Economic Welfare and the Allocation of Resources for Resilience

In 1959 Kenneth Arrow wrote "Economic Welfare and the Allocation of Resources for Invention"

Economic Welfare and the Allocation of Resources for Invention published as part of a 1962 NBER publication The Rate and Direction of Inventive Activity: Economic and Social Factors. Arrow's framework can be used to think about how we allocate resources to create a more resilient economy and society. Here I quote his main points in the context of critical resources and resilience rather than information and invention.

Arrow starts with the classic question of welfare economics: "to what extent does perfect competition lead to an optimal allocation of resources?" ... "competition insures the achievement of a Pareto optimum under certain hypotheses", but a core assumption is that "there is no uncertainty in the production relations and in the utility functions."

Uncertainty is one of the three classical reasons for the possible failure of perfect competition to achieve optimality in resource allocation. Indivisibilities and inappropriability are the other two. If firms can't relieve themselves of all risks (they can't), then any unwillingness or inability to bear risks will give rise to a nonoptimal allocation of resources since there will be discrimination against risky enterprises ... "The inability of individuals to buy protection against uncertainty similarly gives rise to a loss of welfare."

"Unfortunately, it is only too clear that the shifting of risks in the real world is incomplete ... There are a number of reasons why this should be so, but I will confine myself to one, of special significance with regard to invention resilience. In insurance practice, reference is made to the moral factor as a limit to the possibilities of insurance." .. "A fire insurance policy, even when limited in amount to the value of the goods covered, weakens the motivation for fire prevention. Thus, steps which may improve the efficiency of the economy with respect to risk bearing may decrease its technical efficiency."

"The main conclusions to be drawn are the following: (1) the economic system has devices for shifting risks, but they are limited and imperfect: hence, one would expect an underinvestment in risky activities; (2) it is undoubtedly worthwhile to enlarge the variety of such devices, but the moral factor creates a limit to their potential."

If any particular item of information critical resource has differing values for different economic agents, this procedure will lead both to a nonoptimal purchase of information resources at any given price and also to a nonoptimal allocation of the information resource purchased.

"It should be made clear that from the standpoint of efficiently distributing an existing stock of information resources, the difficulties of appropriation are an advantage, provided there are no costs of transmitting information resources, since then optimal allocation calls for free distribution. The chief point made here is the difficulty of creating a market for information resilience if one should be desired for any reason."

Information Resources should be transmitted at marginal cost, but then the demand difficulties raised above will exist. From the viewpoint of optimal allocation, the purchasing industry will be faced with the problems created by indivisibilities; and we still leave unsolved the problem of the purchaser's inability to judge in advance the value of the information resource he buys. There is a strong case for centralized decision making under these circumstances.

To sum up

Arrow writes "we expect a free enterprise economy to underinvest in invention and research resilience (as compared with an ideal) because it is risky, because the product can be appropriated only to a limited extent, and because of increasing returns in use. This underinvestment will be greater for more basic research more infrequent events. Further, to the extent that a firm succeeds in engrossing the economic value of its inventive resilience activity, there will be an underutilization of that information as compared with an ideal allocation."

To paraphrase in the context of pandemic preparedness: we expect the market to underinvest in necessary resources like masks and ventilators because (1) it is unknown when they'll be necessary (due to uncertainty), (2) if they become necessary companies won't be able to charge prices sufficient to make a century of mask-stockpiling worthwhile, and (3) masks, like vaccines, have increasing returns in use (the more people wearing them the better for everyone). Further, to the extent that a firm succeeds in price gouging, not as many masks and ventilators would be distributed as we would want.

Over 90 percent of the masks and 80 percent of the PPE used in American hospitals today are not manufactured in America. Arrow, who set out the general equilibrium equations showing that competitive markets are efficient, gives us a framework for thinking about a more thoughtful industrial policy that relies a little less on an "invisible hand" and more on common sense and strategy.

Government has a revenue problem, not a spending problem

I do not accept the premise of this blog post.

Government debt today is not a bigger problem than collective under-investment in our health, education, infrastructure, and environment. And fortunately, there does seem to be at least a short-term consensus that government should spend what is necessary to protect us during this crisis.

However, policymakers are already preparing for the debt battles to come. Robert Rubin recently wrote that "Congress should commit now to address -- when conditions allow -- the increase in debt as a share of our economy, which was already seriously worsening before the crisis." He recognizes that "There will be ample room to increase revenues, on a highly progressive basis, for example, by increasing corporate taxes, restoring individual rates, repealing pass-through preferences and imposing a financial transactions tax."

Every 18 months or so the Congressional Budget Office publishes an Options for Reducing the Deficit report. It shows how much the CBO estimates that government could save by implementing various policies. A quick glance makes it pretty clear that anyone pushing for deficit decreases without pushing for at least some tax increases is probably not being entirely serious.

