The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
u -- E o c o U LII FEDERAL RESERVE BANK OF DALLAS FOURTH QUARTER 1998 How Increased Product Market Competition May Be Reshaping America's Labor Markets jolm V. Duca Privatization and the Transition to a Market Economy jason L. Sewing Global Warming Policy: Some Economic Implications Stephen P A. Brown This publication was digitized and made available by the Federal Reserve Bank of Dallas' Historical Library (FedHistory@dal.frb.org) Economic Review Federal Reserve Bank of Dallas Robert D. McTeer, Jr. Presldenl and Chief Executive Officer Helen E. Holcomb Firsl Vice President and Chief Operaifng Ollicer Harvey Rosenblum Senior Vice President and Director of Research W. Michael Cox Vice President and Economic Advisor Senior Economists and Assistant Vice Presidents Stephen P. A. Brown John V. Duca Robert W. Gilmer Evan F. Koenig Director, Center for Latin American Economics, and Assistant Vice President William C. Gruben Senior Economist and Research Dfficer Mine K. YOcel Economists Robert Formaini David M. Gould Joseph H. Haslag Keith R. Phillips Slephen D. Prowse Marci Rossell Jason L. Saving Fiona D. Sigalla Lori L. Taylor Lucinda Vargas Alan D. Viard Mark A. Wynne Carlos E. J. M. Zarazaga Research Associates Professor Nathan S. Balke Southern Methodist University Professor Thomas B. Fomby Southern Methodist University Professor Gregory W. Huffman Southern Methodist University Professor Finn E. Kydland Carnegie Mellon University Professor Roy J. Ruffin University of Houston Editors Stephen P A. Brown Evan F. Koenig Publications Director Kay Champagne Copy Editors Jennifer Afflerbach Anne L. Coursey Monica Reeves Design & Production Laura J. Bell The Economic Review is published by Ihe Federal Reserve Bank 01 Dallas. The views expressed are Ihose of Ihe aulhors and do nol necessarily reliecl Ihe positions ollhe Federal Reserve Bank 01 Dallas or Ihe Federal Reserve Syslem. Subscrlplions are available Iree 01 charge. Please send requesls lor single-copy and multiple-copy subscriptions. back issues. and address changes to the Public AIIairs Departmenl. Federal Reserve Bank of Dallas. P.O. Box 655906, Dallas. TX 75265-5906. (214) 922-5257. Articles may be reprinted on Ihe condilion Ihal Ihe source is credited and Ihe Research Deparlmenl is provided wilh a copy 01 the publication conlalning Ihe reprinled malerial. How Increased Product Market Competition May Be Reshaping America's Labor Markets John V. Duca Page 2 Privatization and the Transition to a Market Economy Jason L. Saving Page 17 Global Warming Policy: Some Economic Implications Stephen P. A. Brown Page 26 In this article, John Duca discusses how and why compensation has become more market sensitive in the United States. Specifi ally, he illustrates how fierc I' product market competitio n ca n theoretically reduce the prevalence of nominal wage contracts and of indexation in such contracts, while boosting the use of profit sharing. He also summarizes mpirical findings supporti ng the view that increased competition has reduced the u e of nominal contracts and the indexation of contract wage, and presents limited, inconclusive data supporting the vi w that greater product market competitio n has boosted the overall use of profit haring. Consistent with aggregate movements in labor practices and a measure of the degree of goods market competition, industry-level data are presented that indicate these changes in labor practices are most evident in sectors that have experienced either deregu lation or increased foreign competition since the late 1970s. Whi le more resea rch needs to be done, particularly using industJy-level data, new theoretical arguments and empirica l evidence support, but do not conclusively prove, the view that increased product market competition has been reshaping America's labor markets. The Chinese government has an nounced its intention to privatize thousands of state-owned enterprises. Such an effort wou ld dwarf recent privatizations in the industrialized West and be comparable only to the Eastern European ex:peri ence fo llowing the fall of the Soviet Union. As such, an exa mination of the Eastern European privatization may provide valuable Ie sons for China and any other developing economy that embarks upon a large-sca le privatization program. In this article, Jason avi ng considers three problems with which Ea tern European privatization have had to contend : a sca rcity of information, an inability to exercise managerial overSight, and the ab ence of competitive markets. He suggests that a lack of information need not prevent privatization . He explores the holding com pany as a potential solution to the managerialoversight problem. And he sugge ts that effective privatization requ ires both managerial overs ight and a lega l framework that p 1'mits freedom of entIy for competi ng firms. Many ana lysts believe that the emissio ns of greenhouse ga s resulting from human activity are contributing to global wa l'lTling, but dle linkage i highly un certain . Th largest such source of these gases is carbon dioxide (C0 2) from the growing consumptio n of fossil fu Is. Consequently, the conse rvation of fossil fuel figures prominendy in any strategy to reduce dle threat of globa l warming. Because there is considerable uncertainty about dle ben fits of reducing CO 2 emissions but the cost of conservation can be read ily quantified, some ana lysts have suggested that reducing th em issio ns is like insurance. In this articl e, t phen Brown integrates a growing literature on dle damage ca used by globa l warming with a world en rgy model to do a cost-benefit analysis of U.. comp liance with the accord adopted at th United Nations conference on globa l warming held in late 1997. His analysis shows dlat reducing U.S. e miss ions to comply with th accord wou ld I' present too much insurance aga inst global warming. In recent years, inflation has drifted lower as the unemployment rate has fallen below trigger levels that have been associated with rising inflation. Indeed, since mid-1996 the unemployment rate has been 5 percent or lower —well below the 5.5 percent to 6 percent trigger-point estimates of many economists— while core inflation has remained low.1 Although wage inflation has drifted higher and worker shortages abound, there is less upward wage pressure than prior experience would suggest. These developments have spurred a reexamination of the view that there is a trigger level of unemployment, below which inflation tends to rise and above which inflation tends to fall (see Phelps 1967; Friedman 1968). This trigger level is called the nonaccelerating inflation rate of unemployment (NAIRU). Some critics argue that the concept should be abandoned (Galbraith 1997), while others argue that the NAIRU is so imprecisely estimated it has limited use as a policy guide (Staiger, Stock, and Watson 1997). However, these critics do not shed light on why the relationship between unemployment and changes in inflation has apparently shifted. Others maintain that the NAIRU needs to be modified for factors such as changing demographics (see Gordon 1997), sociological changes (Blanchard and Katz 1997), or the impact of unemployment spells on worker skills (also known as hysteresis effects; see Blanchard and Katz 1997). The weaknesses of these conclusions are that demographic adjustments to the unemployment rate cannot explain the behavior of the 1990s, sociological factors are difficult to track, and the evidence on hysteresis effects for the United States is weak. Another shortcoming of these arguments for modifying the NAIRU is that they are inconsistent with both the U.S. economy’s performance in the 1990s and the perceptions of firms and workers about why the relationship between inflation and unemployment may have changed. For example, while inflation, GDP growth, and unemployment imply that the U.S. economy has performed well at a macroeconomic level in the mid-1990s, there has been an increased sense of job insecurity at the firm or industry level, after adjusting for different phases of the business cycle. An alternative explanation for the combination of these macro- and microlevel developments is that we are in a new era in which technological innovation and global trade are curtailing inflation. Under this new paradigm, two sources of low inflation are (1) cheaper How Increased Product Market Competition May Be Reshaping America’s Labor Markets John V. Duca Senior Economist and Assistant Vice President Federal Reserve Bank of Dallas N ew theoretical arguments and empirical evidence support—but do not conclusively prove—the view that increased product market competition has been reshaping America’s labor markets. 2 Federal Reserve Beige Book Reports of Competition and Inflation imports from higher worldwide capacity and (2) innovations that boost productivity. The jury is still out on these two sources. First, although import prices have fallen, much of the drop is a result of a stronger dollar and declining oil prices. Consequently, it is difficult to decipher any impact of global capacity. Second, it is not yet clear that long-run productivity growth has risen above its post-1972 average. Even if statistics understate growth in productivity, an upward shift in productivity growth does not necessarily mean that the NAIRU is lower. Faster productivity growth may initially offset increasing wage growth resulting from low rates of unemployment. However, the NAIRU hypothesis implies that keeping unemployment at current levels will spur continued increases in wage growth that eventually will more than offset the higher level of productivity growth and will boost inflation.2 Another explanation for the combination of low unemployment and low inflation is that fiercer product market competition has restrained firms from hiking prices and wages and is inducing firms to change the ways they employ and pay workers. This explanation is sometimes associated with “new paradigm” arguments, but it is not based on faster productivity growth (Duca 1998). The “competition” explanation is consistent with reports from Federal Reserve Beige Book respondents (see the box entitled “Federal Reserve Beige Book Reports of Competition and Inflation”). This article reviews how and why increased product market competition may be reshaping America’s labor markets. The analysis focuses on three important trends in U.S. labor markets. The first is the Several Federal Reserve Beige Book reports are consistent with the view that heightened competition has curtailed inflation in recent years. While the anecdotal reports are suggestive and are not sufficient evidence alone, they are consistent with the view that the degree of competition is higher now than in previous expansions. With this in mind, consider the following noteworthy excerpts from recent Beige Books covering the mid-1990s: Manufacturers note that competitive pressures are restraining prices for most products. (April 1994, iii) Most Districts report that competitive pressures continue to temper price increases on the output side. (June 1994, iv) Price increases are noted among a broad range of business materials.... However, virtually all Districts report that competitive pressures are holding prices down at the retail level.... (July 1994, ii – iii) Industrial materials prices edged up further, but businesses say that competitive pressures continue to restrain price increases on finished goods. (September 1994, i) Chicago, Kansas City, and Dallas report that intense competition among retailers kept prices flat despite increased input costs. (March 1995, v) Contacts in the temporary services industry said despite wage pressures they could not raise fees because of fierce competition. (November 1995, v) Competition continues to drive down retail prices in mall stores. (Boston District, November 1995, I-1) Several districts attribute the modest size of upward price movements to competitive conditions. (March 1996, iii) declining use of medium-term nominal wage contracts. This trend is evidenced by the falling share of private workers represented by unions, which negotiate labor contracts that typically set future nominal wages (Figure 1 ).3 Another important trend has been a decline in the share of union contracts that index or adjust wages for inflation according to a preset formula. As Figure 2 indicates, inflation risk, as tracked by the percent change in the U.S. Consumer Price Index (CPI), is positively