Edited by John Siegfried, Vanderbilt University 1

edited by John Siegfried, Vanderbilt University 1. Economic research has had a profound effect on our way of life and well‐being, affecting income and wealth, as well as mortality, health, happiness, and welfare. 2. Economics research success is not widely recognized. ¾ Expectations are so high for an unfettered economy that reality is almost always interpreted as failure ¾ Most economic research produces public rather than private goods ¾ Patents and the history of science do not include economics ¾ Process of moving from innovative economic science to human welfare improvements via policy change is largely a mystery

2 1. Tradable pollution permits Tom Tietenberg

2. Price index reform Michael Boskin

3. Antitrust reform (entry considerations) Lawrence White

4. Matching physicians and students Al Roth

5. Spectrum auction design Preston McAfee, John McMillan, Simon Wilkie

6. Airline deregulation Elizabeth Bailey

7. Welfare to work reform Rebecca Blank

8. Earned Income Tax Credit Robert Moffitt

9. Trade liberalization Anne Krueger

10. Saving for retirement David Laibson, Brigette Madrian, John Beshears, James Choi, Brian Weller

11. Monetary policy John Taylor

12. Voluntary military force Jim Miller, John Warner, Beth Asch

3 Inflation indexed bonds Lawrence Summers

Road congestion pricing Herbert Mohring

Peak load electricity pricing Arthur Hadley, 1880s, Marcel Boiteaux, Peter Steiner

Auctioning airport landing rights Elizabeth Bailey, Alred Kahn

Indexed mutual funds Burton Malkiel

Mutual assured destruction (MAD) Thomas Schelling

National income accounting Ezra Seaman (mid 1800s), Simon Kuznets, George Jaszi

Do-not-call list FTC staff

4 Main idea: • Bureaucrats do not have sufficient information to identify least‐cost pollution avoiders. • A market for pollution permits causes high‐ cost pollution avoiders to buy permits and low‐cost pollution reducers to sell them, so long as they receive more for the permits than their cost of reducing pollution. Having sold their permits, low‐cost pollution avoiders cut pollution, and high‐cost pollution avoiders still pollute. Emissions reduction is achieved at minimum cost.

5 • Ronald Coase (1960): if property rights are explicit and transferrable, markets will assure that pollution is controlled at the least cost. • Connecting the CoaseTheorem to control pollution John Dales (1968) water Tom Crocker (1966) air • William Baumol and Wallace Oates (1971): a uniform pollution tax achieves an environmental target cost‐ effectively, as lower cost abaters choose abatement, and higher‐cost abaters pay the tax. That leaves (only!) to identify the optimal tax. A market for tradable permits does that.

6 • Empirical cost‐effectiveness studies showed that reduced ambient standards could be attained at much lower cost with cap‐and‐trade than command and control when 1970s growth required larger emissions cuts. • 1976 EPA “offset policy” allowed new businesses into local areas (and sustained economic growth) if established firms reduced pollution by 120% of the new pollution that would be added by entrants. New firms had to secure credits for offset. The 20% was “retired.” EPA pressed into policy by areas hitting emissions limit implying no growth. • Soon economic growth reduced air pollution, mitigating growth as evil. • Sulfur allowances for controlling electric utility emissions contributing to acid rain traded since 1980. The allowances now are traded on the Chicago Board of Trade (which lowers costs of locating buyers and sellers). • Caps can be tighter and pollution reduced if abatement is allocated more efficiently, pleasing environmentalists. Remaining issue is how to distribute permits.

7 Main idea:

• Matching people in transactions when each has rank order preferences among those on the other side is challenging. E.g., marriage, school admissions.

• In a deferred acceptance algorithm (one side makes a proposal to agents on the other, proposing first to their first choice), those who receive more proposals than they can accept reject the least preferred, but don’t immediately accept the best received; instead, they hold it without commitment, deferring acceptance until they see all options by the end of the choice period. This produces a match that is better than matches produced by less centralized marketplaces they replace.

• This idea has been applied in recent years to match medical school graduates to hospitals seeking residents, and to allocate students among high schools in New York City and Boston by Al Roth and colleagues.

8 Gale and Shapley (1962) developed an algorithm for two‐sided matching, in which men and women in a marriage market each have preferences over individuals of the opposite sex. It has the attribute that each individual is matched to his or her preferred person among those on the other side of the market that find the individual acceptable, after learning of their options. This is called a stable match, in that no individual can find someone else on the other side of the market that they prefer and who will have them.

Roth (1982) showed that there is no stable matching mechanism that makes stating true preferences always best for each agent, no matter what preferences are stated by others. Roth also showed that it is possible to design the mechanism so that one side of the market can never do any better than to state its true preferences.

9 Medical residency placement used such an algorithm since 1952 to replace a process in which many hospitals whose residency offers were rejected were unable to land their second choices, because the second choices had accepted offers from other hospitals.

The problem that arose in the 1980s is married physician couples, who want to locate in the same place. Roth (1985) showed that if couples were modeled with preferences over pairs of positions, the National Resident Match could be improved. In 1995, he was invited to redesign the resident match.

The school choice problem. In New York, a deferred acceptance algorithm replaced a system that used preferences for only 100,000 of 130,000 students each year. The remainder was allocated administratively, without regard to preferences. Because those who did not get their first choice were allocated arbitrarily, parents often did not state their true first choice, but would substitute a lower acceptable choice where they had a higher probability of success. This led to inefficiencies. The re‐designed algorithm avoids this problem.

10 Main issue:

• Question in early 1970s: could the airline price and entry regulation imposed in the 1930s be replaced by competition, to create better service at lower prices for both passengers and cargo shippers?

11 • Dick Caves (1962): no evidence of economies of scale in airline systems, undermining “natural monopoly” as a basis for regulation.

• Michael Levine (1965) and William Jordan (1970): unregulated intrastate air passenger fares were lower by half than comparable regulated interstate fares. The carrier (PSA) was profitable.

• George Douglas and Jim Miller (1974): price regulation without flight frequency controls leads to too many flights and load factors that are too low. Customers might prefer lower quality service, but with much lower prices.

• Contestability theory (Baumol, Bailey, Panzar, Willig): the threat of entry keeps fares low even if actual concentration is high.

12 In 1975, Stephen Breyer (on‐leave from Harvard, now Supreme Court Justice) organized hearings held by $18.75 Senator Kennedy, asking whether prices were too high. 377 miles

Hearings on the price differential between the LA to SF route and comparable Boston to Washington route used by Senator Kennedy flown on the same aircraft. The former was unregulated, as it was intrastate; the latter was regulated.

Reform became bi‐partisan. Ford Administration CEA chair Alan Greenspan argued for deregulation in 1975 and 1976.

$41.67 409 miles

13 • Existing regulation often granted duopoly authority over routes, and price setting control by duopolists. With no control over flight frequency, if price exceeded costs, frequency increased until load factors declined enough to extinguish all profits. Airlines made no money, planes were half full, service was good, and prices were twice what they might have been. Douglas and Miller (1974).

• Charismatic economist Fred Kahn chaired CAB in 1977. He added economist Elizabeth Bailey as a Commissioner. Economists Michael Levine and Darius Gaskins were senior staff.

• Cargo deregulated in 1977 because no opposition from flight attendants union or passenger groups afraid of losing jobs or service. Cargo carriers’ stock prices rose, ending passenger carrier opposition.

• CAB conducted two experiments: 1) allow discounted fares for 35% of capacity, and 2) allow entry to underserved airports like Newark, Baltimore, Midway, San Jose. Fares dropped; service expanded.

• Passenger air service was deregulated in October 1978; the CAB disappeared in 1985. Fares are about 30 percent lower than what they would have been under regulation. Estimated consumer welfare gain: $28 billion per year.

• Success of Southwest served as catalyst. Today fares are 40% lower where Southwest is likely entrant (where it is already at one end of a point‐to‐point route).

14 Main idea: • In 1970 70% of US pension plans were defined benefit, employers making contribution level and portfolio decisions. By 1990, 70% were defined contribution. • Because of potential liability for choices, employers required employees to opt‐in to defined contribution plans. About 30% of employees opted in. • Behavioral economists discovered the power of inertia—most employees elect the default option, either because it is less costly to do so or because they believe it is endorsed by their employer. • Through research showing that employees elect the default more frequently, economists persuaded Congress to change the law in 2006, limiting employer liability for investment returns in 401(k) plans, and requiring that employees affirmatively opt‐out if they do not wish to participate. This largely reversed participation from 70% opting out to 70% opting in.

15 • Richard Thaler and others discovered that individuals do not make rational economic decisions like homo economicus as modeled by economists. Instead they often use rules of thumb, and make choices based on procedures rather than outcomes. • Laibson, Madrian, and other Harvard colleagues produced empirical evidence that individuals are not saving sufficiently for retirement, and that many are passive in their retirement choices, deferring the choice to their employer.