The top 10 options for raising revenues result in savings almost four times as great as the top 10 options for cutting spending. Not to mention that policies to increase revenues generally don't have much impact beyond slightly smaller numbers in the personal finance spreadsheets of comfortable Americans, while policies to decrease spending generally affect the least fortunate or do not solve problems -- for example cutting Medicaid costs and shifting parts of it onto state budgets.

PS: certain economists might say that the CBO estimates do not account for macroeconomic feedback and incentive effects. This is correct. However, those impacts are much less conclusive than is assumed in mainstream macroeconomic models, and in fact sometimes go the opposite way.

Additionally, the CBO document (and therefore the chart above) does not include estimates for recent larger revenue plans such as a wealth tax, which could raise more than $3 trillion over 10 years.

PPS: Olivier Blanchard basically said as much in his concluding remarks in November 2019 House testimony, Reexamining the Economic Costs of Debt -- government has a revenue problem, not a spending problem.

M2 vs. inflation

The federal government's $2.2 trillion stimulus package* has raised the question of whether we'll face excessive inflation over the next months and years.

One preliminary data point: from the start of 1983 through March 2020, there has been a slight negative correlation (-0.16) between year-over-year changes in the broad money supply (M2) and inflation as measured by the consumer price index (CPI). The following charts all show the same M2 and CPI data -- by levels, changes, and a scatter plot of changes.

The alleged crowding out effect

Let's talk about crowding out. Before our economy recovers, some prominent economists -- America's "serious people" -- are going to be speaking about the risk of government debt crowding out productive private investment in America's economy. They will sound serious and maybe even credible. If they are listened to, millions of Americans will suffer unnecessarily.

Jamie Powell at FT Alphaville introduces the topic nicely:

If you've had the misfortune of having to learn economics, you may remember the theory of the "crowding out effect" from your studies. Popularized in the 1970s, the idea is that an increase [in] public sector investment, and therefore borrowing, has the rather undesirable effect of displacing the private sector's planned borrowing and investment.

Essentially, the theory goes like this: If government finances programs by issuing bonds (i.e., with higher deficits), then interest rates will rise, the projects the private sector is planning will become less profitable due to a higher cost of capital, and some of those projects won't go ahead. The result is lower overall private sector investment, which means lower growth and productivity, ultimately resulting in lower long-run GDP. Fine in theory (with certain very strong assumptions, to be described in a later post), but it doesn't describe reality.

As one example of the fantasy foundations of crowd out theory, Powell quotes a new working paper by Enrico Moretti, John Van Reenen, and Claudia Steinwender. They find that public sector R&D expenditure -- directly or through subsidies -- leads to an increase in private sector spending in R&D, a "crowding in" effect. To those who have followed the work of Mariana Mazzucato and Bill Janeway (and, more recenty, the Council on Foreign Relations), this result will not come as a surprise. A quick glance at the development of industries such as information technology (core technologies of the internet, computers, iPhones, etc. all came through government-sponsored research and were supported by government as an early-stage customer), pharmaceuticals (over 75 percent of new molecular entities -- the truly novel drugs -- are discovered by government-funded researchers), and trains in the 19th century (the railroad boom was made possible in part by government granting 9 percent of US land to corporations) makes clear that there is more to crowd out than the neoliberal story of supply and demand for a fixed amount of national savings in an economy operating at full capacity in which the technological innovations that lead to productivity growth and the corresponding increases in per capita wealth magically occur .

I have nothing to add on those points beyond what is described by economists such as Mazzucato and Janeway. But here I point out one more thing: crowd out theory was created a couple of centuries before it was popularized in the 1970s, when money and taxes were very different from today. Here is Thomas Piketty summarizing 18th tax receipts in Capital and Ideology (p. 364)

In the centuries that followed these sums [tax receipts] would grow spectacularly, mainly due to the intensifying rivalry between England and France: both countries were taking in 600-900 tons of silver in 1700, 800-1,100 tons in the 1750s, and 1,600-1,900 tons in the 1780s, leaving all other European powers far behind. Importantly, Ottoman tax receipts remained virtually unchanged from 1500 to 1780: barely 150-200 tons.

Two points from this:

1. When people say that government spending will "crowd out" public investment, they are reasoning from an intellectual framework developed in the 18th century, when countries measured their tax receipts in tons of silver. That framework has not been updated for today's market-based model of banking and money creation.
2. The empire with the least amount of government spending and hence "crowding out" (even when the theory was more relevant) was the declining one. I don't explore this point further in this post, but there's a chance it's not insignificant.

