Monday, May 17, 2021

What is Behind People's Views on Tax Policy?

When economists study taxation, they typically separate two issues: one is the distributional issue of which groups are paying more or less; the other is the ways in which taxes reduce efficient economic incentives for work, savings, investment, innovation, and so on on. Today is the deadline for when US individual income tax returns are due with the federal Internal Revenue Service as well as with state-level income tax authorities. According to the research of Stefanie Stantcheva, most of those taxpayers focus almost entirely on the distributional question, not the efficiency question. 

Stancheva discusses the topic as part of an interview with David Cutler on the occasion of winning the 2020 Elaine Bennett Research Prize, which is "awarded every two years to recognize and honor outstanding research in any field of economics by a woman not more than seven years beyond her Ph.D." (CSWEP Newsletter, 2021, 21:1, "Interview with Bennett Prize Winner Stefanie Stantcheva"). The underlying research paper by Stantcheva being discussed here is "Understanding Tax Policy: How Do People Reason" (November 2020, NBER Working Paper 27699).  Stantcheva reports in the interview: 

Consider the example of tax policy. Is it that people have different perceptions about the economic cost of taxes? Is it that they think differently about the distributional impacts that tax changes will have? Or is it that they have very different views of what’s fair and what’s not? Could the reason be their views on the government—how wasteful or efficient they think the government is? Or is it purely a lack of knowledge about how the tax system works and what inequality is?

I think of these factors as my explanatory or right-hand side variables. I can decompose a person’s policy views into these various components. What I find is that for tax policy, a person’s views on fairness, and who’s going to gain and lose from tax changes completely dominates all other concerns. This is followed by a person’s views of the government. How much do they think the government should be doing, how efficient is it, how wasteful is it, how much do they trust it? Efficiency concerns are actually quite second-order in people’s minds when it comes to tax policy.

These are all correlations. To see what’s actually causal and what could be shifting views, I show people these short ECON courses, which are two - or three-minute-long videos which explain how taxes actually work. The videos take different perspectives. Although they’re neutral and pedagogical, they don’t tell people what taxes should be or what’s fair or not. They just explain the how taxes work from one perspective. For instance, one version focuses only on the distributional impacts of taxes - who gains and who loses. The other version focuses only on the efficiency costs. Then there is the economist treatment, which shows both and emphasizes the trade-off between efficiency and equity. One can replicate this approach for the other policies such as health policy or trade or even climate change, which all have efficiency and equity considerations.

What I find for tax policy confirms the correlations. What shifts people’s views most is to see the distributional impacts of taxes, not at all the efficiency consequences of it. Even if you put it together and emphasize the trade-off, it’s still the distributional considerations that dominate and outweigh the efficiency concerns.
Distributional concerns about taxes matter to me, as well! But even if you generally agree on the idea that taxes should weigh more heavily on those with higher incomes or different wealth, it doesn't help to distinguish between different ways this might be be done. 

For example, one might have higher marginal tax rates on those with higher income levels. One might reduce the value of tax deductions, like deductions for mortgage interest or state and local taxes, that tend to benefit those with high incomes more. One might insist that taxes be paid on currently untaxed fringe benefits, like employer-purchased health insurance, because exempting those benefits from income tax will provide greater benefit to those with higher incomes. One might want to alter rules that let people make tax-free contributions to retirement accounts, on the grounds that reducing taxes in this way will tend to benefit those with higher incomes. One might think about expansion of "refundable" tax provisions that help the working poor, like the Earned Income Tax Credit and the child tax credit. One might alter corporate taxes, on the theory that this would affect shareholders and top managers more than it will affect wages paid to average employees. One might alter the way in which capital gains are taxed, and one might want to distinguish between capital gains on owner-occupied housing, on family businesses, or on financial assets. One might alter the rules that let high-income people pass wealth to future generations. For example, the current rules are that when financial assets which have gained in value over the lifetime of the owner, those previous gains are not taxed when the asset is passed through an estate. One might want to change other rules on what assets can be passed to the next generation, including other aspects of the estate tax to rules about intergenerational giving, along with rules about using life insurance policies or charitable foundations to pass income between generations. 

For those readers who are sunk deepest into distributional thinking I suspect the honest response to this list is something like: "I'm against anything that would raise my taxes by a single dime, but I'm for anything that would only be paid by high-income, high-wealth individuals, and I don't care about how it affects their incentives." Of course, in a US economy where government debt is ascending to unprecedented levels even before we try to address the middle-term projected insolvency of Social Security and Medicare, that response is just an abdication of analysis. 

Friday, May 14, 2021

Interview with Christopher Pissarides: Unemployment and Labor Markets

Michael Chui and Anna Bernasek of the McKinsey Global Institute interview Christopher Pissarides (Nobel, '10) "about how he developed the matching theory of unemployment, how COVID-19 affected his research, and what might be in store for labor markets after the pandemic" (May 12, 2021, "Forward Thinking on unemployment with Sir Christopher Pissarides"). At the website, audio is available for the half-hour interview, along with an edited transcript, upon which I will draw here.

As a starting point, it's useful to remember that labor markets always have, at the same time, both unemployed workers who are looking for jobs and employers who have job vacancies. For example, the US economy had about 9.7 million unemployed workers in March 2021, and at the same time, employers were listing 8.,1 million job vacancies. Indeed, there was a stretch from April 2018 to February 2020--not all that long ago, where the number of job vacancies for the US economy exceeded the number of unemployed in the monthly data.

At first glance, this combination of millions of job vacancies and millions of unemployed seems like a puzzle. Why don't the unemployed just take the vacant jobs? This is where Pissarides comes in. He has emphasized that unemployment and hiring are not just about raw numbers, but involve a matching process. Most employers, most of the time, don't just hire the first person who walks in through the front door, but instead are looking for a good match for the skills they desire.. Most workers, most of the time, know that they could get certain kinds of low-wage work  pretty quickly, if that was what they wanted, but they instead are looking for a good match for the skills they can offer. Policies to address unemployment, or to help unemployed workers, need to be considered in context of this matching process. Here's Pissarides: 

[U]nemployment is a very serious problem that I think governments should always be dealing with. It’s a cause of poverty, of disenfranchisement from the labor market, of misery.  ... [B]efore we did that work, people were thinking of unemployment as a kind of stock of workers, as a number of workers if you like, who could not get a job. They would start from the top end of the market and say, “This is how much output this economy needs, that’s how much is demanded. Then how many people do you need to produce that output?” Then you would come up with a number. And then they would say, “Well, how many workers want jobs?” If there are more workers that want jobs, you call the difference unemployment. ..

What we did was to start from below, saying the outcomes in the labor market are the result of workers looking for jobs, companies looking for workers.The two need to come together. They need to agree that the qualifications of the worker are the right ones for the firm. That once the firm has the capital, that worker needs to make the best use of his or her skills. That unemployment insurance policy might influence the incentives that the worker needs to take a job. The tax policy might influence the incentives of the company. Once you open the field up like that, it gives you unlimited possibilities for research in that area and working out the impact of these different policies or different features of the labor market on unemployment. ...

[T]he time that it takes to find that job depends on how many jobs are being offered in the labor market, what types of skills firms want, what incentives the worker has to accept the jobs, what’s the structure of production, the profit that the firm expects to make, conditions overall in the market. All those things influence the duration of unemployment. Therefore you could study there—how long does the worker remain unemployed? What could influence that duration? What could make it shorter? What would make it longer if you did

certain things? On that basis, you derive good policies towards unemployment, and they are still the policies that governments use, in fact widely, to work out how long people remain unemployed and what the implications of their unemployment are.

So what are some insights that emerge from this approach? The unemployment that results from this matching process can be a good thing in an economy:

[L]ower unemployment than what might exist is not always a good thing for the labor market, because some unemployment is good because of the matching problem. If a worker becomes unemployed, or if a new worker leaves school, a person leaves school, gets into the labor market, is a new worker, it wouldn’t be a good idea to accept the first job that is offered on day one and get into it. Because it may not be the job that would bring out the best productivity from that worker, or the job that that worker would like best. Now, you might say it’s obvious, and I now think it is, but when we were working on it, this didn’t exist.

On designing unemployment insurance: 

[I]f you offer unemployment compensation, which is necessary to reduce poverty caused by unemployment, then you have to be careful when you’re doing that, because if you just offer it unconditionally, it’s going to create disincentives for people to take jobs, and it’s going to lengthen the duration of unemployment. Therefore it’s going to increase your unemployment incidence. You are going to see more people unemployed, because they stay unemployed longer, collecting benefits. Now, that’s been exploited a lot by politicians. I don’t agree with that way, that they say, we have to cut benefits because of these incentives.

A better way of dealing with it is to say, we need to structure our unemployment compensation policy in such a way that it deals with the poverty issue, but at the same time it doesn’t create those disincentives that you might get if you offer it unconditionally. The leading countries that develop policies that give exactly the answer to the question I’ve just posed, how to structure it, are mainly the Scandinavians—Denmark, Sweden, Norway. And other countries have followed them now, and most of them do follow this advice of structuring the benefit in such a way that the incentives are not harmed very much when you are dealing with the poverty issue of unemployment.
On retraining programs:
Retraining needs to be provided by companies, because they’re the ones that would know in what to train and to what extent. Now, for training to succeed, however, it has to be funded from outside as well, because no company, except for the very big ones, I guess, will take on workers on expensive training programs if they are running the risk that some other company will come and take their workers away from them after they get trained. There is this poaching problem. ...

Then the other issue is that training succeeds when the worker owns the training, in the sense that the worker is doing the training not because someone forced that worker to do the training, but because they believe that it’s good for them and their career, and it’s going to give them career progression and a pay raise. ... Somehow maybe part of the amount should be given to the worker, then the worker chooses how to spend it. They cannot take it as money, but they could draw on a fund, a training fund. Singapore has a very good scheme like that. I think it’s called SkillsFuture. Some other countries are introducing it. It’s not an easy thing, but we have enough experience now to know how to plan those kinds of training support schemes.

Wednesday, May 12, 2021

Political Economy of the Pandemic Response

If economics-minded policy-makers rules made decisions in response to the pandemic, what might they do differently, and why? Peter Boettke and Benjamin Powell suggest some answers to that question in "The political economy of the COVID‐19 pandemic" (Southern Economic Journal, April 2021, pp. 1090-1106). Their paper leads off a symposium on the topic. I'll list all the papers in the symposium below. I'm told that they are all freely available online now, and for the next few weeks, so if you don't have library access to the journal, you might want to check them out sooner rather than later. Boettke and Powell write: 
[F]rom the perspective of promoting overall societal well‐being, we believe that governments in the United States and around the world made significant errors in their policy response to the COVID‐19 pandemic. ... [A] political economy perspective challenges the assumptions of omniscience and benevolence of all actors—politicians, regulators, scientists, and members of the public—in response to the pandemic. We live in an imperfect world, populated by imperfect beings, who interact in imperfect institutional environments ...
What are some ways in which pandemic policies based in micro theory and welfare economics might differ from the policies actually used? The potential answers seems to me both of interest in themselves, but also a good live subject for classroom discussions and writing exercises.  

For example, when discussing the subject of how policymakers should respond to negative externalities, a general principle is that there are a wide array of possible responses, and the least-cost response should be selected. If one thinks of society as divided into the elderly and non-elderly, for example, it seems plausible that the lowest social cost response to COVID-19would involve restrictions on  the elderly. Boettke and Powell write: 

The activities of the young and healthy impose a negative health externality on the old and infirm. But it is equally true that if the activities of the young are restricted because of the presence of the old and infirm, this latter group has imposed a negative externality on the young and healthy. If transactions costs were low, the Coase theorem would dictate that it would not matter to which party the rights to activity or restriction were assigned, as bargaining would reach the efficient outcome. However, in the case of COVID‐19, and large populations, it is quite clear that transactions costs of bargaining would be prohibitive. Thus, the standard law and economics approach would recommend assigning rights such that the least cost mitigator bears the burden of adjusting to the externality. In the case of COVID‐19, it is clear that the low opportunity cost mitigators are the old and infirm. Thus, Coasean economics would recommend allowing the activities of the young and healthy to impose externalities on the old and infirm, not the other way around. Lockdowns and stay at home orders get the allocation of rights exactly backwards and result in large inefficiencies because costs are disproportionately borne by the high cost mitigators.
Another common insight from economics is that those closest to the externality typically know the most about how to respond. In the case of pollution control, for example, there is a standard argument for using a pollution tax or marketable pollution permits, rather than trying to draw up command-and-control rules for  every smokestack or pollution source. Have those creating pollution bear the cost, and they will have an incentive to find ways to reduce those costs. 

Of course, the response of most states and  localities to COVID-19 was very much a command-and-control response, with extensive and ever-changing rules about outdoors and indoors, about restaurants, parks, and churches, about what businesses or school could be open under what conditions. As the authors write: "The thousands upon thousands of varied restrictions are too numerous and diverse for us to comprehensively categorize here. But their sheer number and variability make it obvious that these command and control regulations are not in any way promoting a cost minimizing form of transmission mitigation." The alternative might have been to categorize activities according to their chance of spreading COVID-19, and then impose a tax for participating in such activities.  
The marginal costs of reducing risk‐generating activities are really just the inverse of the subjective marginal benefits of engaging in myriad social interactions in the market place, civil society, families, politics, religious communities, and recreation. No regulator is going to know the value of these diverse activities to those engaged in them. Economists have long appreciated that, in the presence of heterogenous mitigation costs, command and control regulation of much simpler pollution mitigation is less efficient than a pollution tax, because firms know their mitigation costs better than regulators. That informational asymmetry between the economist regulator and people regulated is even greater in this case. Thus, an efficiency maximizing economist policy advisor would recommend leaving people free to choose activities for themselves, while imposing a tax on activities set to reduce the marginal benefit of engaging in activities, proportional to increased risk of COVID‐19 transmission.
Another policy option would be for the government to subsidize activities that would reduce the spread of the externality: for example, "government funding to expand hospital capacity and the purchase of supplies and equipment, and research funding to speed the discovery of new medical treatments and vaccines. They could also include removal of regulatory barriers that impede medical capacity and the development of medicines and vaccines. Unlike efficient policies related to the mitigation activities that risk disease transmission, governments have undertaken these policies to varying degrees." 

But the interesting observation here is that the size of government activities that focused directly on reducing the disease was dwarfed by the size of payments the government made to affected individuals and businesses. For example, the government put $10 billion into the Warp Speed program to produce vaccines and guarantee that certain volumes would be purchased, but has spent trillions of dollars--more than a hundred times more--on payments that do not directly reduce the risk of transmission. 

A final example involves decisions about who would get the vaccine first. For example, should it go to "essential workers"? Or to the elderly or those with greater vulnerability to the disease? Who defines these groups? Will lotteries be involved at certain stages? By the time all the rules are argued over,  spelled out, and  then enforced, an obvious question (to economists) is whether a more flexible and market-oriented system might work better. The authors write: 
Even if policymakers cared more about the welfare of the people that guidelines currently prioritize for vaccination, they could design policy better than the CDC guidelines by allocating a re‐sellable right to receive the vaccination, rather than the vaccination itself. Those prioritized individuals who resell the right will, through their actions, indicate that they are even better off, and the transfers of the right to higher value vaccinators would promote greater efficiency too. No politicians are considering such policies.
What's interesting to me is not that the economics answers here are obviously "right"--one can certainly point out tradeoffs that would be involved--but that the tradeoffs were barely noticed or discussed as real options.  Boettke and Powell point out some underlying issues here of political economy. For example, public health officials "re not necessarily untruthful, but they will be biased against committing an error of over optimism—no forecast or treatment protocol or vaccine will be championed that underestimates the downside risk. Better for them to commit errors of over‐pessimism." 

The combination of media and public attention in the social media age does not seem predisposed to calm consideration of tradeoffs, either. Instead, tradeoffs are typically presented as involving "good people," who are judged leniently , and "bad people," who are judged harshly. The authors write: 
One implication is that fair and balanced reporting may be too boring to grab the attention of the median listener/viewer/reader. Rather than nuanced and subtle discussion of trade‐offs, and the calm calculation of risk, we get extreme projections of nothing here or catastrophe awaits. And, of course, those incentives for attracting an audience have grown more intense in the last decade with traditional print media competing with online sources. ... 

Both politicians and the mainstream media have kept much of the populace in such an alarmed state throughout the pandemic, which has allowed both paternalistic interventions and created bottom up parentalist demands for such interventions, which have nothing to do with efficiently correcting a market failure.
Here's the full Table of Contents for the symposium. Again, I'm told that all the articles will be open access for the next few weeks:

Is the Pandemic Worse in Lower- or Higher-Income Countries?

It seems obvious that the COVID-19 pandemic must be worse in lower-income countries. After all, it seems as if the opportunities for social  distancing must be lower in urban areas in those countries, and the resources for everything from protective gear to hospital care must be lower. There are certainly cases where the pandemic has hit some areas hard outside of  high-income countries: for example, the current situation in India, or the city of Manaus in Brazil that suffered a a first wave, and then suffered a second wave with a new variant of COVID.  

But that said, Angus Deaton (Nobel '15) makes a case that the areas outside the high-income countries of the world have, as a group, been less affected by the pandemic in "Covid-19 and Global Income Inequality" (Milken Institute Review, Second Quarter 2021, pp. 24-35). As a starting point for his argument, consider this figure, which shows countries with higher per capita income have tended to have higher per capita COVID-19 deaths. 

Deaton discusses this figure from a variety of angles,  including the possibility that COVID-19 is less well-measured in lower-income countries. But he argues that a number of other factors may help to explain the pattern of higher COVID-19 deaths in higher-income countries. 

The low number in low-income countries has been linked by Pinelopi Goldberg and Tristan Reed to (the lack of) obesity, to the smaller fraction of the population over 70 and to the lower density of population in the largest urban centers.

Another alternative is to focus on demography. Patrick Heuveline and Michael Tzen provide age-adjusted mortality rates for each country by using country age-structures to predict what death rates would have been if the age-specific Covid-19 death rates had been the same as the U.S. The ratio of predicted deaths to actual deaths is then used to adjust each country’s crude mortality rate. This procedure scales up mortality rates for countries that are younger than the U.S. (Peru has the highest age- and sex-adjusted mortality rate) and scales down mortality rates for countries like Italy and Spain (which had the highest unadjusted rate) that are older than the U.S.

If Figure 1 were redrawn using the adjusted rates, the positive slope would remain, though the slope showing the relationship between death rates and income would be reduced from 0.99 to 0.47 — that is, the relationship would hold but would be less pronounced. ...

[P]oor countries are also warmer countries, where much activity takes place outside, and there are relatively few large, dense cities with elevators and mass transit to spread the virus. It is also possible that Africa’s long-standing experience with infectious epidemics stood it in good stead during this one. People in countries with more-developed economies consume a higher fraction of income in the form of personal services, which makes infection easier.
Deaton further argues that countries with higher death rates, as shown in the figure above, have also tended to have worse economic outcomes. 

All analysis of the pandemic is, as yet, incomplete. Deaton's data goes through the end of 2020. Just as India has recently been clobbered by the pandemic, something similar could happen in other countries. Furthermore, in the poorest countries of the world, even a smaller loss of income may cause extreme human suffering. 

But other possible lessons here are that, just perhaps, the pandemic did not make the world a more economically unequal place. Moreover, having lower deaths from the pandemic appears to be a good way of bolstering a country's economy. asdj

Tuesday, May 11, 2021

The Slow Magic from Agricultural R&D

For much of human history, a majority of people have worked in agriculture. In the countries of sub-Saharan Africa, about half of all workers are currently in agriculture--more in lower-income countries. The process of raising the overall standard of living requires a rise in agricultural productivity, so that a substantial share of workers can shift away from agriculture, and thus be able to work in other sectors of the economy. In turn, rises in agricultural productivity are typically driven by research and development, which has been lagging. Julian M. Alston, Philip G. Pardey, and Xudong Rao make the case in "Rekindling the Slow Magic of Agricultural R&D" (Issues in Science and Technology, May 3, 2021).

The authors discuss CGIAR, which stands for Consultative Group on International Agricultural Research. This system was started in 1971. The authors note: "The CGIAR was conceived to play a critical role, working in concert with national agricultural research systems of low- and middle-income countries, to develop and distribute farm technologies to help stave off a global food crisis. The resulting Green Revolution technologies were adapted and adopted throughout the world, first and foremost in South Asia and parts of sub-Saharan Africa and Latin America where the early centers of the CGIAR were located. In 2019 the CGIAR spent $805 million on agricultural R&D to serve the world’s poor, down by 30% (adjusted for inflation) from its peak of over $1 billion in 2014 ... "

For perspective, total public and private spending by low-income countries on agricultural R&D is roughly equal to what is spent through CGIAR. The payoff from this spending has been on the order of 10:1. 
The CGIAR research record has been much studied, but questions remain about the past and prospective payoffs to the investment. Similar questions have been raised about public investments in the agricultural research systems of various nations—particularly those of poor countries that receive substantial development aid from richer countries. To address those questions, we conducted a comprehensive meta-analysis of more than 400 studies published since 1978 that looked at rates of return on agricultural research conducted by public agencies in low- and middle-income countries. Of that total, 78 studies reported rates of return for CGIAR-related research and 341 studies reported rates of return for non-CGIAR agricultural research. (Full details of the meta-analysis are online at supportagresearch.org.) ...

 Across 722 estimates, the median ratio of the estimated research benefits to the corresponding costs was approximately 10:1 for both the CGIAR (170 estimates) and national agricultural research systems of developing countries (522 estimates). In other words, $1 invested today yields, on average, a stream of benefits over future decades equivalent to $10 (in present value terms). ...  Notably, all these estimated benefits accrued in developing countries, home to the preponderance of the world’s food poor. And yet, rich donor countries also reap benefits by adopting technologies developed by CGIAR research—“doing well by doing good.” For example, the yield- and quality-enhancing traits bred into new wheat and rice varieties destined for developing countries are also incorporated into most varieties used by rich-country farmers.
But as noted earlier, CGIAR funding is down 30% in the last few years. Also, I was surprised to notice that the Gates Foundation alone is more than one-eighth of then entire CGIAR budget. 

We are talking here about quantities measured in hundreds of millions of dollars--not even a single billion, much less the trillions that are being discussed in various pandemic-relief programs. The benefits of agricultural R&D seem enormous, but the world is not stepping up to the opportunity. 

Thursday, May 6, 2021

Interview with Matthew Jackson: Human Networks

David A. Price does an "Interview" with Matthew Jackson, with the subheading "On human networks, the friendship paradox, and the information economics of protest movements" (Econ Focus: Federal Reserve Bank of Richmond, 2021, Q1, pp. 16-20). Here are a few snippets of the conversation, suggestive of the bigger themes.

Homophily
[O]ne key network phenomenon is known among sociologists and economists as homophily. It's the fact that friendships are overwhelmingly composed of people who are similar to each other. This is a natural phenomenon, but it's one that tends to fragment our society. When you put this together with other facts about social networks — for instance, their importance in finding jobs — it means many people end up in the same professions as their friends and most people end up in the communities they grew up in.

From an economic perspective, this is very important, because it not only leads to inequality, where getting into certain professions means you almost have to be born into that part of society, it also means that then there's immobility, because this transfers from one generation to another. It also leads to missed opportunities, so people's talents aren't best matched to jobs.
The Friendship Paradox
This concerns another network phenomenon, which is known as the friendship paradox. It refers to the fact that a person's friends are more popular, on average, than that person. That's because the people in a network who have the most friends are seen by more people than the people with the fewest friends.

On one level, this is obvious, but it's something that people tend to overlook. We often think of our friends as sort of a representative sample from the population, but we're oversampling the people who are really well connected and undersampling the people who are poorly connected. And the more popular people are not necessarily representative of the rest of the population.

So in middle school, for example, people who have more friends tend to have tried alcohol and drugs at higher rates and at earlier ages. And this distorted image is amplified by social media, because students don't see pictures of other students in the library but do tend to see pictures of friends partying. This distorts their assessment of normal behavior.

There have been instances where universities have been more successful in combating alcohol abuse by simply educating the students on what the actual consumption rates are at the university rather than trying to get them to realize the dangers of alcohol abuse. It's powerful to tell them, "Look, this is what normal behavior is, and your perceptions are actually distorted. You perceive more of a behavior than is actually going on."
Causality in Networks
Establishing causality is extremely hard in a lot of the social sciences when you're dealing with people who have discretion over with whom they interact. If we're trying to understand your friend's influence on you, we have to know whether you chose your friend because they behave like you or whether you're behaving like them because they influenced you. So to study causation, we often rely on chance things like who's assigned to be a roommate with whom in college, or to which Army company a new soldier is assigned, or where people are moved under a government program that's randomly assigning them to cities. When we have these natural experiments that we can take advantage of, we can then begin to understand some of the causal mechanisms inside the network.
Live Protests vs. Social Media
[I]t's cheap to post something; it's another thing to actually show up and take action. Getting millions of people to show up at a march is a lot harder than getting them to sign an online petition. That means having large marches and protests can be much more informative about the depth of people's convictions and how many people feel deeply about a cause.

And it's informative not only to governments and businesses, but also to the rest of the population who might then be more likely to join along. There are reasons we remember Gandhi's Salt March against British rule in 1930 or the March on Washington for Jobs and Freedom in 1963. This is not to discount the effects that social media postings and petitions can have, but large human gatherings are incredible signals and can be transformative in unique ways because everybody sees them at the same time together with this strong message that they convey.
If you would like more Jackson, one starting point is his essay in the Fall 2014 issue of the Journal of Economic Perspectives, "Networks in the Understanding of Economic Behaviors." The abstract reads:
As economists endeavor to build better models of human behavior, they cannot ignore that humans are fundamentally a social species with interaction patterns that shape their behaviors. People's opinions, which products they buy, whether they invest in education, become criminals, and so forth, are all influenced by friends and acquaintances. Ultimately, the full network of relationships—how dense it is, whether some groups are segregated, who sits in central positions—affects how information spreads and how people behave. Increased availability of data coupled with increased computing power allows us to analyze networks in economic settings in ways not previously possible. In this paper, I describe some of the ways in which networks are helping economists to model and understand behavior. I begin with an example that demonstrates the sorts of things that researchers can miss if they do not account for network patterns of interaction. Next I discuss a taxonomy of network properties and how they impact behaviors. Finally, I discuss the problem of developing tractable models of network formation.

Wednesday, May 5, 2021

The Great Texas Power Failure of February 2021

In the aftermath of the Texas power failures in February, a number commenters found confirmation that, amazingly, they had been right about everything all along. Thus, those who were against wind power and renewable energy mandated in general blamed the wind farms. Those who are suspicious of competition in markets for generating electrical power blamed deregulation, although blaming "the market" for what happens in a heavily regulated industry seems peculiar to me. Some critics even blamed Enron, a company that has not existed for years. What actually happened seems simpler, if less reinforcing for various preconceptions: It got really cold. 

Michael Giberson provides an overview in "Texas Power Failures: What happened in February 2021 and What Can be Done" (Reason Foundation, April 2021). He describes the weather: 
The temperature in Dallas dipped to -2° F, the coldest it had been in Dallas for 70 years. Snow fell on the beaches on the Gulf Coast at Galveston, south of Houston. Temperatures in Austin remained below freezing for six days at a time of when temperatures usually average in the mid-50s. At Brownsville, near the most southern tip of Texas, February weather typically averages 65° F. High temperatures in Brownsville were in the mid-80s just days before the cold. The temperature in the city did not rise above freezing for nearly 48 hours once the cold settled in. For the first time in history all 254 counties in Texas were under a winter storm warning at the same time. The cold was not unprecedented at any particular location, but it was extreme, widespread, and long lasting in February 2021. ...
The cold affected more than the ERCOT power system. Some power systems in Texas not within the ERCOT system also resorted to rolling outages. Natural gas production and distribution froze up. Municipal water mains froze in cities across the South. Ranchers in the Panhandle lost cattle to the cold. Citrus growers in South Texas saw damage to trees that may last for years. Roads were closed due to ice and storms. Failures were not solely an electric power industry concern or a natural gas failure. The cold was simply worse than almost anyone in Texas was prepared for. ... Clearly, it was not negligent on ERCOT’s part—and maybe anyone’s part—to fail to anticipate such anomalous temperatures.

The  Texas power emergency lasted about four days: at its worst, about 4.5 million people were without power.

Of course, the obvious question is why ERCOT--the ironically named Electric Reliability Council of Texas which has responsibility for regulating Texas electricity---had not already required larger investments against cold weather. After all, there had been a cold snap back in 2011 that also caused power outages, although it was not as extreme as the February 2011 version. The short answer is that the weather turned out to be colder than ERCOT's worst-case scenario. Here's a figure that takes a bit of explaining, but tells much of the story: 
The black line shows the actual electricity load. The thin gray line shows the forecasted demand, if ERCOT had been able to deliver it. In general, electricity demand is typically  higher in Texas in the summer (air-conditioning) rather than the winter. But electricity demand during the cold snap would have broken the all-time summer records, as well as the winter ones. 

But the real problem was on the supply side. The ERCOT "extreme" scenario was that 14 GW of electricity would go off-line; actually, 30 GW went off-line.  The blue dashed horizontal line bottom line shows the 2 GW that ERCOT projected for wind and solar power in its "extreme" scenario. There were a couple of small dips below this level, but a drop in wind power was not the main culprit here. 

In retrospect, some of the problem was poor coordination across the energy system. For example, some natural gas pipeline operators failed to submit the information to their electricity providers so that they could be treated as "critical load" functions, which meant that they couldn't deliver natural gas to generate electricity:" "At its worst, as much as 9,000 MW of generation was sidelined by the lack of gas supplies, in part due to power cut offs at gas pipelines." The drop in electricity produced from natural gas was by far the biggest source of the overall drop in supply. 

Other than better coordination of electricity supplies, what else might be done to avoid similar power failures in the future. The key point here to remember is that we are talking about preparation for a very rare event. 

One option is to invest more in weatherization. Another is to pay some firms for being ready to provide a certain amount of electricity in an emergency, even if most of the time they are not actually doing so--that is, pay for some extra unused capacity. Another option is to build more connections from ERCOT to electricity grids outside of Texas, which could be very valuable in emergencies even if they aren't used much of the time. Yet another option would be to encourage Texas electricity users to maintain some of their own battery storage or generating capacity for emergencies. 

Hindsight is 20:20, but now that Texas has been warned by experience, the case for some mixture of these actions is strong. Garrett Golding, Anil Kumar and Karel Mertens at the Dallas Federal Reserve offer some estimates in "Cost of Texas’ 2021 Deep Freeze Justifies Weatherization" (Dallas Fed Economics blog, April 15, 2021). In measuring economic losses from the power outage, they write: 
The power outages led to widespread damage to homes and businesses, foregone economic activity, contaminated water supplies and the loss of at least 111 lives. Early estimates indicate that the freeze and outage may cost the Texas economy $80 billion–$130 billion in direct and indirect economic loss. These initial calculations come with significant uncertainty. Estimates of insured losses, which are easier to quantify, range from $10 billion to $20 billion.
In terms of what steps might be taken, they note: 
Winterizing standards on new oil and gas wells may offer a targeted and effective approach in the long run. Due to the high initial productivity of shale wells, new wells will eventually make up a large share of overall production. Many companies already implement winterizing measures. With winterizing equipment costing between $20,000 and $50,000 per well, we estimate these measures statewide would total $85 million–$200 million annually. A large and perhaps inexpensive fix would be prioritizing electricity delivery to gas infrastructure. If power plant and pipeline operators improve coordination to identify and constantly monitor the gas infrastructure requiring such prioritization, some of the problems experienced during the freeze could be prevented.

It's also possible to winterize the wind farms that received so much attention. They write about the possibilities of "upgraded blade coatings, cold-weather lubricants and de-icing drones." 

Overall, the lesson here seems to be to move past the blame game, to accept that the cold weather snap was unprecedented, and now to have Texans pay a little more for electricity to fund these kinds of steps.