous generations. We could reach a point where there are no more investment opportunities at the margin.

If we lived in that world, we’d see inflation-adjusted interest rates around the world falling to zero. We’d see cash and short-term investments piling up on corporate balance sheets. We’d see the CEOs of even America’s most innovative companies deciding they have nothing better to do with this cash than to give it shareholders through buybacks and dividends. And we’d see the few companies that do seem to offer the possibility for significant returns earning ridiculously high valuations as too much venture capital chases too few startups.

In other words, a world with too much capital and too few investment opportunities would look a lot like the world we’re in right now.

During the 19th and 20th centuries, interest rates served two important functions. One function was to reward and encourage frugality. If you spent less than you earned, you could put the difference in a savings account and earn a risk-free return higher than the rate of inflation. The money would be leant out to others who wanted to build a house, start a business, or otherwise make a productive investment that grew the economic pie, and you got a share of the gains.

The second, more subtle, function of interest rates was to prevent severe recessions. A healthy economy relies on a steady rate of spending, which requires that people, on average, not hold on to cash for too long. If a lot of people pile up savings that don’t get re-invested, the result will be a reduced rate of spending, a.k.a. a recession. This is why the Fed lowers interest rates during recessions — it’s a market mechanism that encourages savers to save less (and hence spend more) and borrowers to borrow more (and hence spend more).

If society has run out of productive investment opportunities at the margin, then frugality will no longer be socially beneficial (though of course it might be beneficial to the person doing the savings). More savings won’t produce faster growth, and so there’s no reason for society as a whole to encourage people to be frugal. An interest rate at or below the inflation rate is a market signal that society has enough capital, doesn’t need any more, and so won’t be rewarding people who accumulate savings.

The second function of interest rates poses a larger problem when interest rates fall to zero. Remember, recessions happen when spending slows because people are accumulating savings that doesn’t get re-invested in houses, factories, and so forth. When interest rates are positive, then interest rates can fall to encourage less saving and more investing. But when interest rates have fallen to zero, this doesn’t work any more, because it’s safer and easier to just hold onto the money. Central banks have tried to address this issue by flooding the market with more and more cash. But it’s been difficult to get the money to circulate — it just keeps piling up in peoples’ bank accounts.

The problem here isn’t that people are saving too much, it’s that they’re doing their savings in cash rather than by investing in assets like stocks or bonds. This wasn’t an issue in the past because banks would automatically invest the money on their customers’ behalf. But now banks are failing to do that, and so money customers save in bank accounts gets effectively pulled out of circulation.

Another way of looking at this is that we want people to use cash primarily as a medium of exchange rather than a store of value, so it’s important that cash have a lower rate of return than other investments. In the past, non-cash investments tended to have a significant positive rate of return, and so the fact that cash had a rate of return of 0 was plenty of incentive not to hoard cash.

If we can’t raise the rate of return on non-cash investments, then the alternative is to lower the rate of return on cash. The Fed already does this by setting a 2 percent inflation target — essentially that’s a 2 percent penalty for holding cash. In an era where inflation-adjusted interest rates were almost always positive, that was viewed as an ample disincentive for cash-hoarding.

But in a new era of ultra-low interest rates, it’s apparently not enough. So one solution would be for central banks to raise the cash-hoarding penalty — that is, its inflation rate target — perhaps to 4 percent. Then even if non-cash investments don’t produce a positive return in inflation-adjusted terms, there will still be a significant spread between the returns on cash and on other assets. That will discourage people from holding a lot of money in the form of cash, making it easier to prevent major recessions.

One of the big questions about the economy right now is whether monetary policy is a significant factor in the sluggish growth of the last five years. Lots of economists think that easy money was important for speeding the recovery between 2008 and 2010. But even some economists who believe monetary policy mattered in the early post-2008 years think that monetary policy can’t explain the sluggish growth we’ve seen more recently.

To quote the always-insightful Eli Dourado: “Monetary policy is important for 18 months or so after a nominal shock w/ an otherwise functioning financial system. That’s it.” Alex Tabarrok, co-author of the popular Marginal Revolution blog, recently told me it’s “crazy” to think monetary policy could explain the slow growth we’ve seen in the last few years.