Following up on the first point, here is one example of the differences. When the English Crown took in ~1,000 tons of silver from its subjects in 1760, there were 1,000 fewer tons of silver for (a) England's banks to issue currency, (b) its merchants to invest in sailing to India for tea and spices, and (c) its early industrialists to invest in building new textile mills. If King George collected silver to build a new summer palace at the expense of a new textile mill, crowd out was real! Today, if the US Treasury decides to send $1 trillion to Americans, there are not 1 trillion fewer dollars for productive private investment (and, I would argue, this is a better use of money than a new summer palace for George, but that is beside the point here). [I'm still trying to find out if anyone is operating with an updated framework for these types of questions].

Of course, we have to manage our nation's finances wisely. We (probably) can't exceed our productive capacity indefinitely. Eventually, inflation would increase and/or we would face other issues. But that is a discussion for later.

The point here is, if someone says crowd out with a straight face, do not take them seriously. Unless, of course, that person is a "serious person" who many people listen to. In that case, take him/her very seriously: she's dangerous.

Homebase and consumer spending

Two things today:

1. Homebase

Discovered Homebase data. Homebase provides a scheduling and time tracking tool, and they've made a dataset publicly available that comprises 60,000 US businesses and 1 million hourly employees. It's the most impressive high frequency economic dataset I've seen, and the numbers are jarring. Relative to median hours for a given weekday during Jan. 4 - Jan. 31, 2020, 66 percent fewer local businesses are open and 75 percent fewer hourly employees are working as of April 12, 2020. The beauty and personal care industry has been the hardest hit, with only 4 percent of firms still open on April 12, followed by leisure and entertainment (84 percent). Groups such as the Rustandy Center at Chicago Booth have conducted some useful analyses with this data.

The chart below, provided by Homebase, shows a time series of hours worked by hourly employees through April 11:

Central Bank Liquidity Swaps, or Europe to Japan

I've been re-reading Adam Tooze's Crashed and so much is relevant. Today, central bank liquidity swaps.

During the financial crisis, the dollar funding needs of foreign banks (in particular European banks) would have quickly overwhelmed the foreign exchange reserves of their home-country central banks. To deny these commercial banks liquidity assistance could have been disastrous, but lending to the most fragile foreign banks without adequate collateral would have exposed the Federal Reserve to significant risk. As a solution, the Fed provided liquidity assistance to international banks through swap lines set up with other central banks. They were not a new invention in 2007, but the Fed used them on an unprecedented scale.

Tooze writes

“from 2007 the Fed repurposed an instrument that was first developed in the age of Bretton Woods. To manage the fixed currency system in the 1960s the central banks had developed a system of so-called currency swap lines that allowed the Fed to lend dollars to the Bank of England against a reverse deposit of sterling in the accounts of the Fed. Having gone out of use in the 1970s, the swap lines had been briefly revived in 2001 to deal with the aftermath of 9/11. In 2007 faced with the implosion of the transatlantic banking system, they were repurposed and expanded on a gigantic scale to meet the funding needs not of sovereign states but of Europe's megabanks." (p. 209-210)

The total amount outstanding on these dollar swap lines peaked at over $580 billion in December 2008, with over $310 billion outstanding with the European Central Bank. The swap lines prevented a euro-dollar or sterling-dollar crisis. “What the Fed had done for money markets, the central banks now did for the global provision of dollar bank funding. They absorbed the currency mismatch of the European bank balance sheets directly onto their own accounts.”

At the start of the financial strains due to the coronavirus pandemic in February 2020, the Federal Reserve had standing swap arrangements with the central banks of Canada, England, Europe, Japan, and Switzerland. On March 19 it added temporary arrangements with the central banks of Australia, Brazil, Denmark, Korea, Mexico, New Zealand, Norway, Singapore, and Sweden.

As of April 9, 2020, over $396 billion was outstanding. Three charts, using data from the New York Fed, and thoughts below.

Figure 1: This crisis struck funding markets much more quickly than the 2007-08 crisis. 

Buybacks: OK in Theory; Bad in Practice.

In theory, if a company is unable to invest a portion of its retained earnings at the same level of (risk-adjusted) returns available elsewhere, it is better for its shareholders and the economy as a whole for that company to return that money to the shareholders so that they can invest in higher-yielding projects, leading to a more efficient allocation of capital in the economy. The reality, however, is different. When a majority of companies is repurchasing shares at the expense of investment in research and development, and in some cases borrowing money to repurchase shares, it suggests that something is amiss.

S&P 500 firms have repurchased over $5.4 trillion worth of shares since 2009. Well-positioned staff have been profiting at general shareholders' expense while introducing fragility into the system. Despite the theoretical benefits of buybacks in certain circumstances, the risks created by them have become obvious during the coronavirus pandemic. Some of the companies requesting federal loans and grants were the ones aggressively repurchasing shares during the past few years. There is anger because nearly 17 million American have filed for unemployment in the past three weeks and, as William Lazonick mentioned, "If companies are paying dividends and doing buybacks, they do not have to lay off workers."