correlated with the use of indexation. The drop in inflation since the early 1980s likely explains much of the decline in indexation—especially in the 1980s (Holland 1995). However, inflation is not the only factor behind this decline because inflation in the early 1990s was at levels near those of the 1950s, while indexation was twice as prevalent in the earlier period. Duca and VanHoose (forthcoming b) provide theoretical arguments and evidence that this difference may reflect a greater degree of product market competition. The third significant trend has been the rise of profit sharing (Duca and VanHoose, forthcoming a), as shown in Figure 3, which plots the incidence of profit sharing in pension plans (Bell Figure 1 Unions on the Decline Percent of private payrolls 40 35 30 25 20 15 10 ’56 ’61 ’66 ’71 ’76 ’81 ’86 ’91 ’96 SOURCES: Troy and Sheflin (1985); Bureau of Labor Statistics; Duca and VanHoose (1998b). FEDERAL RESERVE BANK OF DALLAS 3 ECONOMIC REVIEW FOURTH QUARTER 1998 market competition can affect labor practices in a world where firms face aggregate and industryspecific shocks. Rather than rederiving the results of other papers, the section below describes the basic theoretical framework common to these studies and then reviews the intuition behind key findings on contracts, indexation, and profit sharing, using charts or simple equations. Figure 2 Fewer Union Contracts Are Indexed for Inflation Than in the 1950s Percent Percent of union contracts 16 80 14 70 12 60 10 Indexation 50 8 Basic Theoretical Framework The economy contains a continuum of sectors, indexed by j, that are distributed uniformly between one and zero, which allows for aggregation of sector outputs. Each sector contains a large number of representative firms and workers, and relationships are expressed in logs (lowercase letters) with constants suppressed. A representative firm in sector j produces output (y) with employment (l ) according to 6 40 4 30 2 Inflation 20 0 ’56 ’61 ’66 ’71 ’76 ’81 ’86 ’91 ’95 SOURCE: Bureau of Labor Statistics. and Kruse 1995). Together, these trends can be viewed as making work and pay more sensitive to market conditions (Duca 1998). I begin with a discussion of the theoretical intuition for how changes in the degree of product market competition may have spurred these changes in labor practices. I then review and interpret empirical evidence on these theoretical implications. yj = αlj + θ, (1) where 0 < α < l, and θ is an aggregate supply shock raising output per hour in all sectors. Following Ball, assume that the demand for a firm’s output in sector j (yj ) as a share of total output (y) depends on relative prices according to the log-linear equation (2) yj – y = –⑀(pj – p) + δj , where ⑀ is the absolute size of the price elasticity of demand, the aggregate price level (P) = (∫Pj(1–⑀)dj )1/(1–⑀), the log of the price level (p) = ∫pj dj , and δj reflects sector-specific (independently and identically distributed) demand shocks that sum to zero. THEORETICAL LINKS BETWEEN LABOR PRACTICES AND THE DEGREE OF MARKET COMPETITION To analyze how goods market competition may affect labor practices, this article draws on several papers by Duca and VanHoose (1991, 1998a, forthcoming a, and forthcoming b), who combine the multisector frameworks of Blinder and Mankiw (1984) and Duca (1987) with the monopolistic competition framework of Ball (1988). The multisector approach permits the analysis of sectoral as well as aggregate shocks and allows for heterogeneity in labor practices. Thus, it is flexible enough to analyze how different susceptibility to sectoral shocks may affect pay practices and why some sectors may adopt nominal wage contracts or index such contracts for inflation while others do not. Incorporating Ball’s monopolistic competition framework permits the analysis of how labor practices are affected by changes in the overall price elasticity of demand. The greater this elasticity, the greater the extent to which firms compete in product markets. Thus, by combining aspects of multisector and monopolistic competition frameworks, it is possible to analyze how changes in the degree of product Figure 3 The Rise of Profit Sharing and Goods Market Competition Percent of workers with pension plans Index 22 8.3 20 8.1 18 7.9 7.7 16 Competition 7.5 14 7.3 12 Profit sharing 10 7.1 8 6.9 6.7 6 ’80 ’81 ’82 ’83 ’84 ’85 ’86 ’87 ’88 ’89 ’90 ’91 ’92 ’93 SOURCES: Bell and Kruse (1995); author’s calculations. 4 try depends more on supply shocks and sectoral demand shocks affecting the MRPL and less on the real purchasing power of wages (see Duca and VanHoose 1991). For simplicity, aggregate demand accords with the quantity theory of money (3) y = m + v – p, where m is the log of the money stock and v is a velocity shock common to all sectors. The shock to aggregate demand shock is (m + v), but for simplicity, money shocks are zeroed out to reduce notation and avoid adding complications associated with different monetary policy rules. For simplicity, each of the aggregate shock variables (θ and v) has an expected value equaling zero in logs, is independently and identically distributed, and has a fixed variance (σ 2θ for θ and σ2v for v). Sectors differ in how much sectoral demand shock variance they face, which is assumed to equal σ 2δ(1 – j )/j. Because the index number j is distributed normally between zero and one, the sectoral demand shock variances range from infinity for sector j = 0 to zero for sector j = 1. Converting Equations 1 through 3 into levels, combining the resulting terms into an expression for profits (πj = PjYj – Wj L j ), and differentiating implies labor demand is (4) Illustrative Arguments A firm’s labor demand depends on the change in total revenue due to a small change in labor hired. Under perfect competition, labor demand equals the price times the marginal physical product of labor (MPPL). The MPPL is related to the marginal cost of producing output (MC), shown in the left panels of Figure 4. Since the MPPL tends to fall as more labor is used, holding other factors constant, the competitive demand for labor (L Dj ) is downward sloping (upper right panel of Figure 4). Holding aggregate prices and price expectations constant, the labor-supply curve is upward sloping and shifts in line with changes in aggregate prices or price expectations. Since wages and employment reflect both labor supply and demand, the competitive outcome is given by point A. Under wage contracts, firms employ as much labor as they desire at a constant nominal wage that equals the prior time period’s expectation (t – 1) of conditions at time t: Et –1(wt ). Thus, contracts replace the spot labor-supply curve with a horizontal wage-contract curve, which only shifts in period t if a contract is indexed for inflation or profits. Under imperfect competition (lower panels), labor demand reflects the MRPL, which equals the marginal revenue of one more unit of output (MR) times the MPPL. Since boosting output lowers prices for an imperfectly competitive firm, marginal revenue is always below price. Thus, such firms produce at an output level below the perfect competition solution such that marginal revenue equals marginal cost (point S ; superscript SP denotes spot market variable solutions). l jd = [–⑀(wj – pj ) + (v + δj – pj ) + (⑀ – 1)θ]/[α + ⑀ – α⑀]. The first term in the numerator reflects that labor demand is declining in the nominal wage but increasing in terms of the sector-specific price. The second term reflects that labor demand rises to the extent that aggregate demand shocks boost overall prices relative to the sector j price or that sectoral demand shocks raise the relative demand for sector j ’s output. The last term in the numerator implies that a supply shock that raises productivity will boost labor demand. To make the model solvable, assume that labor is immobile across sectors in the short run owing to specialization and that labor supply is (5) Nominal Wage Contracting To analyze the extent of nominal wage contracting, assume that a share (Ω) of firms set contracts in period t – 1 that cover period t and set the nominal wage equal to the t – 1 expectation of the next period’s market-clearing wage. The remaining share (1 – Ω) of firms pays spot market wages. For now, assume that contracts are not indexed for inflation and there is no profit sharing. In addition, assume that workers choose between spot and contract wages to minimize a weighted average of the expected squared deviations of employment and the real wage from their market-clearing levels. While other loss functions are plausible, this assump- l j = c (w j – p), s s where c is a parameter > 0, reflecting that labor supply depends on the real purchasing power of wages. Equations 4 and 5 yield a reducedform solution for the market-clearing wage, which reflects a balance between labor supply and labor demand, where the former depends on the real wage and the latter depends on the marginal revenue product of labor (MRPL) in that industry. As goods markets become more competitive, ⑀ is higher and labor demand becomes more sensitive to the MRPL. Consequently, the market-clearing wage in an indus- FEDERAL RESERVE BANK OF DALLAS 5 ECONOMIC REVIEW FOURTH QUARTER 1998 Figure 4 Labor Markets Under Perfect and Imperfect Competition Product Market Labor Market Perfect Competition Pj Wj LjS MC A W jC W jC = E(W ), W jA A PjA D (industrywide) L jD = Pj × MPPLj Yj YjA Lj LjA Imperfect Competition Pj Wj MC LjS PjSP S A S W jC = E(Wj), W jSP W jC D L jD = MRj × MPPLj MR Y SP j Yj LSP j tion yields solutions that permit one to infer the qualitative impact of altering either the degree of market competition or the variances of the different shocks under reasonable parameter assumptions. Solutions reveal that relative to spot wages, wage contracts raise the exposure of real wages to aggregate demand shocks but reduce the exposure to sectoral demand shocks. Contracts also lower the variance of real wages Lj in the face of supply and sectoral demand shocks but raise the variance of employment. Thus, trade-offs emerge, and the extent of contracting depends on how workers weigh realwage versus employment stability.4 This model yields an interior solution such that there is a critical sector that is indifferent between using contract or spot wages, with workers more likely to opt for contracts in industries facing above-average sectoral demand variance. This 6 result arises for two reasons. First, for each sector, wage contracts introduce the same degree of employment and real-wage instability in the face of aggregate shocks. Second, since sectors face differing degrees of sectoral demand variance, contracts pose more real-wage stability relative to higher employment variance for workers in sectors with high sectoral variance. By assumption, the range of sectoral variance is so wide that workers in sector j = 0 face an infinite sectoral variance and will always choose to use contracts because the sectoral variance dwarfs the variances of aggregate shocks, whereas those in sector j = 1 face zero sectoral variance and will never use contracts. While these are theoretical extremes, the qualitative implications accord with the higher unionization rates in more cyclical industries. For some intermediate value of j, the sectoral and aggregate shock variances balance out, implying that there is some sector that is indifferent between nominal wage contracts and spot wages. This balance depends on the degree of product market competition. As the price elasticity of demand rises, the cost of potential job losses induced by nominal wage contracts rises relative to the benefits of realwage stability, thus reducing the net benefits of wage contracts and causing some sectors to shift from wage contracts to spot wages. To analyze how the degree of competition affects the extent of nominal wage contracts, suppose that the price elasticity of demand rises. On the one hand, this reduces the responsiveness of each firm’s price to a given-size demand shock, implying that such shocks have smaller effects on the marginal revenue product (and labor demand) and on spot wages as product demand becomes more price sensitive. At the same time, the greater demand elasticity implies that labor demand falls more in response to a given gap between contract wages and lower spot wages. As a result, greater goods market competition lowers the net benefit of nominal wage contracts in the face of sectoral demand shocks by reducing the volatility in spot real wages