I’m not so sure.

There are two major arguments that monetary policy can’t explain the recent slowdown. First, monetary policy is about as loose as it can be, so there isn’t room for the Fed to do more. And second, even if the Fed could do more, it wouldn’t do any good because with unemployment at 4.9 percent there isn’t any slack left in the economy.

As we’ll see, neither of these claims is very persuasive.

The Fed’s itchy trigger finger is holding back the recovery

After a big demand shock like the 2008 financial crisis, output temporarily falls below the economy’s long-term potential. Workers are laid off, factories get idled, construction projects are slowed down, causing the economy to produce goods and services at less than its full capacity.

But the conventional view holds that this shortfall shouldn’t persist for all that long. Businesses want to increase output. Workers want to work. And so unless there are further shocks, the argument goes, output should return to potential within 18 months or so.

Now imagine it’s 2009 and we have a diabolical Fed chair who wants to prove this conventional view of monetary policy wrong. Her strategy: every time the economy shows signs of catch-up growth — growing faster than the long-term trend to close the output gap — she tightens just enough to prevent that extra growth from occurring. Contrariwise, if the economy looks like it’s tipping into a recession, she eases enough to keep the economy growing, albeit at a sluggish pace.

If you think a sufficiently tight monetary policy can produce a big downturn in real output — and most economists do — then it stands to reason that less-tight monetary policy should be able to produce slow but positive growth. So theoretically, a central bank should be able to maintain a non-trivial output gap indefinitely.

And while I definitely don’t think Janet Yellen has been trying to prevent an economic boom, I think that’s been the practical effect of her decisions since she took office in early 2014.

This might seem implausible since interest rates have been close to zero ever since 2008. But as Milton Friedman emphasized, low interest rates can be a misleading barometer of whether monetary policy is tight or loose.

It’s useful to go back to earlier in the crisis, when the Fed had cut interest rates to 0 percent and still believed more stimulus was needed. One thing the Fed did was a bond-buying program known as “quantitative easing.” But another thing the Fed did was called “forward guidance.”

Forward guidance meant that the Fed promised to keep interest rates low for a long time — even after the economy started to recover. In 2011, the Fed promised to keep rates low for an “extended period” without specifying how long that period would be. In Septemner 2012, the Fed said that it would maintain low rates “for a considerable time after the economic recovery strengthens.” Then in December, it adopted the “Evans Rule,” which said that interest rates would stay low until the unemployment rate fell below 6.5 percent.

In a narrow sense, the Fed wasn’t “doing anything” when it issued this kind of forward guidance. It was just talking. But economists know that this kind of talk can be powerful. Market actors care about future Fed actions almost as much as they care about current ones. A commitment to maintain low rates in the future means a higher probability that there will be a big and long-lived economic boom, which means that investments made today are more likely to pay off in a few years. So when the Fed commits to easy money in the future, it can stimulate more investment in the present.

Since 2014, the Fed has taken the opposite tack, repeatedly talking about “normalizing” monetary policy — that is, raising interest rates. The Fed has been doing this even though inflation remained well below the Fed’s 2 percent inflation target and GDP growth remained sluggish.

The reason the Fed hasn’t followed through on these threats is that the economy keeps under-performing expectations. An expected June 2015 rate hike was delayed until December after disappointing economic performance in the summer and fall of 2015. The December rate hike happened despite the fact that inflation expectations were falling at the time. That was supposed to be followed by a series of rate hikes in 2016, but at each meeting the Fed has flinched in the face of disappointing economic performance and global economic turmoil.

We can view this policy as the opposite of the stimulative forward guidance the Fed was offering between 2011 and 2013. Back then, the Fed said “don’t worry, we’ll keep rates low even after the economy has recovered. Now, the Fed is saying “watch out! We’re going to hike rates as soon as we’re sure it won’t cause a recession — and maybe sooner.” If promising to keep rates low for an extended period makes monetary policy looser, promising to hike rates ASAP makes it tighter.