16 • Elimination of defined contribution pension plan participation from the non‐discrimination test between highly compensated (above $105,000 in 2008) employees and other employees in order for taxes on contributions to be deferred. This allowed employers to require participation, but at a very low level compared to preferences of high income employees, to reduce complaints of low income employees.

• Peter Orzag of Brookings (now Director of OMB) personally lobbied Congress to get the PPA passed. Clear evidence that default enrollment was an effective way to increase savings, without limiting freedom of choice. Orzag personally lobbied Congress hard.

17 Main idea:

• Real economic welfare is greater if GDP grows faster and inflation is lower. • Businesses and investors make better decisions, thus increasing real GDP and reducing inflation, if the volatility of GDP and prices is lower. • Making policy predictable reduces uncertainty, making it easier for the private sector to plan, thereby reducing the volatility.

18 • Lucas (1976) argued that decision makers rationally anticipate future policy in current decisions. Their decisions are better if future policy is transparent, because there will be fewer changes reacting to surprises. Policy can be transparent if it follows rules to minimize the (often weighted quadratic) combination of the volatility of real GDP and inflation (Sargent and Wallace, 1975; Kydland and Prescott, 1977; Taylor, 1979).

• Firms set prices and hold them steady for some period of time. Prices do not adjust instantaneously, as naïve economic models imply. They are staggered and unsynchronized. With a current price decision expected to last into the future, some prices set in the future are relevant for today’s decision. Thus future inflation affects current price decisions. In other words, future inflation affects the current inflation rate, giving credibility to a central bank commitment to price stability and inflation targets.

19 • Regime change catalyst was Paul Volcker’s commitment in 1981 to break inflation expectations because it was causing poor decisions by investors and consumers.

• The new regime came from economics: inflation targeting, based on models and empirical results devised by economists, including FED research economists.

• The models revealed how the policy instrument—first the money supply, and later the interest rate—should be targeted. The models, solved with calculus of variations and estimated with data, showed policy should be more responsive to fluctuations than it had been. The money supply, and interest rates, should not remain fixed (as Friedman had argued). Since 1980 FED policy has about doubled the responsiveness to changes in the inflation rate.

20 • Professional economists at the FED read economics, and communicated with academic monetary economists. Many economists served on the FED since 1980—Lawrence Meyer, Alice Rivlin, Janet Yellen, Edward Gramlich, Roger Ferguson, Fredric Mishkin, Michael Moskow, William Poole, Charles Plosser, Randall Kroszner, Gary Stern, Lawrence Lindsay, Alan Blinder, etc. The FOMC used economic research to make decisions.

• The result was reduced volatility of real GDP, and inflation from 1982 through 2007, called the “great moderation.” Coterminous with the new monetary policy emphasizing inflation targeting.

• Central banks around the world now inflation target, causing predictable, rule‐like decision making, enhancing transparency.

• Today the FED announces its interest rate decisions immediately, and tries to explain its thinking about the future. Before 1980 the FED thought it had to keep monetary policy a secret and surprise markets to be effective.

21 Decline in the Volatility of the Growth Rate of Real GDP

22 Main idea:

Efficiency: A drafted soldier’s military pay does not reflect opportunity cost because the occupation was not chosen voluntarily. A voluntary force, while requiring more compensation to attract volunteers, costs less in real resources because it attracts people with lower opportunity costs.

Fairness: Because pay must rise to attract volunteers, those who serve do not incur economic losses just because of bad luck to be drafted. Others pay higher taxes to compensate the volunteers, taxes that recognize the opportunity costs of soldiers.

23 1. Opportunity cost of a volunteer force is less than opportunity cost of a conscripted force.

2. Change in relative input costs (higher price of labor) encourages substitution of capital for labor, especially capital requiring fewer operators and less maintenance, which also saves lives.

3. Voluntary force has lower turnover, lower training costs. Eliminating some training initially saved 6 percent of total force.

4. A volunteer force eliminates the costs of avoiding conscription.

5. Economic burden of a voluntary force is shared more equitably.

24 Controversy over Viet Nam War was catalyst. Friedman proposed volunteer force in Capitalism and Freedom (1962). An economist working at Department of Defense in 1964, Walter Oi, takes on the issue.

Lee Hansen and Burt Weisbrod (1967), Stuart Altman and Alan Fetcher (1967), Anthony Fisher (1969) and Bob Barro (1971) contributed. Virginia Ph.D. students wrote a collection of essays advocating a volunteer army (James Miller [later Director of OMB], 1968). Spawned by 11:00 pm news broadcast in 1967 reporting the ping‐ pong ball lottery. Both Friedman and Galbraith endorsed the book.

Martin Anderson, an economist at Columbia, suggested ending the draft to the 1968 Nixon presidential campaign. Impressed by arguments, Nixon advocated ending the draft in an October 17, 1968 campaign speech.

25 • Once elected, Nixon established the President’s Commission on an All‐ Volunteer Armed Force. Milton Friedman, who advocated a volunteer military in Capitalism and Freedom, Allen Wallis, economist and president of University of Rochester, and Alan Greenspan were members. Economists William Meckling, Walter Oi and Robert Barro were staff.

• In November 1970 the Commission recommended abolishing the draft, increasing military pay, and improving living conditions.

• In 1971 Nixon sent bill to Congress, which held hearings. Walter Oi’s testimony was persuasive. “His candor, knowledge, and willingness to challenge DOD’s data about the negative consequences of a volunteer force undoubtedly helped Committee members feel more comfortable with an all‐volunteer approach” (John Ford, staff director of House Armed Services Committee at the time, 2004).

• The Committee voted to switch to a volunteer force after two years. The bill passed, was signed in September 1971, and the volunteer military commenced in 1973.

26 Main idea: • Traditional American welfare programs administered by states with federal funding assistance focused on single mothers, who received a grant that declined as earnings increased, creating an incentive for welfare recipients not to work. This affected their experience and lifetime earnings. • Reducing the tax rate on earnings, and requiring labor force participation for eligibility were expected to reduce welfare rolls enough that the cost of the programs would decline, even though working mothers’ benefits increased.

27 If workers retain a greater portion of their compensation after taxes, they will work more. The catalyst for reform was state‐level random assignment experiments which produced compelling evidence supporting the hypotheses.

28 1994 U.S. election led to a conservative Congress, concerned with welfare payments to single mothers not in the labor force, and fatherless families. President Clinton and many Democratic state governors shared these views, creating opportunity for change. Governors saw reform as a way to reduce state spending.

Some states experimented with welfare‐to‐work programs in the 1980s. Economists conducted randomized assignment experiments. They showed that welfare‐to‐work programs reduce welfare use and increase work.

The Personal Responsibility and Work Opportunities Act (welfare reform legislation) passed in August 1996. It left design of welfare programs to states, but required states to move a significant share of welfare recipients into work.

Many states imposed reduced benefits or loss of eligibility on recipients who did not participate in welfare‐to‐work programs. Many reduced the benefit reduction rate, the “tax rate” on earnings.

29 30 Main idea:

• A price index agglomerates information about prices of multiple goods and services over time. It compares similar goods and services, controls for quality changes and shifts in distributional channels, and adjusts for buyer substitution from higher to lower price items. It weighs price changes of more important (larger expenditure) items more heavily.

• The inflation rate is used to ask “How much more income do consumers need to be as well off with a new set of prices as an old set of prices?”

• Economic research has refined the U.S. consumer price index (CPI) for quality changes, changes in distribution channels, and consumer substitution away from items whose prices are rising relatively rapidly.

31 • Irving Fisher, The Making of Index Numbers (1922): Percentage changes can be computed relative to an original base period price, or to a final period price. Theory supports an average of the two, using a geometric mean, that weighs less heavily items that consumers substitute away from as their prices rise relatively faster. BLS now computes a “chained‐ CPI,” that rises slower than a base period price index, because consumers substitute away from items whose prices rise relatively faster. • Erwin Diewert (1976) developed a category of superlative indexes that adjust for consumer substitution to lower priced alternatives. • Hausman and Leibtag (2004, 2005) computed the bias in CPI created as consumers switch to lower margin distribution outlets (e.g. Sam’s Club, WalMart) for items whose prices rise faster, while CPI data are collected within a fixed set of distribution outlets. • Bresnahan and Gordon: Biases resulting from failure to adjust fully for quality change.

32 • Effect of the CPI on union COLA clauses, on social security payments, on income tax brackets, etc.

• Boskin Commission (1996) [including Ellen Dulberger, Robert Gordon, Zvi Griliches, and Dale Jorgenson]: identified several likely biases in the CPI and recommended improvements. Some have been implemented. Boskin pushed this as Bush I’s CEA chairman.

33 Main idea: Competitive pricing is promoted by either existing competition or potential competition. Thus, the ease of entry is as important for creating and preserving competition as is the number and size distribution of existing sellers.

34 • The role of entry in promoting competition was first noted by Joe Bain (1956). By 1970s, entry conditions became important in economic analysis of market power, to a large degree replacing market share and concentration as the central structural characteristic for deciding competitive impact.

• If scale economies are important, and entry is easy, large firm size may lower costs without monopolizing.

35 1) developments in industrial organization thinking 2) elevation of economists’ status and positions in enforcement agencies 3) wider participation of economists in antitrust litigation • In the 1950s Carl Kaysen, a Ph.D. student at Harvard, was appointed by Judge Charles Wyzanski to help him in the monopolization trial of United Shoe Machinery Company. • The Federal Trade Commission (FTC) and Justice Department’s Antitrust Division had economists before 1960. The FTC hired more in the 1960s, under Willard Mueller. In the 1960s Assistant Attorney General for Antitrust, Donald Turner established the Special Economic Assistant. A series of economists served one‐ year terms in the position. In 1973, George Hay convinced the Division to form an Economic Policy Office. • The 1970s major suits against IBM, Kellogg et al., Exxon et al., and AT&T had respected economists on both sides. In the 1980s, for the first time, some FTC Commissioners were economists. • DOJ promulgated Merger Guidelines in 1982, written largely by economists. The Guidelines explain the economic basis for merger enforcement, focusing on coordinated effects and unilateral effects. For the first, cartel theory is relevant. Stigler (1946) explained how the probability of a successful cartel is affected by concentration. For the second, the dominant firm price leadership model. It shows how market share and entry conditions affect the extent to which a firm internalizes the benefits of monopolization. • Spring 1982: Economists at DOJ develop “hypothetical monopolist” paradigm to solve market definition problem. Unbeknownst to them, Morris Adelman (1959) uncovered idea earlier.

36 Main idea:

• With the development of cell phones, in 1993 President Clinton signed a law granting the Federal Communication Commission (FCC) authority to auction spectrum licenses. Prior allocation methods—administrative allocation by comparative hearings until 1982, and then allocation by lottery—ran into difficulty as license values increased so much that large expenditures were devoted to securing the licenses by hearing (rent seeking), or lottery allocation of licenses did not get them to where they were valued most.

• The idea was to use an auction to allocate spectrum rights to their most valuable use without squandering the surplus value on rent‐seeking.

• Originated with Coase (1959), separate article from Theory of Social Cost.

37 The format of the auctions is important for allocating rights to their most valuable use. Design reflected trade‐offs that were understood only because of the development of auction theory in the 1980s.

1. One‐time sealed bid vs. multiple ascending rounds. One‐time sealed bid reduces collusion among bidders but increases malfeasance by government officials (bribery). Ascending bids reveal information and increase bids (Milgrom and Weber, 1982). Chose ascending multiple round bidding.

2. Simultaneous vs. sequential bidding. In sequential bids, bidders may lay back and wait to see others’ bids. Solution is an activity rule (Milgrom and Wilson) requiring each bidder to either hold a high bid or make a new bid each period. Now known as “FCC auction.”

3. Combinatorial bids. Bidders allowed to bid on combinations as a unit.

4. Reserve prices. Insures value maximization in case seller has highest value, either in current use or in expected future sale value, taking into account future technology.

5. Size of minimum increase between bids. Too small wastes time. Too large and the auction allocation may be inefficient, as bidder with the 2nd highest value wins because the high value bidder is unwilling to pay the second‐highest value plus the increment. FCC adopted rule, that increment declines as competitors dropped out.

6. Ownership caps. Limit on the amount of spectrum any one participant may hold in a given geographic area to insure competition in the final product market. Caps set to guarantee at least three new entrants in most areas.

38 • FCC economist Evan Kwerel was the driving force. The FCC hired John McMillan to advise in design. The bidders hired game theorists—Paul Milgrom, Robert Wilson, Preston McAfee, Barry Nalebuff, Peter Cramton, David Salant, Robert Gertner. • The absence of a history of auctions helped. No need to push out an incumbent system. All designs were on the table. • Economists emphasized four goals: simplicity, efficiency, revenue, diversity. The economists focused on allocating licenses to those who valued them the most—efficiency. • Revenue was important, not only to assist Treasury, but also as a signal that the licenses went to their most valuable use.

39 Main idea: • Helping the poor and creating incentives for poor people to work conflict. If welfare is reduced by earnings, a recipient faces a 100% tax, and has no incentive to work. • The EITC idea started in the 1960s as a “negative income tax,” (NIT) in which benefits decline by less than earnings as earnings increase. • Reducing welfare benefits by less than earnings as earnings rise creates a work incentive.

40 Milton Friedman proposed NIT in Capitalism and Freedom (1963), devised with Stigler in the late 1940s at the University of Chicago. Robert Lampman of Wisconsin (1965) and James Tobin of Yale (1966) also advocated an NIT. Supported by both conservative and liberal economists, the idea was popular.

An earnings subsidy has even greater work incentives [Kesselman (1969); Barth and Greenberg (1971)], but provides no benefits for those who are most destitute. Robert Haveman pushed earnings subsidies in 1970s.

Friedman’s proposal did help the most destitute; this is the difference between NIT and EITC.

Solution is to combine fixed supplement and EITC, with the supplement declining with earnings by less than the EITC subsidies them. This increases work incentives and helps the most destitute.

¾ The EITC is administered by the IRS, thus costs less than state welfare programs.

¾ Everyone is eligible, so other incentives (e.g. to have more children) are reduced

¾ EITC is a national program, reducing incentives to move to high‐benefit states; benefits are higher where wages are higher (and possibly higher costs of living).

¾ Because it is a cash‐based system, if it replaces some in‐kind systems, it is of more value to recipients than the cost of the in‐kind subsidies.

¾ The stigma of participating is low, and participation high.

41 Economists in 1960s found evidence of strong work responsiveness of women to financial incentives. President Nixon recommended a NIT to Congress in 1969. It passed the House, but failed the Senate, opposed by ideologically conservative Senators (whose constituents had the most to gain!)

Field experiments on the NIT conducted at the University of Wisconsin in the 1960s found considerable work responsiveness to financial incentives.

Senator Russell Long (D, La.), who voted against the NIT, subsequently introduced several positive work incentive bills for families with children.

His staff member, Michael Stern, was influenced by research by Robert Haveman at Wisconsin . After failing three times, EITC passed in 1975 with a 10% subsidy up to $4,000 of earnings, phased out by $8,000 in earnings.

EITC has since been expanded, and indexed; 2007 subsidy rate is 40% on earnings up to $11,790 for families with 2 children, stays constant for earnings up to $15,390, and phases out when earnings reach $37,783.

42 Main idea: • Stagnant Economic growth in low income countries in 1950s; lowest income countries even had falling per capita incomes. Infant mortality over 10%, life expectancy low (34 in India). • Policies were similar—import substitution to develop manufacturing—in these 60‐70% rural economies, helped by barriers (import tariffs and licensing requirements) to protect manufacturing industries. • Knowing they were protected, firms ignored costs and quality, producing substitutes for imports rather than goods they could export, where they had a comparative advantage. Balance of payments crises evolved, as little foreign exchange was earned from exports, stifling purchases of inputs to produce import substitutes. Protection increased in order to help them. Development stagnated. Currencies devalued. The idea: abandon import substitution in favor of freer trade.

43 Comparative advantage (David Ricardo, early 1800s). All countries can be made better off if each specializes in producing that for which it has the lowest opportunity cost, and trades with others. Not everyone has to manufacture to prosper.

44 In the 1950s, S.C. Tsiang, a Chinese‐American economist, persuaded Taiwan of the advantages of trade. Seeing that foreign aid was ineffective, in 1958 Congress reduced it, creating a crisis. Korea noticed Taiwan, and followed its lead. By 1960, both countries had liberalized trade; growth rates skyrocketed. During 1960s import duties were further reduced. Soon, Singapore and Hong Kong followed.

In the 1970s and 1980s more countries abandoned import substitution. Living standards rose; hundreds of millions of people lifted out of poverty. New Zealand, India and China followed in 1990s. India’s life expectancy is now above 60.

NBER organized a 10‐country research project in late 1960s to identify difficulties in developing countries. Nine country studies were published. They identified the effect of protectionism on incentives, quantifying the effects of the trade regimes on exchange rates, balance of payments, and efficiency. The totality of findings was dramatic. Inefficiencies encouraged by import substitution were glaring.