relative to the volatility of employment under nominal wage contracts. On the other hand, in the face of aggregate demand shocks, greater competition implies that wage contracts pose somewhat lower downside costs in terms of real wage and employment volatility. However, greater competition clearly boosts the degree to which labor demand responds to aggregate supply shocks and thereby the extent to which contracts boost the volatility of employment—a key downside cost of contracts. FEDERAL RESERVE BANK OF DALLAS To demonstrate these results, focus on the case in the upper panels of Figure 5, where the economy starts in a zero shock equilibrium (point O) and the price elasticity of demand is low. Suppose an unanticipated sectoral demand shock puts downward pressure on the price of industry j ’s product but has no effect on overall prices. Since aggregate prices and price expectations are unaffected, the labor-supply curve is unaffected. However, the drop in the relative demand for industry j ’s product shifts the demand and marginal revenue curves for industry j left (shift 1). Since profit-maximizing firms produce at levels where marginal cost equals marginal revenue, output falls in the upper left panel from Y Oj (point O) to Y SPj if contracts are not used. In labor markets, market-clearing or spot wages and employment fall from W Oj and L Oj (point O) to W SPj and L SPj (point S ), respectively. However, if wage contracts are used, employment falls further to L Cj (point C ), where the marginal cost of labor equals its marginal revenue product, and output falls further to Y Cj . Thus, nominal wage contracts protect real wages from sectoral demand shocks but at the cost of increased volatility of employment. Now consider what happens under a high price elasticity of demand, where the MRPL (marginal revenue times MPPL) curve is flatter and prices fall less. The MRPL curve in the lower right panel shifts by less than in the upper right panel. Despite the flatter slope of the MRPL curve in the lower panel, it can be shown that wages and employment fall by slightly less than in the upper panel under each wage practice. Nevertheless, when relative demand shocks occur, the flatter MRPL curve implies that less real-wage stability is gained relative to extra employment instability induced by using wage contracts. Now consider an aggregate demand shock (left panels of Figure 6), where the analysis differs in that the labor-supply function shifts out to the extent that a negative aggregate demand shock lowers aggregate prices. For this reason, the labor supply curve shifts right, leaving spot employment unchanged at L Oj , spot nominal wages lower at W SPj , and spot real wages (W/P, not shown) unchanged. Since contracts prevent a decline in wages relative to sectoral demand shocks, aggregate demand shocks induce bigger employment and real-wage deviations from spot market and initial values under wage contracts. However, wage contracts pose less realwage volatility in the face of aggregate demand shocks because such shocks affect prices less under tougher market competition. 7 ECONOMIC REVIEW FOURTH QUARTER 1998 Figure 5 Negative Sectoral Demand Shock and Nominal Wage Contracts Product Market Labor Market Low Price Elasticity of Demand Pj PjO PjC PjSP Wj LjS (P ) 1 MC O C S O C ^ W jO = E(Wj ) WC WjSP S 1 MR O MRPLOj = MRjO × MPPLO j DO D′ 1 MRPLj′ = MRj′ × MPPLjO Yj YjC YjSPYjO LCj LSP j LOj Lj High Price Elasticity of Demand Pj Wj LjS (P ) MC 1 PjO PjC PjSP O C ^ W jO = E(Wj ) WjSP S C O WC S MRPLOj = MRjO × MPPLO j 1 Dj′ 1 MR MR ′ YjC YjSP YjO DjO MRPLj′ = MRj′ × MPPLjO O Yj O LCj LSP j Lj In contrast, the downside costs of wage contracts stemming from aggregate supply shocks are much greater when product markets are more competitive. Consider a negative aggregate supply shock that shifts the MRPL curves of every industry inward by a given distance (right panels of Figure 6, shift 1). In addition, the aggregate supply shock causes an overall rise in prices, which induces an upward shift (shift 2) in the spot labor supply curve equal to the aggregate rise in prices. Spot wages and employment fall from W Oj and L Oj (point O) to W SPj and L SPj (point S ), respectively, while contract employment falls further to L Cj (point C ). Comparing the upper and lower right panels, Lj the extent to which nominal contracts exacerbate employment losses when supply shocks occur is much higher when competition is more intense. Fundamentally, the fiercer the competition, the more falling spot wages cushion the drop in output. Thus, the downside of using contracts when supply shocks occur increases as the price elasticity increases. In summary, on the one hand, greater competition slightly lowers the downside risk of using contracts in the face of aggregate demand shocks. On the other hand, it also cuts the net benefit of contracts when sectoral demand shocks occur and clearly raises the disincentive to contract because of supply shocks. If sectoral 8 Figure 6 Negative Aggregate Shocks and Nominal Wage Contracts Aggregate Demand Shock Aggregate Supply Shock Low Degree of Competition Wj Wj LSj (P O ) LSj (P SP ) LSj (P SP ) 1 LSj (P O ) 1 2 2 O ^ W jO = E(Wj ) ^ W jO = E(Wj ) W jC C C S MRPLOj = MRjO × MPPLjO MRPLOj = MRjO × MPPLjO WjSP W jC O WjSP S MRPLj′ = MRjO × MPPLj′ MRPLj′ = MRj′ × MPPLjO LjC Lj LjO = LjSP LjC LjSP Lj LjO High Degree of Competition Wj LSj (P O ) Wj LSj (P SP ) LSj (P SP ) 2 1 ^ W jO = E(Wj ) WjSP 2 1 O ^ W jO = E(Wj ) W jC C S LSj (P O ) C MRPLj′ = MRj′ × MPPLjO LjC LjO = LjSP MRPLj′ = MRjO × MPPLj′ Lj LjC LSP j LO j Lj Indexation protects real contract wages from aggregate demand variations by keeping contract wages close to their market-clearing level, which moves with inflation. The intuition behind this result is that an expansionary aggregate demand shock will push up inflation and the spot nominal wage. Hence, indexation helps limit deviations of contract wages from marketclearing wages when aggregate demand shocks occur. However, as Gray (1976) shows, indexation moves contract wages further from their full information level when aggregate supply shocks happen. Intuitively, a negative aggregate supply shock boosts inflation and lowers the demand and aggregate supply shocks are large enough relative to aggregate demand shocks, then use of nominal wage contracts will arguably fall as the degree of product market competition rises. Wage Indexation Thus far I have assumed that workers and firms choose only between spot wages or wages that are set ahead of market conditions. In practice some contracts include clauses that adjust wages for inflation or profits, according to a previously arranged formula. For now, consider whether a firm indexes nominal wages for inflation. FEDERAL RESERVE BANK OF DALLAS W jC MRPLOj = MRjO × MPPLjO S WjSP MRPLOj = MRjO × MPPLjO O 9 ECONOMIC REVIEW FOURTH QUARTER 1998 W SPj since prices fall by that magnitude. But when demand is more elastic (lower left panel), the drop in contract employment and the change in the real contract wage are smaller. This occurs because the MRPL curve shifts less in response to aggregate demand shocks when demand is more elastic. Thus, the fiercer the degree of goods market competition, the lower the incentive is to index in the face of aggregate demand shocks because indexing poses smaller benefits. To analyze the downside cost of indexing when aggregate supply shocks occur, consider a common supply disturbance that shifts the MRPL curves downward, shown in the right panels of Figure 7 (shift 1). Under indexation, the higher price level would shift the wage curves in these panels upward (shift 3) so that the nominal wage would rise as much as the overall price level, leaving the real wage unchanged if employment remained at L Oj . However, because labor demand has shifted left (shift 1), the nominal spot wage (W SPj at point S ) is below the level that maintains a constant real wage (W jI at point I ). If contracts are not indexed, the labor-supply curve does not shift and employment falls to L cj (point C ). However, if contract wages were indexed for inflation, the wage-contract curve (right panels) would shift up so that the real wage would be unchanged (point I ), inducing a rise in wages to W jI and a bigger fall in employment to L jI . While aggregate supply shocks cause employment under wage contracts to deviate from its market-clearing level in both cases, the deviation is larger under indexation. In addition, the greater the degree of market competition, the flatter the MRPL curve is (lower left panel) and the more indexation pushes employment and wages from their full information levels (point S ). Since increased employment variation in the face of aggregate supply shocks is the major downside cost of indexing and less employment/real-wage variation in the face of aggregate demand shocks is the major benefit, there is less incentive to index when the degree of product competition is greater. As a result, the prevalence of indexation clauses in nominal wage contracts should fall as the degree of product market competition rises. real marginal product of labor. In such cases, the spot real wage would fall, whereas cost-ofliving escalator clauses would prevent a decline in the real contract wage toward its new spot market level. As a result, the presence of aggregate supply shocks makes incomplete indexation optimal, and a trade-off emerges in which indexation is more attractive the greater the ex ante variance of aggregate demand shocks relative to the ex ante variance of aggregate supply shocks. For plausible variation in aggregate demand and supply shocks, this trade-off implies that a sector exists that is indifferent between indexing and not indexing. Suppose the degree of product market competition rises. Then each firm’s price and marginal product of labor curve become less sensitive to aggregate demand variations. Thus, employment at contract firms is also less sensitive to demand shocks, reducing the incentive to index. Increased competition also means that aggregate supply shocks pose greater downside costs in the form of employment losses from indexing wages. In analyzing the impact of the degree of competition on indexation choices, it is relevant to consider aggregate rather than sectoral shocks because, by definition, only aggregate shocks affect the overall price level. For ease of exposition, comparisons are made between nonindexed wage contracts and contracts fully indexed for inflation. First, consider a negative aggregate demand shock that shifts the MRPL curve leftward (left panels of Figure 7, shift 1). The flatter slope of the marginal revenue curves in the high price elasticity case (lower panel) implies a smaller shift of the MRPL curve. Because indexation shifts the wage-contract curve (W Cj ) down (shift 3) in line with the shifts in the MRPL curves and spot labor-supply curve (shift 2), work hours and real wages are unaffected whether the price elasticity is high or low (depicted by point I ). However, if contract wages are not indexed, these inward shifts imply a decline in employment and a rise in real wages (depicted by point C in both left-hand panels) because the wage contract curve does not move. The extent to which either employment or real wages are affected by not indexing can thus be seen as proportional to the benefit of indexing. Suppose the price elasticity of demand is relatively low, as in the upper left panel. If contract wages are not indexed, then employment falls by the gap between L Oj and L Cj in the upper left panel and real wages rise by the gap between W Cj and Profit Sharing To analyze how increased product market competition affects the incidence of profit sharing in labor contracts, assume that the share of the labor force covered by contracts is constant, as in Duca and VanHoose (forthcoming 10 Figure 7 Negative Aggregate Shocks and Wage Indexation Aggregate Demand Shock Aggregate Supply Shock Low Price Elasticity of Demand Wj 1 LSj (P ′) Wj LSj (P O ) 1 LSj (P ′) 2 LSj (P O ) 2 WjI O ^ E(Wj ) WjI I 3 W jC C 3 C W j,E(Wj ) W jC O S WjSP 3 S,I WjSP = WjI WjI MRPLOj = MRPLj′ = MRj′ × LjC LjO = LjSP MRPLOj = MRjO × MPPLjO × MPPLjO MPPLjO MRjO Lj = LjI LjI LCj LjSP LjO MRPLj′ = MRjO × MPPLj′ Lj High Price Elasticity of Demand Wj LSj (P ′) Wj LSj (P O ) LSj (P ′) 2 1 ^ E(Wj ) ^ W C = E(Wj ) W jC C S,I WjSP = WjI 2 I WjI O 3 LSj (P O ) 1 WjI 3 C WjSP WjI 3 O W jC S MRPLOj = MRjO × MPPLjO MRPLOj = MRjO × MPPLjO MRPLj′ = MRj′ × MPPLjO LjC LjO = LjSP = LjI MRPLj′ = MRjO × MPPLj′ Lj LjI b). As is standard in much of the wage indexation literature (Gray 1976; Karni 1983), assume also that the optimal contract minimizes deviations from the market-clearing wage and employment levels. Under these conditions, the optimal wage contract adjusts the contract wage to mimic the spot wage, which equates labor supply with labor demand. The market-clearing wage reflects that labor demand depends on the marginal product of labor in that industry and labor supply depends on the real wage in terms of the overall price level. A higher degree of product market competition makes labor demand more FEDERAL RESERVE BANK OF DALLAS LjC LSP j LO j Lj elastic, as represented by a larger magnitude of ⑀ in Equation 4. As a result, equilibrium wages and employment at firms with contracts become more sensitive to changes in the MRPL that arise from sectoral demand shifts. For this reason, the relative degree to which spot market wages reflect overall prices, rather than the sectoral marginal product of labor, is higher the greater the sensitivity of labor supply to the real wage and the greater the significance of labor in production. By contrast, profit sharing becomes more important and inflation indexation less appropriate as the extent of goods market competition rises, which makes labor demand more 11 ECONOMIC REVIEW FOURTH QUARTER 1998 this issue. The few related studies done before the late 1990s were largely microeconomic studies that compared cross-sectional patterns of unionization or wage determination with industry patterns of market power. However, these studies did not assess the impact of macroeconomic factors. Inspired by Fischer (1977a, 1977b) and Gray (1976), empirical macroeconomic studies of wages and indexation have focused on traditional macroeconomic factors (such as inflation and aggregate supply shocks) but have ignored changes in market structure or the degree of product market competition (for example, see Ghosal and Loungani 1996). sensitive to the MRPL. Although the marketclearing solution for sectoral wages is complicated, as markets become perfectly competitive (⑀→∞), the equilibrium wage implied by Equations 4 and 5 approaches wj = γ (pj + θ) + (1 – γ)p, (6) where (pj + θ) can be interpreted as the sectorspecific price adjusted for positive supply shocks, γ = 1/[c (1 – α) + 1] < 1 is the weight on the MRPL in determining wages,5 and (1 – γ) is the relative importance of overall prices for spot wages. In practice, one does not observe the explicit indexation or adjustment of wages to the MRPL in a particular industry, mainly because it is difficult to measure and verify this variable. However, as the degree of product market competition rises, the MRPL moves more in tandem with firm profits, a result formally shown by Duca and VanHoose (forthcoming b). Basically, greater competition reduces the extent to which firms boost prices rather than output when the relative demand for that sector’s product rises. As a result, wages and profits more closely reflect the MRPL. For these reasons, greater market competition boosts both the incentive to index wages to the MRPL and the desirability of using profit sharing as a means of doing so. Nominal Wage Contracts The theoretical framework implies, under the assumption of constant money growth, that the use of nominal wage contracts is declining in the variances of aggregate supply and demand shocks and in the degree of product market competition. Consistent with this model, Duca and VanHoose (1998b) find that changes in the unionized share of private-sector workers (a proxy for nominal wage contracts) are negatively related to the variance of real oil prices (a proxy for the aggregate supply shock variance), the inflation rate (a proxy for aggregate demand shock variance), and the inverse of an adjusted profit-share measure (a proxy for the degree of goods market competition).6 Changes in the cross-sectional pattern of unionization also support the view that greater market competition reduces the use of nominal wage contracts. Some sectors are more suited to such contracts than others due to sector-specific conditions, which, at a point in time, can account for differences in unionization across industries. However, changes in how much competition a sector faces relative to others may explain why unionization rates have declined faster in some sectors. Indeed, data available since the early 1980s show that the largest declines in unionization have been in sectors facing increased foreign competition or in deregulated sectors (Figure 8 ).7 Table 1 shows that many U.S. industries have been deregulated since the late 1970s, suggesting that competition has become tougher in product markets, not only for traded goods but also for nontraded services. Note that most of the overall drop in unionization since the early 1980s is the result of falling unionization rates within sectors, rather than shifts in jobs from more unionized industries to less unionized ones.8 The results are also consistent with a “rent-sharing” theory of unions (Layard, Nickell, EMPIRICAL EVIDENCE ON PRODUCT MARKET COMPETITION AND LABOR PRACTICES Although increased product market competition can theoretically affect labor practices, there has been little empirical macro work on Figure 8 Unionization Rates by Industry Groups Percent of private payrolls 35 30 25 Manufacturing 20 15 Deregulated industries 10 Other private industries 5 ’83 ’84 ’85 ’86 ’87 ’88 ’89 ’90 ’91 ’92 ’93 ’94 ’95 ’96 SOURCE: Bureau of Labor Statistics. 12 Table 1 Many U.S. Industries Have Been Deregulated Since the 1970s and Jackman 1994; Oswald 1982) and with microeconomic studies that test for links between firm monopoly power and unionization (for citations, see Mason and Bain 1993). According to the rent-sharing approach, imperfect competition or pricing power creates excess profits for a firm, which gives workers an incentive to incur the costs of unionizing. By forming a union, current workers can restrict the potential entry of other workers and thereby gain enough negotiating leverage to induce the firm to share excess profits by paying abovemarket wages. However, if the market structure changes so that new firms can more easily enter, then excess profits are bid down and there are fewer rents to share. From this perspective, deregulation and foreign competition have greatly reduced the economic benefits of unionizing. The shortcoming of rent-sharing models relative to the framework sketched here is that existing rent-sharing models tend to ignore that American unions generally negotiate only partially inflation-indexed wage contracts whose optimality is also affected by supply shocks. The advantage of the rent-sharing approach is that its assumption of restricted labor supply within a given sector can account for the tendency of union wages to be higher than nonunion wages. Major Deregulatory Steps Agriculture — Mining Oil and gas: oil prices deregulated by a series of presidential executive orders beginning in 1976; natural gas prices deregulated by the Natural Gas Policy Act of 1978. Construction — Manufacturing Increased openness to trade, partly from the General Agreement on Trade and Tariffs (1979, 1993), the Canada– U.S. Free Trade Agreement (1989), and the North American Free Trade Agreement (1994). Transportation Trucking: truck rates liberalized in the late 1970s and deregulated by the Motor Carrier Act (1980). Airlines: the Airline Deregulation Act (1978) allowed entry in 1982 and deregulated airfares in 1983. Railroads: deregulated by Interstate Commerce Commission liberalization of rail rates in the late 1970s and the Staggers Rail Act (1981). Communications Telephones: largely deregulated following the AT&T court settlement of 1982. Cable television: deregulated in a series of Federal Communications Commission rulings in the late 1970s and by the Cable Television Deregulation Act (1984). Telecommunications: partly deregulated by the Telecommunications Act (1996). Wage Indexation The theoretical model implies that the use of indexation in wage contracts should be increasing in the variance of aggregate demand shocks, decreasing in the variance of aggregate supply shocks, and decreasing in the degree of product market competition. Consistent with these predictions, Duca and VanHoose (forthcoming b) find that the overall incidence of indexation clauses in union contracts is negatively related to the variance of real oil prices, positively correlated with the inflation rate and squared inflation expectation errors of households, and negatively related to the inverse of an adjusted profit-share measure. Indeed, Duca and VanHoose (forthcoming b) find that changes in the overall use of indexation in contracts are better tracked by an empirical model that adds the degree of product market competition to a more conventional empirical model having only measures of aggregate supply and demand shock variances. As with unionization rates, the cross-sectional pattern of declines in indexation are most pronounced in industries that either face foreign competition or have been deregulated since the late 1970s. FEDERAL RESERVE BANK OF DALLAS SIC Industry* Wholesale — Retail — FIRE (finance, insurance, and real estate) Banking: partly deregulated by the Depository Institution Deregulation and Monetary Control Act (1980) and the Garn – St. Germain Depository Institutions Act (1982). * Standard industry classification (SIC) of sectors at the one-digit-level classification code. SOURCES: Winston (1993); author’s compilations. Profit Sharing Profit sharing has risen dramatically since the early 1980s, as seen in Figure 3. As shown by Bell and Kruse (1995), most of these profitsharing provisions include employee stock ownership plans or profit-based contributions to thrift plans but make relatively little use of nondeferred forms of profit sharing, such as cash bonuses. Deferred profit sharing is more common because most workers do not have sufficient wealth to see their weekly take-home pay vary with market conditions. They are, however, better able to handle profit volatility over the long run, such as in the form of variable, but cumulative, contributions to their retirement accounts. Nevertheless, recent salary and Federal Reserve Beige Book surveys indicate that annual base/hourly pay is increasingly being comple- 13 ECONOMIC REVIEW FOURTH QUARTER 1998 more in some industries than in others. Indeed the biggest increases in profit sharing through 1993 have occurred in sectors facing increased foreign competition, such as manufacturing, or in deregulated sectors, such as transportation (Figure 9 ). While inconclusive, these trends are loosely consistent with the view that more intense product market competition is boosting the use of profit sharing. Figure 9 Profit Sharing Rises in Manufacturing and Deregulated Industries Percent of workers with pension plans 45 40 Manufacturing 35 30 25 Deregulated industries 20 CONCLUSION 15 Fiercer product market competition can theoretically reduce the prevalence of nominal wage contracts and of indexation in such contracts while boosting the use of profit sharing. Arguably, product markets have generally become more competitive in the United States since the late 1970s, owing to increased foreign competition in traded goods markets and the deregulation of many nontraded sectors. Consistent with this view, the after-tax profit share of nonfinancial corporations has moved within a lower range since the late 1970s, after adjusting for swings in temporary factors and net interest (Duca 1997). Aggregate time-series evidence supports the view that increased competition has reduced the use of nominal contracts and indexation, as reflected in the declining rate of unionization and the falling incidence of CPI indexation clauses in union contracts. Limited, inconclusive data also support the view that greater product market competition has boosted the overall use of profit sharing. Consistent with aggregate movements in labor practices and a measure of the degree of goods market competition, industry-level data indicate that all three of these trends are most evident in sectors that have experienced either deregulation or increased foreign competition since the late 1970s. While more research needs to be done, particularly using industry-level data, new theoretical arguments and empirical evidence support—but do not conclusively prove—the view that increased product market competition has been reshaping America’s