In Why Stock Buybacks Are Dangerous for the Economy, Lazonick, Mustafa Erdem Sakinç and Matt Hopkins explain that "When companies do these buybacks, they deprive themselves of the liquidity that might help them cope when sales and profits decline in an economic downturn." During the drafting of the first coronavirus response bill, The New York Times summarized the awkwardness of companies like the major airlines asking the federal government for bailout money after spending $19 billion repurchasing shares over the last three years. President Trump said, "I don't want to give a bailout to a company and then have somebody go out and use that money to buy back stock in the company and raise the price and then get a bonus. OK?" In response to buybacks' financial and political downsides, the Coronavirus Aid, Relief, and Economic Security (CARES) Act signed March 27 precludes certain businesses that receive relief loans through the Act from repurchasing equity securities for up to "12 months after the date on which the direct loan is no longer outstanding."

The negative aspects of buybacks should not come as a surprise. In 2014's Profits Without Prosperity, Lazonick describes how "the corporate resource allocation process is America's source of economic security or insecurity". The shift from the retain-and-reinvest approach, in place between WWII and the 1970's, to the current downsize-and-distribute regime reflects the shift from an economic model in which corporations create value to one in which they extract it. Not surprisingly, the value creation setup is more stable and sustainable. The value extraction setup contributes to the employment instability and income inequality (and slower productivity growth) evident today.

In January 2018, Dan McCrum examined whether US companies' huge repurchases made sense, considering that investment in development was stuck at pre-crisis levels and stock prices were high (buybacks make more sense when a corporation considers its shares to be undervalued by the market). Given repatriation rule changes in 2017's TCJA, it was clear that buybacks were going to remain near record levels. He examines five charges :

Charge	Verdict
1. Staff benefit more than shareholders	Guilty
2. Opportunity for manipulation	Not Guilty
3. Executives can be a bad judge of value	Guilty
4. Buybacks reduce real investment	The Jury is Out
5. Encouraging fragility	Guilty as Hell

Following up on the charge that buybacks encourage fragility, Lazonick, Sakinç and Hopkins (2020) highlight the issue with the corporate buyback boom: up to 30 percent of buybacks in 2016 and 2017 were financed by corporate bonds. These "debt-funded" payouts, encouraged by low interest rates, are a form of financial risk-taking that "can considerably weaken a firm's credit quality." (IMF GFSR)

The Atlantic summarizes the investor-level and economy-level impacts: "The proliferation of stock buybacks is more than just another way of feathering executives’ nests. By systematically draining capital from America’s public companies, the habit threatens the competitive prospects of American industry—and corrupts the underpinnings of corporate capitalism itself."

-------------------------------------------------------------------------------------------

Other things to look into:

There are of course factors within the legal framework that help explain why buybacks have increased. The main example is the SEC's Rule 10b-18, which creates a "safe harbor" in which companies are free from risk of liability for stock price manipulation as long as they follow conditions laid out in the Rule.

But I wonder to what extent the economic situation has led to more buybacks. Decreased productivity and population growth, among other factors pulling the "natural rate of interest" down, might indicate that the marginal product of capital is lower today than it has been at any other point since the end of the 1930's, when Alvin Hansen was worried about secular stagnation. Companies might simply not have investment opportunities that stand up to the ROI expectations ingrained in their executives' minds since their 1990's MBA programs where their professors were teaching corporate finance classes and the 10-year Treasury yield of around 7 percent was 10 times higher than it is today. Part of this is a question of capital accumulation and how our society structures itself for the Japanese experience of low growth -- which does not have to be bad!

Lastly, to the extent that federal spending on research crowds in private spending on research and development, how much of the growth in buybacks is a result of the public sector stepping away from its active role in promoting growth? The work of economists such as Enrico Moretti, Mariana Mazzucato (The Entrepreneurial State), and Bill Janeway (Doing Capitalism in the Innovation Economy) shows how government spending on research (and government's ability to create a market for technologies before they are ready for commercial success) "crowds in" private sector spending on research and development.

Bank capital and German Google searches

Two thoughts today.

1. Are the banks really safer

People are saying the post-financial crisis regulations have done their job and the banking sector is much safer now. In some ways yes, banks now have better liquidity positions (and the Fed acts more quickly when companies need cash), stress tests, resolution plans, more active risk management departments, and a slew of other new regulations. But the most important metric, bank equity levels, are not so much higher.

The Federal Reserve Bank of Minneapolis' Plan to End Too Big to Fail calculated that to keep the probability of a bank crisis in the next century below 10 percent, banks would need capital amounting to 38 percent of assets. Here is how bank capital has progressed against that since 2007: