Category Archive for: Economics [Return to Main]

Tuesday, January 29, 2019

Links (1/29/19)

Thursday, January 24, 2019

Links (1/24/19)

Tuesday, January 22, 2019

Links (1/22/19)

Tuesday, January 15, 2019

Links (1/15/19)

Friday, January 11, 2019

Links (1/11/19)

Monday, January 07, 2019

Links (1/7/19)

Monday, December 31, 2018

Links (12/31/18)

Monday, December 17, 2018

Links (12/17/18)

Couple of links:

Friday, December 07, 2018

Links (12/7/18)

Tuesday, December 04, 2018

Links (12/4/18)

I am in Kenya (was teaching time-series econometrics in Nairobi) and got a bit behind. For pictures of Maasai Mara, Nairobi, and other stuff, see here:

Wednesday, November 28, 2018

Links (11/28/18)

Monday, November 26, 2018

Links (11/25/18)

Wednesday, November 21, 2018

Links (11/21/18)

Monday, November 19, 2018

Links (11/19/18)


Thursday, November 15, 2018

Links (11/15/18)

Tuesday, November 13, 2018

Links (11/13/18)

Friday, November 09, 2018

Links (11/9/18)

Wednesday, November 07, 2018

Links (11/7/18)

Monday, November 05, 2018

Links (11/5/18)

Wednesday, October 31, 2018

Links (10/31/18)

Wednesday, October 24, 2018

Links (10/24/18)

Monday, October 22, 2018

Links (10/22/18)

Tuesday, October 16, 2018

Links (10/16/18)

Thursday, October 11, 2018

Links (10/11/18)

Monday, October 08, 2018

Links (10/8/18)

Friday, October 05, 2018

Links (10/5/18)



Wednesday, October 03, 2018

links (10/3/18)


Tuesday, October 02, 2018

links (10/2/18)



Sunday, September 30, 2018

Links (9/30/18)

My computer died. I may have lost a few links (and doing this on an iPad is less than convenient)

Wednesday, September 26, 2018

Links (9/26/18)

Thursday, September 20, 2018

Links (9/20/18)

Monday, September 17, 2018

Links (9/17/18)

Thursday, September 13, 2018

Links (9/13/18)

Tuesday, September 11, 2018

Links (9/11/18)

Friday, September 07, 2018

Links (9/7/18)

Wednesday, September 05, 2018

Links (9/5/18)

Tuesday, September 04, 2018

  New Evidence that Unions Raise Wages for Less-Skilled Workers

From Steve Maas at the NBER:

New Evidence that Unions Raise Wages for Less-Skilled Workers: Tapping into eight decades of private and public surveys, a new study finds evidence that unions have historically reduced income inequality.

For Unions and Inequality over the Twentieth Century: New Evidence from Survey Data (NBER Working Paper No. 24587), Henry S. Farber, Daniel Herbst, Ilyana Kuziemko, and Suresh Naidu assembled a household-level database on union membership dating back to 1936.


The U.S. Bureau of the Census has tracked wages and education consistently since 1940. Aggregate data on union membership goes back to the early 20th century, but data on individual workers were not readily available until the Census Bureau started asking about union affiliation in 1973. By that time, unions were already in decline, and higher-skilled workers accounted for an increasing share of their membership.

The researchers draw on more than 500 surveys conducted by Gallup and other pollsters from 1936 through 1986, extending their dataset into the present day with information from government surveys and other sources.

Their study finds that the salary premium for union members compared to workers with comparable skills and demographic characteristics has remained relatively steady over the last 80 years despite large swings both in the overall number of union members and in their education levels. The less skilled the workers were, the greater the wage premium associated with their union membership. The researchers find a negative correlation between unionization rates and measures of inequality such as the Gini coefficient.

Between 1940 to 1970, when unionization peaked and income inequality narrowed, unions were drawing in the least-skilled workers. Before and after that period, unions were smaller and a higher fraction of their members were drawn from the ranks of high-skill workers. The 1940 - 1970 period also coincided with the highest share of union members drawn from minority groups.

The clear implication of the researchers' analysis is that, because unions offer a larger wage premium to less-skilled workers, unions have an important equalizing effect on the income distribution to the extent that they are successful in organizing the less-skilled. Recent decades have seen growth in educational attainment in the workforce, and, importantly, not only has the overall share of workers who are unionized declined, but unions have also become relatively less successful in organizing less-skilled workers. The remaining unionized workforce is more highly educated than it was earlier. The combination of the declining presence of unions in the labor market and the increased skill level of the remaining union workers means that the important equalizing effect of unions on the income distribution that was seen in the middle of the 20th century has diminished substantially.

Monday, September 03, 2018

Links (9/3/18)

Saturday, September 01, 2018

Links (9/1/18)

I have been in Hawaii since before the hurricane and kind of fell behind. Oops:

Thursday, August 23, 2018

Links (8/23/18)

Tuesday, August 21, 2018

Links (8/21/18)

Sunday, August 19, 2018

Credit Supply and Housing Speculation

Atif Mian and Amir Sufi at VoxEU:

Credit supply and housing speculation, by Atif Mian and Amir Sufi, VoxEU: Charles P. Kindleberger, who was the world’s leading expert on financial crises, wrote that “asset price bubbles depend on the growth in credit” (Kindleberger and Aliber 2005). Nobel prize winner Vernon Smith described evidence from experimental settings showing that that the size of a bubble increased when individuals were allowed to borrow (Porter and Smith 1994). Economic theorists have taken this lesson to heart, writing down models in which easier credit helps fuel asset prices through an increase in speculative buying (Allen and Gorton 1993, Allen and Gale 2000).
A core idea in the theory of credit and bubbles is that easier credit allows optimists with high asset valuations to aggressively buy assets, and therefore boost the price (Geanakoplos 2010, Simsek 2013). Even if optimists form a small part of the overall population, easier credit can allow this small group to have a large effect on the market. Further, if the optimists suddenly lose access to credit, the price of the asset will collapse before more pessimistic individuals can be induced to buy the asset. As a result, fluctuations in credit availability increase the amplitude of fluctuations in asset prices.
Our recent study tests this idea, focusing on the boom and bust in house prices from 2000 to 2010 in the US (Mian and Sufi 2018). The study focuses on a natural experiment: the sudden acceleration of the private label mortgage securitisation (PLS) market in the late summer of 2003. The sudden rise in the PLS market, which was part of the broader global rise in shadow banking during this period, disproportionately reduced the cost of financing by lenders that did not traditionally rely on deposit financing for mortgage lending. The study shows that lenders who traditionally relied on non-deposit financing, such as CountryWide and Ameriquest Mortgage Company, suddenly boosted mortgage lending in the late summer of 2003, just as the PLS market accelerated.
To test the effect of this sudden increase in credit availability on the housing market, we exploit variation across geographic areas in the US in the location of these lenders as of 2002. Zip codes where lenders traditionally relied on non-deposit financing witnessed a sudden and large relative increase in mortgage lending just as the PLS market accelerated in 2003. Our study shows several results that suggest this is a clean experiment – the sudden and large expansion of mortgage lending in these zip codes was due to the acceleration of the PLS market, as opposed to some other factor such as a change in income prospects or beliefs about house prices among those living in these zip codes.
Consistent with models in which credit availability affects asset prices, the sharp rise in mortgage lending in these zip codes generated a boom and bust in house prices. In fact, exposure of a zip code to non-traditional lenders in 2002 predicted the severity of the collapse in house prices from 2006 to 2010.
Furthermore, US cities that had greater exposure to these lenders were more likely to experience a simultaneous increase in both house prices and construction activity during the boom. The presence of such bubble cities, such as Las Vegas and Phoenix, has puzzled economists because in most standard models the ability to easily construct more housing units should put a lid on house price growth. The results of our study suggest that easier credit was a crucial ingredient in explaining bubble cities that had both house price and construction booms. We further show that these cities witnessed a particularly painful bust from 2006 to 2010.
A unique advantage of our study is the ability to track the marginal buyers of homes that were brought into the market by easier credit. Zip codes more exposed to the acceleration of the PLS market witnessed a substantial increase in transaction volume from 2003 to 2006, and this increase in volume was almost completely driven by flippers (i.e. individuals that buy and sell multiple homes in a short period of time). Such flippers were a small fraction of the overall population – by our estimate, flippers made up less than 1% of the overall adult population in 2005 and 2006. Despite being a small part of the overall population, flippers had a disproportionate effect on the housing market because they were able to easily obtain credit.
The results support models in which easier credit can boost asset prices by giving a small group of aggressive buyers the ability to affect the overall market. In the presence of easy credit, it is not necessary for there to be a widespread increase in optimism about the housing market to generate a large increase in house prices.
Evidence from the Michigan Survey of Consumers supports this conclusion. As has been shown in previous research (Piazzesi and Schneider 2009), the fraction of the overall population who said “it is a good time to buy a home” actually declined substantially from 2003 to 2006 during the heart of the housing boom. We add to this evidence by showing that the share of individuals saying “now is a good time to buy a home” declined most in cities that experienced a large rise in house prices fuelled by the PLS market. On average, individuals became increasingly pessimistic about the housing market in cities where the PLS market fuelled a trading frenzy by flippers. Easy credit allowed a small group of individuals to boost house prices in some cities even though the average individual in these cities soured on the housing market.
Flipping fuelled by the PLS market was a crucial factor that instigated the mortgage default crisis. As early as 2007, flippers in zip codes most exposed to the PLS market had default rates above 20%. The share of all mortgage defaults from zip codes most exposed to the PLS market increased in 2007. By 2008 and 2009, defaults were rising throughout the country, but the evidence suggests that the mortgage default crisis was triggered by defaults emanating from the PLS market.
The bust also provides important lessons for the interaction of credit and asset prices. While almost all buyers in zip codes most exposed to the PLS market used a mortgage to buy a home from 2003 to 2006, the share of cash-buyers increased sharply in 2007 and afterward. This pattern is consistent with the idea that prices collapsed in part because tighter credit prevented optimists from buying homes during the sell-off, which meant more pessimistic cash-buyers became the marginal price setters. Loose credit boosted prices during the boom, and tight credit exacerbated the bust. Credit fluctuations and asset price fluctuations are closely connected.
Allen, F and D Gale (2000), “Bubbles and crises," The Economic Journal 110(460): 236-255.
Allen, F and G Gorton (1993), “Churning bubbles," The Review of Economic Studies 60(4): 813-836.
Geanakoplos, J (2010), “The leverage cycle," NBER Macroeconomics Annual 24(1): 1-66.
Kindleberger, C P and R Z Aliber (2005), Manias, panics and crashes: A history of financial crises, Palgrave Macmillan.
Mian, A and A Sufi (2018), “Credit Supply and Housing Speculation,” NBER Working Paper 24823.
Piazzesi, M and M Schneider (2009), “Momentum traders in the housing market: survey evidence and a search model," American Economic Review 99(2): 406-11.
Simsek, A(2013), “Belief disagreements and collateral constraints," Econometrica 81(1): 1-53.

Saturday, August 18, 2018

Links (8/18/18)

Wednesday, August 15, 2018

Links (8/15/18)

Monday, August 13, 2018

FRBSF: The Current Economy and the Outlook

From the FRBSF:

FedViews: Kevin J. Lansing, research advisor at the Federal Reserve Bank of San Francisco, stated his views on the current economy and the outlook as of August 9, 2018.

Economy running above sustainable rate

The initial estimate of real GDP growth in the second quarter of 2018 came in at 4.1% at an annual rate, resulting in a growth rate of 2.8% over the past four quarters. For 2018 as a whole, we expect growth to come in just under 3%, well above our estimate of the economy’s long-run sustainable growth rate. Given the diminishing effects of federal fiscal stimulus over the next few years and the expected tightening of financial conditions, we project that growth will slow to just under 2% by 2020.

Job growth remains strong

The Bureau of Labor Statistics reported that payroll employment increased by 157,000 jobs in July. Data for the previous two months were revised upward, resulting in an average job gain over the past three months of 224,000. Over the past year, the average monthly gain was 203,000 jobs. The unemployment rate edged down to 3.9% from 4% in June. We expect monthly job gains to remain above the breakeven level needed to keep pace with the growth rate of the labor force. Consequently, we expect the unemployment rate to decline further below our 4.6% estimate of the natural rate of unemployment.

Unemployment below sustainable levels

Inflation expected to modestly exceed target

Inflation over the past year is close to the Federal Open Market Committee’s (FOMC’s) 2% target. With unemployment below the natural rate and real GDP growth above its long-run sustainable pace, we expect some upward pressure on inflation over the medium term, causing the four-quarter inflation rate to slightly overshoot the 2% target in 2020.

Interest rates up but remain accommodative

Following the conclusion of its latest meeting on August 1, the FOMC announced its decision to maintain the target range for the federal funds rate at 1¾ to 2%. The Committee noted that recent economic activity has been strong and that risks to the economic outlook appear roughly balanced. The Committee expects further gradual increases in the target range for the federal funds rate.
Interest rates have continued to increase with the ongoing monetary policy normalization. Nevertheless, the current level of the federal funds rate remains accommodative as it stands about 50 basis points below our estimate of the “neutral” federal funds rate.
Few, if any, past recessions have been successfully predicted either by the Federal Reserve or professional forecasters. Forecasting recessions is difficult because each one tends to differ in important ways from previous episodes. Past recessions have been triggered by upward spiking oil prices, increases in policy interest rates designed to bring down high inflation, and bursting asset price bubbles.
Despite the varied triggers, recessions are typically preceded by some characteristic interest rate configurations. These include an inverted Treasury yield curve, an elevated real short-term interest rate, and a compressed credit spread (as measured by the yield difference between Baa corporate bonds and 10-year Treasury bonds).
An inverted yield curve is often observed after a sustained series of monetary policy tightening actions that serve to raise real short-term Treasury yields. Long-term Treasury yields, which reflect expectations of future economic conditions, tend to move up with short-term yields during the early phases of an economic expansion, but may stop doing so (resulting in a flat or inverted yield curve) if investors’ economic outlook becomes more pessimistic.
Corporate bond yields are typically higher than Treasury bond yields because corporate yields must compensate investors for the risk of default. During an economic expansion, default risk declines which causes the credit spread to compress. But a sustained expansion may cause investors to underestimate the risk of default, contributing to weak lending standards, excessive borrowing, and a credit spread that is too low. The onset of an economic slowdown or a recession would trigger the unwinding of such conditions. Research shows that optimistic credit market sentiment, as measured in part by a compressed credit spread, tends to predict slower economic activity at a two-year horizon.

Yield curve is flat but not inverted

Real short−term interest rate remains low

Credit spread is compressed

The current interest rate configuration can be described as follows: (1) The Treasury yield curve is relatively flat but not inverted, (2) the real short-term interest rate has been increasing but is still low by historical standards, and (3) the credit spread is compressed. This configuration can be described as providing mixed signals about the future. While the first two observations suggest that the risk of a recession remains relatively low, factoring in the third observation would suggest a somewhat higher risk of a recession than otherwise.
The views expressed are those of the author, with input from the forecasting staff of the Federal Reserve Bank of San Francisco. They are not intended to represent the views of others within the Bank or within the Federal Reserve System.

Links (8/13/18)

Friday, August 10, 2018

Links (8/10/18)

Wednesday, August 08, 2018

Links (8/8/18)

The Impact of Higher Temperatures on Economic Growth

This is from Riccardo Colacito, Bridget Hoffmann, Toan Phan and Tim Sablik:

The Impact of Higher Temperatures on Economic Growth, by Riccardo Colacito, Bridget Hoffmann, Toan Phan and Tim Sablik, Economic Brief, Richmond Fed: June 2018 was the third-warmest on average across the contiguous forty-eight states since record keeping began in 1895, according to the National Oceanic and Atmospheric Administration (NOAA). Only 1933 and 2016 saw hotter starts to the summer.
Climate scientists project that average global temperatures will rise over the coming decades, which could have a variety of environmental impacts. But what impact would higher temperatures have on the economy? To date, studies of this question have largely focused on developing countries, under the assumption that those countries are more exposed to the effects of higher temperatures. The economy in developing countries is often more reliant on agriculture or other outdoor activities, and those countries have fewer resources to devote to mitigating the effects of heat through technologies such as air conditioning. Indeed, researchers have found that higher temperatures have significant negative effects on the economic growth of developing nations.1
In the case of developed countries, such as the United States, researchers have focused largely on measuring the impact of warming on outdoor economic activities, such as agriculture.2 Since these sectors make up a relatively small share of the U.S. economy, it has generally been assumed that the economic effects of global warming for the United States would be relatively small. As Nobel prize winning economist Thomas Schelling observed in a 1992 article, "Today very little of our gross domestic product is produced outdoors, susceptible to climate."3
However, research by three authors of this Economic Brief  (Colacito, Hoffmann, and Phan) finds that the consequences of higher temperatures on the U.S. economy may be more widespread than previously thought. By examining changes in temperature by season and across states, they find evidence that rising temperatures could reduce overall growth of U.S. economic output by as much as one-third by 2100.4
Warming across Seasons and across States
Attempting to measure the relationship between temperature and growth by looking at the whole United States can hide important variations. Some parts of the country have higher average temperatures. Further increasing temperatures in those areas may be more harmful than rising temperatures in parts of the country that are generally cooler. In fact, higher temperatures in colder regions or during colder seasons actually may have positive effects on economic activity because extreme cold can be as much an impediment to certain activities as extreme heat.
Highlighting the importance of these seasonal and regional variations, Colacito, Hoffmann, and Phan find no statistically significant relationship between temperature and economic growth when looking across the whole United States. But measuring the impact of temperature in different seasons and across individual states yields different results. The authors take the average of daily weather observations from NOAA for each season for 1957–2012. They define each season as a quarter of the calendar year: January through March is winter, April through June is spring, July through September is summer, and October through December is fall. This definition aligns the temperature data with the quarterly periods used for economic data.
Colacito, Hoffmann, and Phan find that temperature increases in the summer are associated with a decline in gross state product (GSP), which is the value added in production by the labor and capital of all industries in a given state. On average, each 1˚F increase in the mean summer temperature reduces the annual GSP growth rate by 0.154 percentage points. A reduction in the growth rate, as opposed to the level of economic output, has important implications for the impact of temperature changes in the long run. Changes to the growth rate compound over time and, as a result, are more lasting.
As theory would suggest, Colacito, Hoffmann, and Phan also find that higher temperatures during the colder fall months have a positive effect on growth. On average, each 1˚F increase in the mean fall temperature increases the annual GSP growth rate by 0.102 percentage points. This finding is smaller and less statistically robust than their finding for the summer effect, but it may help explain why temperature changes do not appear to have a significant effect on growth when averaged across the whole year and across the whole country: the effects in the summer and fall partly offset. The authors do not find any significant effects for temperature increases in the spring or winter.
Measuring the impact of temperature changes on states as opposed to the country as a whole also reveals significant variations. Colacito, Hoffmann, and Phan divide the country into four regions — North, South, Midwest, and West — using classifications from the U.S. Census Bureau. Average temperatures are highest in the South, and the authors find that the economies of southern states are the most sensitive to changes in summer and fall temperatures. Further investigation shows that this effect is not driven by a larger role of agriculture in southern states. In fact, the authors find that the economic effects of temperature are widespread across a variety of industries.
Rising Temperatures Hurt Many Industries
One might easily presume that higher temperatures would only affect agriculture. But in fact, studies have documented the effects of extreme temperatures on other industries. For example, temperatures above 90˚F have been found to reduce production at automobile manufacturing plants in the United States.5 Another study published by the Chicago Fed found that severe winter weather has a significant, albeit short-lived and generally small, negative effect on a variety of industries.6 In line with these findings, Colacito, Hoffmann, and Phan find that higher temperatures in the summer have a negative effect on labor productivity generally, while higher fall temperatures have a positive impact.
Losses in labor productivity have the potential to impact a wide range of industries, which is exactly what Colacito, Hoffmann, and Phan find. (Figure 1 shows results for 1998–2012.) The two largest sectors of the U.S. economy — services and FIRE (finance, insurance, and real estate) — make up half of national GDP and are both hurt by higher summer temperatures. More housing transactions take place in the spring and summer, perhaps because house shopping involves travel and outdoor activity. As temperatures rise, potential homebuyers may tend to stay inside, which could help explain the finding that higher summer temperatures negatively impact the real estate sector.7


Studies also have documented that high temperatures negatively affect health, resulting in increased hospitalizations.8 Colacito, Hoffmann, and Phan hypothesize that this connection may explain the finding that higher summer temperatures have a substantial impact on the insurance sector. As health outcomes worsen, insurers would face increased claims. Overall, the authors find that a 1˚F increase in temperature is associated with a 1.30 percentage point decline in output growth for the insurance sector.
As expected, the authors also find that higher summer temperatures have a large negative effect on agriculture, forestry, and fishing. Although this sector accounts for only about 1 percent of national GDP, losses in this area may spill over to other sectors of the economy, such as retail food services. Higher summer temperatures do have a positive effect on some industries, including utilities and mining, benefits that may stem from increased energy consumption during hotter days.
Looking Ahead
Although the effects estimated by Colacito, Hoffmann, and Phan are robust, they are also small in the short term. Over a longer horizon, however, the impact on GDP growth rates may be substantial. The authors study the effects of rising temperatures in the future using projections for average temperatures in the United States over the years 2070–99.9 These estimates use three different scenarios of future greenhouse gas emissions (high, medium, and low) by the Intergovernmental Panel on Climate Change. The authors apply these estimates to their analysis, assuming that states do not make any changes to adapt to or mitigate the effects of higher temperatures and that the effects of temperature on economic growth that they found in their state-by-state analysis do not change.
Under the low-emissions scenario, the authors estimate that rising temperatures would reduce the growth rate of GDP by 0.2 to 0.4 percentage points from 2070 through 2099, or as much as 10 percent of the historical average annual growth rate of 4 percent. Under the high-emissions scenario, rising temperatures could reduce the growth rate by up to 1.2 percentage point, or roughly one-third of the historical average annual GDP growth rate. (See Figure 2.) The authors note that these estimates should be "interpreted with caution," since future adaptations to changing temperatures may mute the long-run effects they calculate.


While the impact of future climate adaptations is unknown, Colacito, Hoffmann, and Phan do examine whether more widespread climate adaptation within their sample period may have reduced the impact of temperature on growth. In fact, they find that the negative impact of higher summer temperatures is larger and still statistically significant after 1990, while the positive fall effect becomes smaller and statistically indistinguishable from zero. Thus, if anything, they find that the negative impact of temperature increases on GDP growth has become more pronounced in recent decades despite advances in adaptive measures.
Overall, these findings suggest that rising temperatures in the future could hamper economic growth in a variety of industries even in developed nations such as the United States.

Riccardo Colacito is an associate professor of finance and economics at the University of North Carolina, Chapel Hill, and Bridget Hoffmann is an economist in the Research Department at the Inter-American Development Bank. Toan Phan is an economist and Tim Sablik is an economics writer in the Research Department at the Federal Reserve Bank of Richmond.
1John Luke Gallup, Jeffrey D. Sachs, and Andrew D. Mellinger, "Geography and Economic Development," International Regional Science Review, August 1999, vol. 22, no. 2, pp. 179–232; William D. Nordhaus, "Geography and Macroeconomics: New Data and New Findings," Proceedings of the National Academy of Sciences of the United States of America, March 2006, vol. 103, no. 10, pp. 3510–3517; Melissa Dell, Benjamin F. Jones, and Benjamin A. Olken, "Temperature Shocks and Economic Growth: Evidence from the Last Half Century," American Economic Journal: Macroeconomics, July 2012, vol. 4, no. 3, pp. 66–95.
2See, for example, Marshall Burke and Kyle Emerick, "Adaptation to Climate Change: Evidence from U.S. Agriculture," American Economic Journal: Economic Policy, August 2016, vol. 8, no. 3, pp. 106–140.
3Thomas C. Schelling, "Some Economics of Global Warming," American Economic Review, March 1992, vol. 82, no. 1, pp. 1–14.
4Riccardo Colacito, Bridget Hoffmann, and Toan Phan, "Temperature and Growth: A Panel Analysis of the United States," Federal Reserve Bank of Richmond Working Paper No. 18-09, March 2018.
5Gerard P. Cachon, Santiago Gallino, and Marcelo Olivares, "Severe Weather and Automobile Assembly Productivity," Columbia Business School Research Paper No. 12/37, December 2012.
6Justin Bloesch and François Gourio, "The Effect of Winter Weather on U.S. Economic Activity," Federal Reserve Bank of Chicago Economic Perspectives, First Quarter 2015, vol. 39, no. 1, pp. 1–20.
7 L. Rachel Ngai and Silvana Tenreyro, "Hot and Cold Seasons in the Housing Market," American Economic Review, December 2014, vol. 104, no. 12, pp. 3991–4026.
8See, for example, Ekta Choudhary and Ambarish Vaidyanathan, "Heat Stress Illness Hospitalizations — Environmental Public Health Tracking Program, 20 States, 2001–2010," Morbidity and Mortality Weekly Report, Surveillance Summaries, December 12, 2014, vol. 63, no. 13.
9Temperature estimates come from Evan H. Girvetz, Chris Zganjar, George T. Raber, Edwin P. Maurer, Peter Kareiva, and Joshua J. Lawler, "Applied Climate-Change Analysis: The Climate Wizard Tool," PLoS One, December 2009, vol. 4, no. 12, e8320.

This article may be photocopied or reprinted in its entirety. Please credit the authors, source, and the Federal Reserve Bank of Richmond and include the following statement. Views expressed in this article are those of the authors and not necessarily those of the Federal Reserve Bank of Richmond or the Federal Reserve System.

Economic Policy for Artificial Intelligence

Ajay Agrawal, Joshua Gans, and Avi Goldfarb at VoxEU:

Economic policy for artificial intelligence: Artificial intelligence (AI) technologies advanced rapidly over the past several years. Governments around the world responded by developing AI strategies. France released its national AI strategy in March 2018, emphasising research funds, ethical issues, and inequality. China stated a goal of being the top AI country by 2030. The EUCanada, Japan, the Obama administration, the Trump administration, and many others have put forth their own plans (Sutton 2018).

Pessimistic views of the impact of AI on society are widespread. Elon Musk, Stephen Hawking, Bill Gates, and others warn that rapid advances in AI could transform society for the worse. More optimistically, AI could enhance productivity so dramatically that people have plenty of income and little unpleasant work to do (Stevenson 2018). Regardless of whether one adopts a pessimistic or optimistic view, policy will shape how AI affects society. 

What is AI?

While the Oxford English Dictionary defines artificial intelligence as “the theory and development of computer systems able to perform tasks normally requiring human intelligence”, the recent excitement is driven by advances in machine learning, a field of computer science focused on prediction. As machine learning pioneer Geoffrey Hinton put it: “Take any old problem where you have to predict something and you have a lot of data, and deep learning is probably going to make it work better than existing techniques”.1 Recent advances in AI can therefore be seen as a drop in the cost of prediction. Because prediction is an important input into decision-making, in recent work we discuss how AI is likely to have widespread consequences as a general purpose technology (Agrawal et al. 2018a, 2018b).

There are two aspects of AI policy. 

  • First, regulatory policy has an impact on the speed of diffusion of the technology and the form that the technology takes. 
  • Second, a number of policies focus on mitigating potential negative consequences of AI with respect to labour markets and antitrust concerns. 

Policies that will influence the diffusion of AI

Machine learning uses data to make predictions. The biggest constraint on AI in many settings is the ability to acquire useful data. This creates privacy concerns. Therefore, privacy policy has a direct impact on the ability of organisations to build and implement AI. Too little privacy protection means that consumers may be unwilling to participate in market transactions where their data are vulnerable. Too much privacy regulation means that firms cannot use data to innovate. While the existing empirical work does not focus on AI specifically, the evidence to date suggests that most government-mandated privacy regulation slows technology adoption and innovation, suggesting a trade-off between the right to privacy and the speed of innovation (Goldfarb and Tucker 2012). This means that any government strategy focused on AI – particularly with the aim of fostering a local AI industry – should weigh the potentially conflicting interests of data producers and users, especially with respect to privacy. Perhaps more than any other regulation, rules around privacy are likely to influence the speed and direction of the application of AI in practice. 

Liability rules will also impact the diffusion of AI (Galasso and Luo 2018). Firms will be less likely to invest in the development of AI products in the absence of clear liability rules. Autonomous vehicles provide a useful example. A number of different companies will participate in the development of a self-driving car. If a car gets into an accident, would the sensor manufacturer be liable? The telecommunications provider? The vehicle manufacturer? Or perhaps an AI software firm? Without clear rules on who is liable, all may hesitate to invest. If autonomous vehicles would save lives, should manufacturers of non-autonomous vehicles be held to higher standards than current law requires? This would accelerate diffusion of the safer technology. In contrast, if the increases in liability focus is primarily on newer technology, then diffusion will slow.

In addition, similar to other technologies, advances will be faster with more research support, well-balanced intellectual property law, and the ability to experiment in a safe way.

Policies that address the consequences of AI

A common worry about AI concerns the potential impact on jobs. If machines can do tasks normally requiring human intelligence, will there be jobs left for humans? In our view, this is the wrong question. There are plenty of horrible jobs. Furthermore, more leisure is generally considered to be a positive development, although some have raised concerns about the need to find alternate sources of meaning (Stevenson 2018). The most significant long-run policy issues relate to the potential changes to the distribution of the wealth generated by the widespread use of AI. In other words, AI may increase inequality. 

If AI is like other types of information technology, it is likely to be skill-biased. The people who benefit most from AI will be educated people who already are doing relatively well. These people are also more likely to own the machines. Policies to address the consequences of AI for inequality relate to the social safety net. While some have floated relatively radical ideas to deal with the potential increase in inequality – such as a tax on robots and a universal basic income – the AI context is not unique in weighing the costs and benefits of social programmes from progressive taxation to universal healthcare.

In the shorter run, if AI diffuses widely, the transition could mean temporary displacement for many workers. Acemoglu and Restrepo (2018) emphasise a short- and medium-term mismatch between skills and technology. This means that policy preparation in advance of the diffusion of AI should consider both business cycles and education policy. Technology-driven layoffs concentrated in location and time are not unique to AI. They were a feature of factory automation and the mechanization of farming. For education policy, there are many open questions. Should we emphasise social skills and the humanities if machines increasingly are able to do technology-related prediction tasks? Should the education system evolve to focus more on adults? How do the skills needed as AI diffuses differ from the skills currently provided through the education system?

Another policy question around the diffusion of AI relates to whether it will lead to monopolisation of industry. The leading companies in AI are large in terms of revenue, profits, and especially market capitalisation (high multiples on earnings). This has led to an increase in antitrust scrutiny of the leading technology firms from governments (particularly the European Commission) and in the press (see, for example, The Economist’s 20 January 2018 cover story, “The new titans, and how to tame them”, and their subsequent story, “The market for driverless cars will head towards monopoly”, on 7 June 2018). Much of this antitrust scrutiny focuses on the role of these firms as platforms, not on their use of AI per se. The feature that makes AI different is the importance of data. Firms with more data can build better AI. Whether this leads to economies of scale and the potential for monopolisation depends on whether a small lead early in the development cycle creates a positive feedback loop and a long-run advantage. 

Much of economic policy for AI is simply economic policy. For the diffusion of AI, it resembles innovation policy. For the consequences of AI, it resembles public policy (the social safety net) and competition policy (antitrust).  We summarise aspects of economic policy for AI in Table 1.

Table 1 Aspects of economic policy for artificial intelligence


Although AI is like other technologies in many respects, it is unusual in a few important dimensions.  Specifically, AI is both a general purpose technology (GPT) – i.e. it has a wide domain of applications – as well as an ‘invention of a method of invention’ (IMI) (Cockburn et al., 2018; Agrawal et al. 2018). Cockburn et al. assert that “… the arrival of a general purpose IMI is a sufficiently uncommon occurrence that its impact could be profound for economic growth and its broader impact on society.” They assemble and analyse the corpus of scientific papers and patenting activity in AI, and provide evidence consistent with the characterisation of machine learning as both a GPT and IMI.

The implication concerns the returns to investments in AI policy design. Due to the breadth of applications, the cost of suboptimal policy design will likely be significantly higher than with other technologies – or the benefits of optimal policy greater. Furthermore, the returns to investments in policy design are not only a function of the directeffects, where AI “directly influences both the production and the characteristics of a wide range of products and services”, but also the indirecteffects because “AI also has the potential to change the innovation process itself, with consequences that may be equally profound, and which may, over time, come to dominate the direct effect” (Cockburn et al. 2018).

Authors’ note: The points we raise in this column are based on Agrawal et al. (2018a), which in turn builds on discussions at the 2017 NBER Conference on the Economics of AI in Toronto and the associated conference volume (Agrawal et al. 2018c). 


Acemoglu, D, and P Restrepo (2018), “Artificial Intelligence, Automation and Work”, in A Agrawal, J Gans, and A Goldfarb (eds), The Economics of Artificial Intelligence: An Agenda, University of Chicago Press. 

Agrawal, A, J Gans, and A Goldfarb (2018a), “Economic Policy for Artificial Intelligence”, NBER Working Paper 24690. 

Agrawal, A, J Gans, and A Goldfarb (2018b), Prediction Machines: The Simple Economics of Artificial Intelligence, Harvard Business School Press.

Agrawal, A, J Gans, and A Goldfarb (eds) (2018c), The Economics of Artificial Intelligence: An Agenda, University of Chicago Press. 

Agrawal, A, J McHale, and A Oettl (2018), “Finding Needles in Haystacks: Artificial Intelligence and Recombinant Growth”, in A Agrawal, J Gans, and A Goldfarb (eds), The Economics of Artificial Intelligence: An Agenda, University of Chicago Press.

Cockburn, I, R Henderson, and S Stern (2018), “The Impact of Artificial Intelligence on Innovation”, in A Agrawal, J Gans, and A Goldfarb (eds), The Economics of Artificial Intelligence: An Agenda, University of Chicago Press.

Galasso, A, and H Luo (2018), “Punishing Robots: Issues in the Economics of Tort Liability and Innovation in Artificial Intelligence”, in A Agrawal, J Gans, and A Goldfarb (eds), The Economics of Artificial Intelligence: An Agenda, University of Chicago Press.

Goldfarb, A, and C Tucker (2012), “Privacy and Innovation”, in J Lerner and S Stern (eds), Innovation Policy and the Economy, Volume 12, NBER, University of Chicago Press: 65-89.

Sutton, T (2018), “An Overview of AI Strategies”, Medium, 28 June.

Stevenson, B (2018), “AI, Income, Employment, and Meaning”, in A Agrawal, J Gans, and A Goldfarb (eds), The Economics of Artificial Intelligence: An Agenda, University of Chicago Press.


[1] (accessed 22 May 2018).