labor markets. Other private industries 10 5 0 ’80 ’81 ’82 ’83 ’84 ’85 ’86 ’87 ’88 ’89 ’90 ’91 ’92 ’93 SOURCES: Bell and Kruse (1995); author’s calculations. mented by variable cash bonuses. This shift suggests that pay may be becoming more market responsive in both the short run and long run. The theoretical framework presented in this article implies that under fiercer competition, profits are more closely aligned with workers’ market value, with profits more closely reflecting prices minus unit labor costs (wage costs adjusted for productivity). As a result, profit sharing should trend upward with a measure of market competition, as seen in Figure 3, where the competition measure rises as the relative demand for goods becomes more price elastic. In this figure, the price elasticity of demand is measured using the inverted after-tax profit share of nonfinancial corporations, adjusted for swings related to the business cycle, oil prices, and exchange rates (Duca and VanHoose, forthcoming a, forthcoming b). Nevertheless, since available data cover a short period and have a missing data point (1987), the evidence is supportive, not conclusive. Thus, it is unclear whether greater competition rather than other factors has induced a rise in profit sharing. One way around this inference problem is to compare deregulated or traded goods industries with others. Some sectors are more suited to profit sharing than others because the nature of work and the ability to monitor work vary across sectors. Such factors would account for why differences exist across sectors at a point in time, while changes in the attitudes of different generations might account for why profit sharing has risen overall. However, changes in the relative degree of competition across sectors might account for why profit sharing has risen NOTES I would like to thank John Benedetto for providing research assistance; several human resource professionals, in particular Denise Duca and Joel Koonce, for helpful discussions about changing norms in labor compensation; and Evan Koenig, Stephen Prowse, and David VanHoose for helpful comments. Any errors are my own. 14 1 2 3 4 5 6 7 8 1990s?” Federal Reserve Bank of Dallas Economic Review, Fourth Quarter, 1–14. This is true if core inflation is adjusted for methodological changes in recent years that have reduced the extent to which statistics have overstated consumer price inflation. This may have occurred in the 1960s when a run-up in productivity growth initially offset rising wages stemming from low unemployment, but eventually accelerating wage growth overtook productivity growth, causing unit labor costs and inflation to rise. Accompanying this change has been an increased use of temporary workers, a phenomenon examined by Segal and Sullivan (1995, 1997). Nevertheless, the incidence of sectoral demand shocks may reduce contracting in extreme cases in which workers place much more emphasis on minimizing deviations from market-clearing employment than on minimizing deviations from the expected real wage. In logs, the MRPL equals marginal revenue (pj ) plus the marginal physical product of labor (the sum of the two log-linear supply shocks). Prior studies (Evans 1991; Holland 1986) have found inflation uncertainty to be increasing in the level of inflation, which is consistent with arguments for pursuing price stability. For details on deregulation, see Duca and VanHoose (forthcoming a) and Winston (1993). Following Duca and VanHoose (1998b), the time series splices data from Troy and Sheflin (1985) with data from the Bureau of Labor Statistics. Similar trends in unionization rates are evident in recent estimates by Freeman (1998). ——— (1998), “The New Labor Paradigm: More MarketResponsive Rules of Work and Pay,” Federal Reserve Bank of Dallas Southwest Economy, Issue 3, May/June, 6 – 8. Duca, John V., and David D. VanHoose (1991), “Optimal Wage Indexation in a Multisector Economy,” International Economic Review 32 (November): 859 – 67. ——— (1998a), “Has Greater Competition Restrained U.S. Inflation?” (Unpublished manuscript). ——— (1998b), “The Rise of Goods Market Competition and the Fall of Nominal Wage Contracting: Endogenous Wage Contracting in a Multi-Sector Economy” (Unpublished manuscript). ——— (forthcoming a), “Goods Market Competition and Profit-Sharing: A Multisector Macro Approach,” Journal of Economics and Business. ——— (forthcoming b), “The Rise of Goods Market Competition and the Decline in Wage Indexation,” Journal of Macroeconomics. Evans, Martin (1991), “Discovering the Link Between Inflation Rates and Inflation Uncertainty,” Journal of Money, Credit, and Banking 23 (May): 169 – 84. Fischer, Stanley (1977a), “Long-Term Contracting, Sticky Prices, and Monetary Policy: A Comment,” Journal of Monetary Economics 3 (July): 317– 23. REFERENCES Ball, Laurence (1988), “Is Equilibrium Indexation Efficient?” Quarterly Journal of Economics 103 (May): 299 – 311. ——— (1977b), “Long-Term Contracts, Rational Expectations, and the Optimal Money Supply Rule,” Journal of Political Economy 85 (February):191– 205. Bell, Linda, and Douglas Kruse (1995), “Evaluating ESOPs, Profit Sharing, and Gain Sharing Plans in U.S. Industries: Effects on Worker and Company Performance” (Manuscript, U.S. Department of Labor, March). Freeman, Richard B. (1998), “Spurts in Union Growth: Defining Moments and Social Processes,” in The Defining Moment: The Great Depression and the American Economy in the Twentieth Century, eds. Michael D. Bordo, Claudia Goldin, and Eugene N. White (Chicago: University of Chicago Press), 265 – 95. Blanchard, Olivier, and Lawrence F. Katz (1997), “What We Know and Do Not Know About the Natural Rate of Unemployment,” Journal of Economic Perspectives 11 (Winter): 51–72. Friedman, Milton (1968), “The Role of Monetary Policy,” American Economic Review 58 (March): 1–17. Blinder, Alan, and N. Gregory Mankiw (1984), “Aggregation and Stabilization in a Multi-Contract Economy,” Journal of Monetary Economics 13 (January): 67– 86. Galbraith, James K. (1997), “Time to Ditch the NAIRU,” Journal of Economic Perspectives 11 (Winter): 93 –108. Duca, John V. (1987), “The Spillover Effects of Wage Rigidity in a Multi-Sector Economy,” Journal of Money, Credit, and Banking 19 (February): 117– 21. Ghosal, Vivek, and Prakash Loungani (1996), “Evidence on Nominal Wage Rigidity from a Panel of U.S. Manufacturing Industries,” Journal of Money, Credit, and Banking 28 (November): 650 – 58. ——— (1997), “Has Long-Run Profitability Risen in the FEDERAL RESERVE BANK OF DALLAS 15 ECONOMIC REVIEW FOURTH QUARTER 1998 Gordon, Robert J. (1997), “The Time-Varying NAIRU and Its Implications for Economic Policy,” Journal of Economic Perspectives 11 (Winter): 11– 32. Oswald, Andrew J. (1982), “The Microeconomic Theory of the Trade Union,” Economic Journal 92 (September): 576 – 95. Gray, Jo Anna (1976), “Wage Indexation: A Macroeconomic Approach,” Journal of Monetary Economics 2 (April): 221– 35. Phelps, Edmund S. (1967), “Phillips Curves, Expectations of Inflation, and Optimal Unemployment Over Time,” Economica 34 (August): 254 – 81. Holland, A. Steven (1986), “Wage Indexation and the Effect of Inflation Uncertainty on Employment: An Empirical Analysis,” American Economic Review 76 (March): 235 – 44. Segal, Lewis M., and Daniel G. Sullivan (1995), “The Temporary Labor Force,” Federal Reserve Bank of Chicago Economic Perspectives 19 (March/April): 2 –19. ——— (1997), “The Growth of Temporary Services Work,” Journal of Economic Perspectives 11 (Spring): 117– 36. ——— (1995), “Inflation and Wage Indexation in the Postwar United States,” Review of Economics and Statistics 77 (February): 172 – 77. Staiger, Douglas, James H. Stock, and Mark W. Watson (1997), “The NAIRU, Unemployment, and Monetary Policy,” Journal of Economic Perspectives 11 (Winter): 33 – 51. Karni, Edi (1983), “On Optimal Wage Indexation,” Journal of Political Economy 91 (April): 282 – 92. Layard, Richard, Stephen Nickell, and Richard Jackman (1994), Unemployment: Macroeconomic Performance and the Labour Market (New York: Oxford University Press). Troy, Leo, and Neil Sheflin (1985), Union Sourcebook (New Brunswick, N.J.: Industrial Relations Data and Information Services). Mason, Bob, and Peter Bain (1993), “The Determinants of Trade Union Membership in Britain: A Survey of the Literature,” Industrial and Labor Relations Review 47 (January): 332 – 51. Winston, Clifford (1993), “Economic Deregulation: Days of Reckoning for Microeconomists,” Journal of Economic Literature 31 (September): 1263 – 89. 16 Chinese President Jiang Zemin announced in late 1997 that his nation would soon privatize thousands of state-owned enterprises. The precise meaning of privatization in the Chinese context is open to debate: one deputy minister explained that the state might retain a majority stake in the firms, and the editor of a communist newspaper said that expectations of a Western-style ownership structure in the firms had arisen from a “misunderstanding” (Lyle 1997). Still, while information is scarce about the details of the Chinese plan, it is clear that any privatization effort in the world’s most populous country could have a major impact on the global economy. Given the distinct possibility of massive privatization in China, it is natural to reexamine economic reform in the postcommunist nations of Eastern Europe and determine what lessons may be drawn for China. The reform process in these nations has been substantial: the privatesector contribution to gross domestic product (GDP) now exceeds 50 percent in nineteen of the twenty-six states that once formed the Soviet empire (European Bank for Reconstruction and Development 1997), including each of the Eastern European states outside the volatile Balkan region (Table 1 ).1 The short-term pain caused by economic reform has generally given way to significant gains in per capita GDP (Figure 1 ) and the possibility of membership in multilateral institutions such as the European Union and NATO.2 But can any lessons from the Privatization and the Transition to a Market Economy Jason L. Saving Economist Federal Reserve Bank of Dallas A n effective privatization must transfer ownership in such a way that the new private-sector owners and managers have an incentive to maximize profit, and this incentive must be reinforced by a legal structure that encourages private-sector competition for these firms. Table 1 Share of GDP Derived from Private Sources Czech Republic Hungary Poland Romania Russia Slovakia United States 1980 <1.0 3.5 15.6* 4.5 <1.0 <1.0 79.4 1988 <1.0 7.1 18.8* — <1.0 <1.0 79.6 1994 65 55 55 35 50 55 81.1 1997 75 75 65 60 70 75 82.0 * This is almost exclusively agricultural production (Slay 1993). NOTE: Czechoslovakia dissolved in 1993 and was replaced by Slovakia and the Czech Republic. SOURCES: Patterson 1993; European Bank for Reconstruction and Development 1994, 1997; Bureau of Economic Analysis. FEDERAL RESERVE BANK OF DALLAS 17 ECONOMIC REVIEW FOURTH QUARTER 1998 support” (Bird 1998), and it is estimated that one-sixth of Albania’s population may become bankrupt because of investments in pyramid schemes (Percival 1997). Such examples suggest that economic education may be important to the success of privatization. In addition, formerly communist citizenries face special problems in attempting to evaluate the net worth of to-be-privatized firms. Balance sheets for such firms were not kept according to Western norms, often deliberately, so as to shield the poor performance of highranking party functionaries who were appointed as managers during the communist era but wish to retain their positions after privatization (Brada 1996, Tirole 1994). Moreover, the division of assets among such companies is not clearly delineated, so that a potential investor during the transition period could not be certain which assets would end up with a company and which would remain with the state (Bolton and Roland 1992). In addition, the transition to a market economy would almost certainly cause an enormous reallocation of resources across sectors of the economy, making it difficult for citizens to gauge a firm’s future performance even if communist-era balance sheets were available. A final problem relates to communist-era informational asymmetries between the ordinary citizen and the party elite. Information is a closely guarded commodity under communist systems, and government officials commonly possess an enormous amount of information of which ordinary citizens are unaware (Blanchard and Layard 1992). In the aftermath of communism, then, a citizen’s information about staterun enterprises would be primarily a function of past affiliations with a now-discredited state. This implies that the group of citizens best able to evaluate the profit potential of state-run firms is composed of precisely those individuals society would least like to prosper. Yet there is widespread agreement that shares in state-owned enterprises should be given to the public whose efforts created the enterprises. These “mass privatization” programs are regarded as the most egalitarian method of privatization because each citizen profits from them (Grime and Duke 1993). Also, giving each citizen a stake in privatization increases public support for the program and reduces the government’s ability to renege on its privatization commitments (see the box entitled “Credible Commitments and Privatization”). Is there a socially optimal distribution of ownership shares when ordinary citizens are Figure 1 Real GDP Per Capita U.S. dollars 6,000 5,000 Hungary Czech Republic Slovakia Poland Romania 4,000 3,000 2,000 1,000 0 1990 1991 1992 1993 1994 1995 1996 experiences of these Eastern European nations be applied to other countries embarking on the road of economic reform? Also, was the road to privatization in these nations paved with serious mistakes others might be able to avoid? In discussing these questions, it is important to define precisely what is meant by privatization. While privatization occurs whenever the ownership of a firm is transferred from the government to the private sector, the purpose of this transfer is to help create a competitive economic environment in which firms strive to increase efficiency and maximize profit. Thus, an effective privatization must transfer ownership in such a way that the new private-sector owners and managers have an incentive to maximize profit, and this incentive must be reinforced by a legal structure that encourages private-sector competition for these firms. These are the issues this article discusses. INFORMATION AND PRIVATIZATION IN PRACTICE In theory, privatization is a relatively simple process in which potential investors evaluate the profit potential of each firm, choose those in which to acquire shares, and then manage the firms in a way that maximizes their return.3 In practice, however, this model suffers from limitations when applied to the transition from communism to a market economy. In general, formerly communist citizenries are likely to be less familiar with market economics than their capitalist counterparts, and some commentators have suggested that this unfamiliarity could lead them to make bad investment decisions (Brada 1992). This perception is not without foundation. For example, a 1998 Moldovan newspaper headline marveled that the “private sector survives even though it gets no state 18 Credible Commitments and Privatization Privatization helps promote economic growth by ensuring that firm owners have an incentive to maximize profit. However, one important aspect of any privatization program is the possibility of revocation (Weingast 1995). Unless those who become owners of privatized enterprises believe the government will allow them to keep the fruits of their labors, there is little reason to believe these owners will invest time and effort in their enterprises. The importance of political commitment to the privatization process can be illustrated through an examination of Russia’s New Economic Policy.* Shortly after the communist takeover of 1917, Bolshevik leaders confiscated private farmland in what they called the “crusade for bread.” Though the peasantry was ordered to continue working the fields, all farm output would go to the state rather than the workers. This policy reduced peasant effort to such an extent that mass starvation ensued, followed by peasant riots that began to threaten the survival of the Bolshevik regime. In response, Lenin introduced a partial privatization of farmland and a partial restoration of farmers’ right to sell excess produce to ordinary Russians, arguing that peasants would not work at maximum efficiency unless they were able to reap the rewards of their labors. But would the communist regime — which had previously viewed private property as anathema — be willing to commit to the New Economic Policy for the foreseeable future? Nove (1992) recounts the words of a communist official who complained New Economic Policy supporters “demand of us a promise that we will never, i.e. not in 15 or 20 years, confiscate or expropriate” farmland, to encourage peasant farmers to work without fear that the state would seize their crops and confiscate their land. This the official refused to do: in his words, expropriation of farmland by the state would occur “when the time comes for so doing.” The productivity gains from the New Economic Policy were short-lived, partly because of these fears. Farmland was then reconfiscated in a new effort to abolish private farming. unable to evaluate the prospective financial performance of state-run enterprises but communist functionaries are capable of making such evaluations? Perhaps surprisingly, there are some conditions under which this question can be answered in the affirmative. For example, suppose that shares are to be divided among the citizenry subject to three conditions: 1. Every share must be distributed to a citizen. 2. The expected profit of each citizen must be equal. 3. The risk (variance) borne by each citizen must be equal and must be as low as possible, subject to all other conditions. Only one portfolio of shares meets these three criteria: a distribution in which each citizen receives an equal percentage of each enterprise.4 Moreover, such a distribution can be implemented even if those in charge of privatization cannot evaluate the financial viability of state-owned firms—an important consideration in countries where government economic decision-making prowess is at best questionable. There are no examples of large-scale privatizations in which each citizen was given an equal share in each state-owned enterprise. However, the equal-share distribution is a special case of the voucher-based privatization programs implemented by, among others, Poland and Czechoslovakia (which split into Slovakia and the Czech Republic on January 1, 1993). In these mass privatization programs, each citizen received an equal amount of purchasing power, which could be used to obtain shares of state enterprises. However, these and other mass privatization programs must confront certain problems. * Details of the New Economic Policy are taken from Nove (1992). might perform no better after privatization than they had before privatization.5 Such a situation points to the difference between the act of privatization (transferring ownership to the private sector) and an effective privatization. The oversight issue is especially problematic because of the temptation to retain communist-era managers during the transition period and let investors decide later whether to remove them. In Russia and the Ukraine, such holdover managers routinely used bribes and intimidation to forestall their replacements (Roland 1994), a maneuver that has been dubbed “grab-ization” by the media (Seely 1993). In Hungary, managers often held positions in local government, and the patronage opportunities afforded by such an arrangement encouraged local governments to resist both managerial replacement and the overall privatization effort (Bolton and Roland 1992). Finally, some managers across Eastern Europe simply seized firms from the state and began operating them as private businesses (Voszka 1993). This “spontaneous privatization” of firms was aided by a lack of information, even on the part of government, about which assets belonged to the state (Boycko, Shleifer, and Vishny 1994). When owners were able to replace management despite these problems, empirical evidence suggests such replacement helped HOLDING COMPANIES AND EFFECTIVE PRIVATIZATION Problems of Dispersed Ownership Mass privatization carries with it a serious problem: no single investor has the ability to oversee management. Since managers do not own their firms, shareholders must oversee managers to ensure that management makes every effort to maximize profit ( Jensen and Meckling 1976). However, mass privatization can produce a situation in which an enormous number of individuals hold only a small number of shares. Under this scenario, no single investor would have sufficient incentive to monitor managerial activity, even though every investor would gain by such monitoring (Gray 1996). Without a resolution to this problem, firms FEDERAL RESERVE BANK OF DALLAS 19 ECONOMIC REVIEW FOURTH QUARTER 1998 era banks had no incentive to investigate the financial viability of firms to which they made loans or to hire people who possessed any particular aptitude for such investigations. Moreover, since it is almost impossible for bank privatization to precede the privatization of other enterprises, distributing firms across the banking sector would amount to continued nationalization rather than privatization. With neither the national government nor the banking sector able to fulfill the responsibilities of the powerful shareholder during the transition from communism, privately run holding companies would become the most natural alternative. Econometric evidence from Britain suggests that holding companies can be an effective means of enforcing managerial productivity after privatization (Blanchard and Layard 1992). However, holding companies alone are not sufficient to ensure effective privatization, as recent experiences in Chile and Russia have demonstrated. Chile attempted large-scale privatization shortly after the ascension of Pinochet, and the result was an enormous concentration of ownership shares among private financial institutions that were subsequently rendered bankrupt by an economic downturn (Bolton and Roland 1992). In Russia, voucher privatization was hampered by the presence of bogus investment funds that absorbed citizen investments and acquired firms in the name of those citizens, but then vanished when payment was requested by the state (Seely 1993). Both Chile and Russia were forced to embark upon costly reprivatization programs. These examples suggest that the proper implementation of holding companies is of crucial importance. How did the design of the Polish and Czechoslovak privatizations compare in this respect? Recognizing the pitfalls associated with mass privatization, the Polish proposal mandated the creation of between ten and twenty holding companies (Blanchard and Layard 1992). Responding to concerns about a lack of information about the financial performance of state-owned firms, each holding company was to be assigned a distribution of ownership shares by the government (Boycko, Shleifer, and Vishny 1994). Finally, to ensure managerial oversight, at least one-third of the shares of each state-owned firm would be given to a single holding company that would be expected to oversee the firm’s management. Each citizen would receive one share in each holding company but none in individual firms. The Czechoslovak proposal, in contrast, did not mandate the creation of holding com- increase the economic performance of privatized firms in Eastern Europe (Dyck 1997). But in a world of dispersed ownership, such replacement may be unlikely to occur. Holding Companies as a Solution To address these concerns, many economists have recommended the creation of a small number of powerful shareholders (Bolton and Roland 1992). Because firm performance would substantially impact the profit of these shareholders, they would (at least theoretically) have an incentive to control management (Gray 1996). At the same time, citizens could invest in the diverse portfolios of these powerful shareholders, rather than in a single company (Boycko, Shleifer, and Vishny 1994). Further, the powerful shareholders would presumably have a much greater ability to obtain financial information about the firms to be privatized and, therefore, a greater ability to pick a reasonable ownership portfolio. Perhaps the most obvious candidate for the role of “powerful shareholder” would be government. As the single most powerful entity in a country, government would have the necessary size and scope to exert pressure on management. Moreover, because governments in formerly communist countries are far more intrusive into the lives of citizens than are governments in the West, government officials in these nations would be relatively adept at obtaining information and pressuring firms. However, governments have a multiplicity of interests, which include running an efficient firm but also include minimizing unemployment and facilitating political patronage. Presumably, there would be no reason to discuss privatization unless governments were failing to operate efficient enterprises, but a government that fails in this capacity would not be an appropriate choice for powerful shareholder. The other obvious candidate for fulfilling the responsibilities of a powerful shareholder would be the indigenous banking sector.6 After all, these banks had made loans to newly privatized enterprises before those enterprises were private. This suggests that the banking sector would have knowledge of the firms’ financial status and future prospects (to gauge the ability to repay a loan) as well as an especially strong interest in the firms’ financial future (to ensure such repayment). However, lending decisions under communism were based on political rather than economic considerations, and it was known that loans not repaid by borrowers would be covered by the state. Thus, communist- 20 panies. Instead, each citizen was to receive a certain number of voucher points, which could be used to obtain shares in individual firms. These points could also be entrusted to holding companies if such companies were to emerge from the private sector, but the government would not itself create them (Brada 1992). Voucher holders would then expend points to acquire shares in particular enterprises. The Polish plan might appear to be the more effective means of privatization because it mandated a reasonable ownership structure and minimized the informational problems associated with citizen participation. In reality, however, the Polish plan was stymied for years because political pressures prevented the government from choosing how to allocate ownership shares to holding companies (Roland 1994). In Czechoslovakia, on the other hand, privately run investment funds quickly arose in response to an enormous demand for such funds by consumers. Indeed, more than half of all Czechoslovak citizens entrusted their vouchers to these funds, whose portfolios were determined through market mechanisms rather than the political process (Kotrba and Svejnar 1994, Grime and Duke 1993). Ironically, Czechoslovakia emerged with an ownership structure reasonably close to the structure desired by Poland without any government mandates to that effect, while Poland’s attempt to mandate such a structure almost derailed its privatization program entirely. The examples given here permit certain conclusions to be drawn regarding holding companies. First, citizens recognize the problems associated with mass privatization and, as the Czechoslovak experience demonstrates, will solve them through holding companies if given the opportunity to do so. Second, government can assist in this process by guarding against fraudulent holding companies, something the Russian state failed to do in its initial privatization effort. However, efforts to move beyond this oversight could harm a privatization program even if such efforts were well-intentioned, such as the Polish government’s desire to ensure that holding companies would be created and would receive the proper distribution of shares. When states seek to create holding companies, the Polish experience suggests that policymakers must have a firm consensus on precisely how shares will be distributed—perhaps by implementing an equal-share distribution at the holding company level and then giving each citizen an equal share of each holding company. FEDERAL RESERVE BANK OF DALLAS LEGAL STRUCTURE AND PRIVATE BUSINESS CREATION A successful privatization program transfers ownership from government to the private sector in a way that ensures proper managerial oversight of the privatized firms. However, these firms will be tempted to behave as monopolies unless they face competition from other private-sector firms. Such competition cannot form as long as communist-era constraints on private business creation are maintained. These constraints range from a ban on the private ownership of land to the harassment of private-sector banks to confiscatory tax regimes. They even include price controls, which are not typically associated with private business creation but nevertheless play an important role in facilitating competition. Direct Constraints There are several reasons to suppose that constraints on private business creation would diminish the economic effects of privatization. One reason is that constraints on private business creation lead to monopolistic industries (Feinberg and Meurs 1994). Although privatization programs move government-run monopolies into the private sector, they do not in and of themselves create competitive markets. Unless entrepreneurs are free to create competing firms, newly privatized companies will not face sufficient financial pressure to improve quality and reduce prices. In such a noncompetitive environment, privatization could change firm ownership without changing firm behavior. Constraints on private business creation are also likely to harm consumers who wish to obtain goods deemed unimportant during communist rule. Communism emphasizes heavy industry over agriculture and the service sector, which suggests that the public sector would contain a disproportionately high number of heavy-industry firms but a disproportionately low number of service-sector and agricultural firms (Bolton and Roland 1992). In fact, the employment share of heavy industry relative to these sectors was estimated to be 50 percent higher in Eastern Europe than in comparable developed economies (European Bank for Reconstruction and Development 1997). Thus, private business creation is needed to shift the overall composition of firms in the economy toward services and agriculture and thereby satisfy consumer demand. Another effect of constraints on private business creation is to exacerbate unemploy- 21 ECONOMIC REVIEW FOURTH QUARTER 1998 Table 2 Qualitative Competitiveness Measures, 1997 Czech Republic Hungary Poland Romania Russia Slovakia Privatization Removal of entry barriers Price liberalization Near OECD Near OECD Near OECD Substantial Substantial Near OECD Substantial Substantial Substantial Some Some Substantial Substantial Substantial Substantial Substantial Substantial Substantial Solidarity government in Poland and a high priority in most other Eastern European states (Vickers and Yarrow 1991). While price controls are most commonly criticized for obscuring producers’ ability to know which products to produce, price controls also act as a barrier to private business creation and therefore affect whether an entrepreneur produces at all. When the state sets artificially low prices in a particular industry, entrepreneurs have little incentive to set up firms in that industry because they cannot earn a sufficiently high rate of return on their investment. Thus, raising prices can actually help consumers by hastening the arrival of competitors, which will in turn eliminate the shortages endemic to communism and ultimately lower prices as well. NOTE: Near OECD means near normal Western levels. SOURCE: European Bank for Reconstruction and Development 1997. ment problems that occur during the privatization process. Because communist doctrine emphasizes full employment, state-owned firms are encouraged to employ workers even when those workers become unprofitable for their companies (Feinberg and Meurs 1994). This suggests that private owners would almost certainly need to lay off workers. Unless these individuals are free to start their own businesses and free to work for newly created businesses, the unemployment created by privatization could cause rather severe social difficulties, especially given the near-total absence of a social safety net in these countries. Finally, constraints on private business creation place an inordinate emphasis on successful privatization. Even though the theoretical issues involved in privatization are relatively simple, designing a privatization program that successfully translates theory into effective privatization is extremely difficult. By definition, private business creation lessens the economic influence of newly privatized firms. Therefore, business creation acts as a sort of safety net for poorly designed privatization programs and thereby assists in the transition to a market economy. Eastern European Experience The privatization process is nearing completion across the core states in Eastern Europe, and constraints on private business creation in these states have been substantially reduced (Table 2 ). However, private business creation was hindered in a variety of ways during the transition period. In Russia, market entry was discouraged “by a variety of technological, administrative, and other obstacles” (Capelik 1992). In Hungary, market entry was hampered by a government policy under which large state banks ignored the private sector in favor of state-owned enterprises; additionally, any loans by private-sector banks to private-sector businesses could be deemed improper banking practices punishable by closure of the bank in question (Pataki 1993). And in Romania, government regulations impede market entry to such an extent that many privatized firms continue to be run as monopolies (Melloan 1998). On the other hand, private-sector economic activity exploded in the Czech Republic as a result of rapid postcommunist market liberalization (European Bank for Reconstruction and Development 1997). For example, the Czech Republic’s private-sector share of GDP is almost on a par with that of the United States (Table 1 ). Private-sector economic activity also rose in Poland and Hungary, where successful communist-era experiments in private-sector legalization helped persuade postcommunist policymakers to adopt market liberalization shortly after the fall of the Soviet Union (Johnson 1994). These three countries have had the fastest-growing economies in Eastern Europe since the transition from communism (Figure 1). While one cannot demonstrate conclusively that a favorable climate for private Price Controls Before the fall of the Iron Curtain, goods communist officials deemed necessary were often priced significantly lower than comparable goods in the West, while goods deemed unnecessary were often priced significantly higher (Arrow and Phelps 1993). It is, therefore, not surprising that price liberalization was one of the first issues with which the postcommunist states of Eastern Europe had to contend. Economists have argued that price liberalization is essential to the development of a market economy because prices are the mechanism through which entrepreneurs determine what to produce. Based on this advice, price reform was made the first priority of the postcommunist 22 as China’s? One such lesson is that the design of a privatization program is as important as the economic theories the program incorporates. Unlike privatization in the West, which typically involves a single firm whose financial performance has been evaluated by independent auditors, the transition from communism involves the simultaneous privatization of hundreds of firms for which even the most basic balance-sheet information is likely to be unavailable. Additionally, a dispersal of ownership across many citizens could produce a situation in which shareholders would be unable to exercise effective oversight of management. One way to overcome this problem would be through the creation of holding companies in which each citizen could invest. For example, heeding the suggestion of economists, both Poland and Czechoslovakia made holding companies a centerpiece of their privatization programs. However, equally strong commitments to the principle of holding companies did not result in equal success with their implementation. Holding companies played an enormous role in the success of the Czechoslovak program but nearly destroyed the Polish scheme. Such seemingly small changes in implementation could have enormous consequences for Chinese privatization. Of particular importance is the claim that Chinese privatization might not move a majority interest in state-owned firms into the private sector. If the Chinese government were to maintain a majority stake in a particular “privatized” enterprise, it could thwart efficiency-enhancing