And because the Fed’s statements about future policies can effect the economy in the present, promising to raise rates soon can actually cause rates to stay low for longer. The Fed’s signals about impending rate hikes makes markets more pessimistic about the economy’s trajectory. That depresses spending and investment, weakening the economy enough that the Fed is forced to delay its rate increase.

In short, the Fed’s policy over the last couple of years has effectively been to tap on the brakes any time the economy shows signs of producing faster-than-trend growth. And because markets know that’s the Fed’s policy,
they’ve come to expect slow growth, making them less likely to invest in boosting output.

“Full employment” isn’t necessary full employment

OK, so money isn’t as loose as it could be. But the conventional view holds that this doesn’t matter because the unemployment rate is 4.9 percent. That’s very close to the level that economists consider to be “full employment.” With very few idled workers, the argument goes, there’s no room for higher demand to produce higher real output. If the Fed made monetary policy looser, it would just produce more inflation.

But I think this takes too simplistic a view of the concept of full employment. There are a number of ways an economy can be at 4.9 percent unemployment and still be well below its potential output.

One way is known in economics jargon as “scarring.” A bunch of people lost their jobs in 2009. As the recession dragged on, a lot of them couldn’t find work for several years. And once a worker had a multi-year gap in his resume, it became harder and harder to find a job, since employers worried that the worker’s skills were stale or that there was something wrong with the worker that led to the prolonged spell of unemployment. Eventually, many of these workers gave up looking for work altogether. If they were in their 50s or 60s, they may have been forced into a premature retirement. If they were younger, they might have gone on disability, become stay-at-home parents, or moved in with family.

The conventional view treats scarring as an irreversible event: once a worker has left the workforce, he or she’s no longer counted in the unemployment rate and is effectively excluded from the concept of full employment.

But these workers do exist, and they could starting working again with some extra encouragement and training. What they need is a 1990s-style boom, in which employers are so desprate to find workers that they’re willing to hire people they wouldn’t have considered in leaner times. In tight labor markets, employers start asking friends and family if they know of anyone looking for work. They set up training programs to teach skills that are scarce on the ground. And so the labor force can become partially “unscarred,” with thousands of people who had given up on finding work returning to the workforce.

A boom could can also help workers shift into jobs where they’ll be more productive. During a recession or a slow recovery, many workers feel that their job situation is precarious — that if they lose their current job, they’re unlikely to get another one like it. That makes them reluctant to make risky career moves — like switching to a new field, going back to school, or starting a business — that could boost their long-run productivity. As a result, many workers wind up in career ruts, spending years in jobs where they’re performing at well below their level of potential.

In a 1990s-style boom, employers become desperate for workers and workers become confident that they’ll be able to find a decent job if they need one. At the same time, employers become more open to hiring people with unconventional backgrounds, making it easier for people to hop from one industry to another. In short, the labor market becomes more fluid, and this dynamism allows more workers to find jobs where they can maximize their long-term productivity and pay.

A similar point applies to capital investment. Say you’re in charge of a factory and you have a idea for an overhaul that could double output with the same number of workers. However, the refit would be expensive and there’s a risk that the gains won’t pan out.

In a weak recovery, you (or your bosses) might be reluctant to give it a try. If the investment works, you might find the sales force can’t sell all the extra units you’re producing. If the investment doesn’t work, you worry that you (and whoever approved the investment) could lose your jobs. So you might decide that it’s safer to stick with current technology, satisfying small demand increases by hiring more workers and buying more of existing equipment.

In a boom, the calculation looks different. The entire industry is struggling to keep up with soaring demand so there’s little doubt you’ll be able to sell the extra units. You’re not too worried about losing your job because there’s an industry-wide shortage of experienced managers. Indeed, if the project fails but demand is still strong, the CEO might be eager for you to take what you’ve learned and try again.

In short, the fact that the unemployment rate is 4.9 doesn’t prove that the economy is operating at — or even close to — its maximum output. I suspect that if the Fed were to step on the gas, we’d see a sudden surge in labor force participation, investment spending, and total factor productivity. The economy might be at “full employment,” but that doesn’t necessarily mean it’s actually at full employment.

In a recent Vox post I tried to combine a headline designed to troll my least favorite website — “Pokémon Go is everything that is wrong with late capitalism” — with a serious policy argument. Unfortunately, I think the silly headline overshadowed the content of the article. So let me try to make the point again in a more systematic and (hopefully) clear way.

Suppose you have two regions, Region A and Region B. And suppose that a productivity shock causes output in Region A to grow much faster than output in Region B. Money starts to flow from Region A to Region B much faster than in the opposite direction. Logically speaking, one of five things has to happen:

(1) Workers can move from Region B to Region A.

(2) Companies can move jobs to from Region A to Region B.

(3) A government can tax people in Region A and give the money to people in Region B.

(4) Wages and prices can rise in Region A, causing people in both regions to buy fewer products from Region A and more products from Region B.

(5) Region B can enter into a deflationary spiral — with either falling prices or falling real output.

If Region A is Germany and Region B is Greece, then I’ve just provided a thumbnail explanation of the eurozone crisis. The problem in the Eurozone is that none of these five options is working very well:

(1) Language and cultural barriers make it hard for workers to relocate from Greece to Germany

(2) Language and cultural barriers — and perhaps a lack of relevant skills among Greek workers — make it hard for German companies to move jobs to Greece.

(3) The EU doesn’t have independent taxing authority, and of course the German government doesn’t want to tax Germans and send the money to Greece.

(4) Tight money policies by the European Central Bank have prevented significant price increases in Germany. (The Mercatus Center recently did a great report on this.)

That leaves (5), a deflationary spiral in Greece. Deflationary spirals are painful because because (a) workers really hate taking nominal wage cuts, and (b) Greece has a lot of euro-denominated debt, so falling prices means that Greece’s debt keeps growing as a share of GDP.

This kind of issue, incidentally, is why Milton Friedman warned against the creation of the euro back in 1997. He pointed out that if Greece and Germany have different currencies, then you can make the necessary adjustment by devaluing Greece’s currency — with much less severe consequences than the deflation spiral Greece is suffering today.

OK, so what does all this have to do with Pokemon Go? Well, let’s now change examples and say that Region A is “big American cities” like San Francisco (and their metropolitan areas) and Region B is “real America” — everywhere else. In the last couple of decades, powerful technological and economic forces — of which Pokémon Go is a small but evocative example — have caused productivity in regions like Silicon Valley to skyrocket. The rest of the country has gotten left behind.

The same analysis applies in this case:

(1) Strict regulations have caused severe housing shortages in cities like San Francisco and New York, limiting the number of people who can move there to find work.

(2) Agglomeration effects make it hard to relocate high-skilled, high-paying jobs outside of the biggest cities.

(3) The federal government does tax rich people (who are disproportionately in big cities) and give the money to others (think Social Security, Medicare, food stamps, farm subsidies, etc), but recent economic data suggests that these policies have not been sufficient to prevent the economies of big cities from diverging from the rest of the country.

(4) Thanks to tight-money policies by the Federal Reserve, inflation has been below the Fed’s 2 percent target, leaving little room for inflation in big cities to accomplish the needed economic adjustments.

(5) That leaves sluggish growth in the rest of the US. To be sure, the economies of Cleveland or Memphis aren’t in nearly as bad shape as the economy of Greece. But small cities and rural areas are suffering the worst economic recovery in decades even as big cities are booming.

To solve the problem, policymakers could attack any one of the first four issues: (1) They could deregulate housing. (2) They could invest more in education and infrastructure so more jobs could move to smaller cities (thought his could take decades and may or may not work). (3) They could raise taxes on the rich and make government transfer programs more generous. (4) The Fed could cut interest rates to boost economic growth generally.

Personally, I’m most enthusiastic about options (1) and (4). I’m skeptical that option (2) is going to work, though it’s probably worth trying. And I worry that these options may not be enough, and we may ultimately be forced to adopt some of option (3).

Ultimately the point here isn’t that Pokémon Go is bad. Obviously, consumer surplus is a good thing. The point is that our existing policies and economic institutions aren’t set up to deal with the emerging internet economy, in which a hugely disproportionate share of the wealth is created by a handful of big cities.

After my last post, reader Joe pointed out the parallels to Fred Hirsch’s classic book the Social Limits to Growth. I hadn’t read Hirsch’s book, but the similarity isn’t entirely a coincidence, since my argument relies heavily on the concept of positional goods that Hirsch invented.

I just finished reading the book, and it seems to me that he takes the the same basic idea — some goods have inherently limited supply — and takes things in a needlessly gloomy direction. Americans might have a chicken in every pot and a car in every garage, Hirsch argues, but it’s never going to be possible for everyone to have a vacation home in the Hamptons or send their kids to Harvard. For Hirsch, the growing importance of these kinds of positional goods will mean more frustration and social conflict.

But Hirsch never really makes a convincing case that we should view this as a problem we need to solve rather than just a fact of life that people should learn to live with.

One way I think Hirsch goes astray here is by adopting a definition of positional goods that’s far too broad. One of his key examples of how the struggle for positional goods is making us miserable is the growth in college education. He points out that as college educations have become more common, jobs that once required only a high school diploma now require a college degree. If he’d been writing today he might have pointed out that some jobs that once required a bachelor’s degree now require a master’s degree.

It’s not crazy to believe that the inflation of educational requirements is a problem, but it’s not clear what the positional good in this story is supposed to be. There’s a limited supply of seats at particular schools — like Harvard or Yale — but there’s no necessary limit on the number of college degrees awarded in general. And while having a degree from Harvard gives applicants a leg up in the job market, this effect becomes much weaker once you move a few rungs down the college ranking charts. Is the University of Northern Iowa more or less prestigious than the University of Minnesota — Duluth? Most employers don’t know or care, they just want to know you earned a college degree somewhere.

Hirsch also seems to imply that good jobs are a scarce positional good, and that people have to go to more and more educational effort to secure one of those scarce good jobs. But here again, good jobs are positional only at the very high end of the job market. By definition, there can only be 500 CEOs of Fortune 500 companies, for example. And there are a few professions like medicine where high barriers to entry make the jobs effectively positional.

But this isn’t really true of the economy as a whole. The number of companies isn’t fixed, nor is the number of positions in any given company. In many skilled professions, employers are constantly looking for more skilled practitioners they can hire.

That’s just one example of my broader point, which is that Hirsch sees positional goods everywhere and claims that as we get richer, human lives will become increasingly consumed by our struggle to gain positional goods. But I think that’s unduly pessimistic, because it’s almost always possible to opt out of positional competitions and still live a happy life. If you can’t afford to live in Manhattan, you can live a perfectly fulfilling life in Minneapolis or Omaha. If you can’t afford to send your kid to Harvard, she’ll do just fine at Penn State or the University of Northern Iowa. Nobody’s happiness depends on owning a Rembrandt.

People sometimes make themselves miserable by entering positional competitions they can’t afford to win. But the right lesson to draw here isn’t that positional goods are causing an inevitable social crisis — it’s that people should learn to keep positional goods in perspective. You can live a perfectly happy life without them.

Here are some thoughts on economic growth that aren’t yet coherent enough to be a Vox article…

1. Technological progress in a particular industry often has diminishing returns, and it’s possible to reach the point where it’s hard to imagine significant further improvements.

2. Clothing is the best example. Americans tend to run out of closet space before we run out of money to buy clothes. The fraction of our household incomes we spend on clothing has been steadily declining, even as we own dozens of outfits each.

3. As the production costs of clothing have continued to fall, a larger and larger fraction of the value people get from the clothing they buy — especially at the high end of the market — reflects social factors rather than economic ones. Someone might pay $40 for a T-shirt that cost $5 to produce because it carries the label of a famous designer.

4. For products like this, there’s little room for technological progress to lower clothing costs further. A 2x improvement in textile manufacturing productivity might reduce the shirt’s $5 manufacturing cost to $2.50. But we shouldn’t expect technological progress to reduce the $35 that goes to the designer and retailer. They’re selling exclusivity as much as they’re selling a piece of clothing.

5. A similar point can be made about food. The average American family has been able to comfortably afford more than enough food for many decades. And the quality and variety of food available to the average American has been steadily improving over time. Supermarkets now offer such a wide variety of high-quality, convenient food that there seems to be little room for further improvement. As with clothing, food prepared at home has become a smaller and smaller fraction of households budgets, even as the quality and variety of the food we consume has improved.

6. As ingredients have gotten cheaper and incomes have risen, we’ve spent less at the grocery store and more at restaurants. As with high-end clothing, most of the value of a restaurant meal comes from factors that can’t easily be improved by technology. Restaurants with human waiters tend to be more prestigious than restaurants that make you order at the counter precisely because people like to have other people serving them. If you figured out a way to serve fancy restaurant food from a vending machine, people would not see that as an improvement.

7. So families have been spending a shrinking share of their incomes on basic necessities like food and clothing. Where has their income gone instead? During the 20th century, there was a steady stream of new inventions — cars, televisions, washing machines, refrigerators, telephones, electric lighting, personal computers, and so forth — that soaked up peoples’ growing disposable income.

8. Over the last 30 years, this process has continued for information technology — we’ve seen the invention and widespread adoption of personal computers, gaming consoles, DVD players, smartphones, and so forth. VR headsets seem to be the next big thing. But outside of the IT sector, significant new inventions have been few and far between. Today’s kitchens have the same suite of labor-saving appliances — a refrigerator, oven, dishwasher, microwave, blender, and so forth — as the kitchens of the 1980s.

9. There has been a big debate about whether there has been a “slowdown in innovation” — with the implication that this represents a flaw in the way our economic system is working. But maybe we’re just running out of big problems that could be solved with technology.

10. One way to see this is to look at how wealthy Americans spend their money. A century ago, rich people could spend their money on a wide variety of technological luxury goods — electric lighting, telephones, automobiles, indoor plumbing — that substantially improved their quality of life. Today, very wealthy people have private jets, but otherwise it’s hard to think of examples of major technologies that are available to them but not to Americans with more modest incomes.

11. Instead, wealthy people spend money on two things that are not really amenable to technological improvement: positional goods (famous paintings, Manhattan real estate, Harvard tuition) and labor-saving services (nannies, housekeepers, chess tutors, art dealers).

12. As we get wealthier, I expect the previous point to describe the budgets of more and more Americans. People in large coastal cities are spending more and more money on housing in desirable locations — a positional good. And as the cost of food, clothing, furniture, and other goods has declined, child care costs have loomed larger and larger as a factor in the budgets of two-income households.

13. Education also fits this pattern. People are spending more and more money to send their children to fancy schools and colleges. And I while some aspects of the educational process can be improved by technology, elite schools mostly have the characteristics of a positional good. You can view a lot of MIT classes online for free, but people still seem to be willing to pay hundreds of thousands of dollars for their kids to be members of MIT’s undergraduate class — because what they’re really buying is access to an exclusive club.

14. I think we’re running out of room for technological improvements in most areas of economic life, with three big exceptions: IT, medicine and transportation. The IT part is obvious — smartphones were just invented recently, and VR seems likely to become a big market in the next few years. Obviously, if someone finds a cure for cancer, heart disease, or AIDS, that would create a tremendous amount of value. It’s also easy to imagine transportation technologies that people would pay a lot of money for: self-driving cars, affordable private airplanes, personal helicopters, supersonic airplane flights, space travel. It’s possible that physics or logistical constraints will prevent these from ever coming to fruition, but we can at least imagine ways these products could get better.

15. Energy is a third area where there seems to be a lot of room for progress — especially in solar panel and battery technology, as well as electric cars. But this is an interesting case because the primary selling point isn’t that they will make our lives qualitatively better so much as that they’ll help prevent a worsening of our collective living standards due to climate change.

16. This isn’t to say that there’s no room for further economic growth. Most American families can comfortably afford food, clothing, and shelter, but we’d all like these things to consume a smaller share of our incomes. More important, there are still some people in the United States and billions of people outside the United States who have yet to achieve the standard of living most of the people reading this post take for granted. It will take several more decades, at least, for the median income in countries likes India and Nigeria to reach American levels.

17. What this does mean, however, is that in the future most growth may be “catch-up growth,” in the sense that the economy will be focused on providing more and more people with the same standard of living that someone in the top income quintile of the United States enjoys today. That’s different from the 20th century, when even wealthy families could look forward to inventions (like air conditioning, televisions, and the internet) that would provide dramatic improvements in their standard of living.

18. This also means that we should expect a gradual slowdown in productivity growth rates. As people get wealthier, a smaller and smaller share of their household income will be devoted to goods and services that are amenable to technological improvement.

19. This could be seen as a pessimistic take, but the optimistic way to think about it is that Americans in the top half of the income distribution have arrived: we’re getting pretty close to the highest level of material comfort and security that it’s possible for a human civilization to have. Our children and grandchildren probably won’t enjoy a much higher standard of living than we do, but that’s mostly because it’s hard to imagine what a much higher standard of living would look like.

Update: I tweaked the example in point 3 after Saku Panditharatne convinced me that the original version was overstating my case.

Most of my experience with PR people involves them cluttering my inbox with unwanted pitches. But yesterday I had a PR experience that was so positive that I thought it was worth sharing. In an ideal world, this is how most PR professionals would do their job.

I was working on a story about mattress shopping, so I sent out a tweet:

If anyone has experience in the mattress business and is willing to talk about it, please email tim@vox.com.

— Timothy B. Lee (@binarybits) March 3, 2015

I was mostly hoping that a former mattress salesman would see it and give me an inside perspective on the negotiation process. But I was also interested in hearing from others with experience in the industry.

Just 90 minutes later, I got an email from Phil Krim, CEO of the mattress startup Casper:

Hi Tim-

How are you? Lindsay, our VP of Communications, sent me your tweet about
looking for mattress industry experts. I have had a great deal of
experience in the space. Anything I can do to help you?

Thank you.

Best,
PK

I don’t know Lindsay Kaplan, Casper’s VP of communications, but she did two things that are rare in the PR world.

First, she paid attention to what I was doing and figured out a way to help me out. I was looking for information about the mattress business. Casper is in the mattress business. So talking to Casper’s CEO was actually useful to me.

It’s astonishing how rare this is. There are dozens of companies in the mattress business. Presumably all of them would like favorable press coverage. Yet Casper was the only company that noticed my tweet and got in touch with me.

And the email was written in a way that made my job easier. Krim didn’t send me a wall of text explaining how great Casper mattresses were. He just let me know that he was available to talk.

Second, Kaplan stayed out of the way. Instead of sending me an email offering to put me in touch with Krim, she had Krim email me directly. That signalled that Krim was actually interested in talking to me and actually available to talk. And he was — when I responded with my phone number, he called me in a few minutes.

I’m way more likely to respond to a personal email from a potential interview subject than I am to a PR person trying to arrange an interview for someone else. Long experience has taught me that these third-party pitches are usually a hassle to deal with. The subject might not actually be that interested in talking to me, or it might take several hours to find space on his calendar.

Kaplan’s work got results. If Krim hadn’t contacted me, I wouldn’t have written about Caplan or its “bed in a box” competitors, because I simply didn’t know they existed. Krim’s email (and subsequent phone interview) convinced me to learn more. And my independent research found that these companies had a lot of satisfied customers. So I wound up adding a whole section discussing this product category.

If you’re a PR person, you should be doing your job more like Lindsay Kaplan. Instead of sending out press releases indiscriminately, learn about the specific reporters who cover the topics you’re working on and look for opportunities to help them out. And don’t get in the way. Whenever possible, pitches should come directly from the would-be interview subject.