measures at that company if the measures were likely to increase unemployment. The government could also block the firm from entering new markets if those markets were already served by a government-owned firm. It could even bankrupt the firm if the private-sector shareholders were to resist the directions in which the government wished to take the firm. One might hope the government would behave as if it were a private-sector individual, but the only way to guarantee such behavior by the owners of privatized firms would be to ensure that every owner is a private-sector individual— which would be inconsistent with a privatization plan in which the government releases only a minority of ownership shares to the public. Another lesson is that privatization, while important, is not sufficient to create a market economy. Eastern European governments focused on privatization as a quick fix, something that could be executed quickly and would bring about effective competition just as quickly. Yet business creation is responsible, the experience of these countries is consistent with the idea that constraints on private business creation can hinder growth. The abolition of price controls occurred in a more uniform fashion in Eastern Europe. Within a few years after the fall of communism, price controls had been sharply reduced across the region (Table 2 ). This resulted in the virtual elimination of shortages and an eventual increase in product quality to near-Western levels. Reversion to the previous regime has been a possibility only in Russia, where many expressed hope that firms would voluntarily limit their price increases in the name of protecting the consumer (Arrow and Phelps 1993). Indeed, policymakers almost reinstated strict price controls in 1993 to counter what one legislator called the “greed” of producers, and the legislature enacted a number of provisions to punish those deemed to have raised prices too far or too fast (Bush 1993). This lack of political commitment to reform deterred entry and shifted business toward the black market (which has seen astonishing growth in Russia) while prolonging the presence of queues that in turn have generated widespread public dissatisfaction with the reform effort.7 Several Russian regions reintroduced price controls in 1998 to quell this dissatisfaction, and the national government has discussed doing the same (LaFraniere 1998). Such a move would reduce competition in the Russian economy and, ironically, exacerbate many of the economic problems to which the public objects. CONCLUSION This article explores problems and possibilities associated with privatization and the transition to a market economy. In it, I suggest that privatization at its most basic level—the distribution of shares from the government to the private sector—is difficult to achieve and insufficient to bring about economic growth. I also consider the importance of appropriate owner/manager incentives and conclude that privately created holding companies may be the best way to ensure that de jure privatization becomes effective privatization. Finally, I consider the importance of legal constraints on private business creation and conclude that a competition-enhancing legal framework is vital to the success of privatization. What lessons can be learned from the Eastern European privatization experience and, perhaps, applied to emerging economies such FEDERAL RESERVE BANK OF DALLAS 23 ECONOMIC REVIEW FOURTH QUARTER 1998 privatization does not by itself create competition. If communist-era restraints on private business creation were to be maintained in China, entrepreneurs would be unable to compete against newly privatized companies. This could mean that privatization would only convert government-run monopolies into privately held monopolies, with minimal employment opportunities for workers laid off during the privatization process. Such a result might exacerbate the unemployment shocks associated with the transition to a market economy without providing any of the efficiency gains that result from competition—probably the worst imaginable outcome for those who wish to help unleash the Chinese economy through privatization. Chinese policymakers have much to learn from the Eastern European privatization experience. As China embarks upon its ambitious privatization program, it remains to be seen whether its leaders will look to Eastern Europe for guidance on how to enact a successful program— or whether they will simply repeat Eastern Europe’s mistakes. REFERENCES Ali, Abdiweli (1997), “Economic Freedom, Democracy and Growth,” Journal of Private Enterprise 13 (Winter): 1– 20. Arrow, Kenneth J., and Edmund S. Phelps (1993), “Proposed Reforms of the Economic System of Information and Decision in the USSR: Commentary and Advice,” in Privatization Processes in Eastern Europe: Theoretical Foundations and Empirical Results, ed. Mario Baldassarri and Luigi Paganetto (New York: St. Martin’s Press), 15–47. Bird, Chris (1998), “Voters in Quandary over Moldova’s Political Zoo,” Reuters Online News Service, March 19 <http://www.infoseek.com>. Blanchard, O., and R. Layard (1992), “How to Privatise,” in The Transformation of Socialist Economies, ed. Horst Siebert (Tubingen, Germany: Mohr [Siebeck]), 27– 44. Bolton, Patrick, and Gerard Roland (1992), “Privatization Policies in Central and Eastern Europe,” Economic Policy: A European Forum 15 (October): 275 – 309. NOTES 1 2 3 4 5 6 7 Boycko, Maxim, Andrei Shleifer, and Robert W. Vishny (1994), “Voucher Privatization,” Journal of Financial Economics 35 (April): 249 – 66. I would like to thank Steve Brown, Steve Prowse, Lori Taylor, and Mine Yücel for helpful comments. These Eastern European states are the Czech Republic, Hungary, Poland, Romania, and Slovakia. A study by Ali (1997) finds that increased economic freedom generally leads to increased economic growth. One example is the landmark 1984 privatization of British Telecom (Financial Times 1995). This is a standard portfolio-theory result applied to privatization. See Sharpe (1970) for a more general discussion of portfolio theory. Hansen (1997) presents an alternative view in which mass privatization is desirable precisely because no single investor has the power to oversee management. According to this view, managers who face no shareholder oversight can transfer any profit from the shareholders to themselves, which gives managers a powerful incentive to maximize profit. See Dittus and Prowse (1996) for an interesting and informative discussion of this issue. Responding to this dissatisfaction, the Russian legislature passed a nonbinding resolution in June 1997 that called for the renationalization of many previously privatized companies. The nonbinding measure passed by a vote of 288 to 6 (Radio Free Europe/ Radio Liberty 1997). Brada, Josef C. (1992), “The Mechanics of the Voucher Plan in Czechoslovakia,” RFE/RL Research Report 1 (17): 42 – 45. ——— (1996), “Privatization Is Transition— Or Is It?,” Journal of Economic Perspectives 10 (Spring): 67– 86. Bush, Keith (1993), “Chernomyrdin’s Price Control Decree is Revoked,” RFE/RL Research Report 2 (5): 35 – 37. Capelik, Vladimir (1992), “The Development of Antimonopoly Policy in Russia,” RFE/RL Research Report 1 (34): 66 – 70. Dittus, Peter, and Stephen Prowse (1996), “Corporate Control in Central Europe and Russia: Should Banks Own Shares?” in Corporate Governance in Central Europe and Russia, Volume 1, Insiders and the State, ed. Roman Frydman, Cheryl W. Gray, and Andrzej Rapaczynski (Budapest: Central European University Press), 20 – 67. Dyck, I. J. Alexander (1997), “Privatization in Eastern Germany: Management Selection and Economic Transition,” American Economic Review 87 (September): 565 – 97. 24 European Bank for Reconstruction and Development (1994), Transition Report (London). Pataki, Judith (1993), “Hungary: Domestic Political Stalemate,” RFE/RL Research Report 2 (1): 92 – 95. ——— (1997), Transition Report (London). Patterson, Perry L. (1993), “Overview: The Return of the Private and Cooperative Sectors in Eastern Europe and the Soviet Union,” in Capitalist Goals, Socialist Past, ed. Perry L. Patterson (Boulder, Colo.: Westview Press), 1–16. Feinberg, Robert M., and Mieke Meurs (1994), “Privatization and Antitrust in Eastern Europe: The Importance of Entry,” Antitrust Bulletin 39 (Fall): 797– 811. Financial Times (1995), “A Decade of Privatization,” December 27, 9. Percival, Lindsay (1997), “Albania: Pyramid Schemes Common Across Europe,” RFE/RL Research Report 6 (1): 31– 33. Gray, Cheryl (1996), “In Search of Owners: Privatization and Corporate Governance in Transition Economies,” World Bank Research Observer 11 (August): 179 – 97. Radio Free Europe/Radio Liberty (1997), “Duma Slams Privatization Results,” Newsline on the Web, June 6 <http://www.rferl.org/newsline>. Grime, Keith, and Vic Duke (1993), “A Czech on Privatization,” Regional Studies 27 (8): 751– 57. Roland, Gerard (1994), “On the Speed and Sequencing of Privatisation and Restructuring,” Economic Journal 104 (September): 1158 – 68. Hansen, Nico A. (1997), “Privatization, Technology Choice and Aggregate Outcomes,” Journal of Public Economics 64 (June): 425 – 42. Seely, Robert (1993), “Shops Go on Block as Ukraine Privatizes,” Los Angeles Times, February 21, A4. Sharpe, William F. (1970), Portfolio Theory and Capital Markets (New York: McGraw-Hill Book Company). Jensen, Michael C., and William H. Meckling (1976), “Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure,” Journal of Financial Economics 3 (October): 305 – 60. Slay, Ben (1993), “The Indigenous Private Sector in Poland,” in Capitalist Goals, Socialist Past, ed. Perry L. Patterson (Boulder, Colo.: Westview Press), 19 – 39. Johnson, Simon (1994), “Private Business in Eastern Europe,” in The Transition in Eastern Europe Volume 2: Restructuring, ed. Olivier Jean Blanchard, Kenneth A. Froot, and Jeffrey D. Sachs (Chicago: University of Chicago Press), 245 – 91. Tirole, Jean (1994), “Western Prudential Regulation: Assessment, and Reflections on Its Application to Central and Eastern Europe,” Economics of Transition 2 (June): 124 – 49. Kotrba, Josef, and Jan Svejnar (1994), “Rapid Multifaceted Privatization: Experience of the Czech and Slovak Republics,” Economic Policy in Transitional Economies 4 (2): 147– 85. Vickers, John, and George Yarrow (1991), “Economic Perspectives on Privatization,” Journal of Economic Perspectives 5 (Spring): 111– 32. LaFraniere, Sharon (1998), “Legislative Leaders Decry Moscow’s Political Vacuum,” Washington Post, September 10, A26. Voszka, Eva (1993), “Spontaneous Privatization in Hungary,” in Privatization in the Transition to a Market Economy: Studies of Preconditions and Policies in Eastern Europe, ed. John S. Earle, Roman Frydman, and Andrzej Rapaczynski (New York: St. Martin’s Press), 89 –107. Lyle, Robert (1997), “China: ‘Survival of the Fittest’ Drives Privatization,” Radio Free Europe/Radio Liberty Weekday Magazine, September 15 <http://www.rferl.org/nca/ features>. Weingast, Barry R. (1995), “The Economic Role of Political Institutions: Market-Preserving Federalism and Economic Development,” Journal of Law, Economics and Organization 11 (April): 1– 31. Melloan, George (1998), “Communism’s Ghosts Trouble Europe’s Grand Plan,” Wall Street Journal, April 7, A19. Nove, Alec (1992), An Economic History of the USSR: 1917–1991 (New York: Penguin Books). FEDERAL RESERVE BANK OF DALLAS 25 ECONOMIC REVIEW FOURTH QUARTER 1998 Many analysts believe that adverse climate changes in the form of global warming are— or soon will be—under way as the result of anthropogenic emissions of greenhouse gases. (See the box entitled “What Is Global Warming?”) The largest such source of these gases is carbon dioxide (CO2 ) resulting from the growing consumption of fossil fuels (petroleum products, natural gas, and coal).1 Consequently, the conservation of fossil fuels figures prominently in strategies to reduce CO2 emissions. Increasing concerns about the extent of global warming and its potential consequences culminated in a United Nations conference in Kyoto, Japan, in late 1997. Prior to the conference, President Clinton proposed that the United States join with other industrialized countries in setting a target for reducing CO2 emissions in each country to 1990 levels by 2010. By the end of the conference, emissaries from the industrialized countries had agreed to a target 7 percent below 1990 levels by 2010 and to make further reductions in subsequent years. Developing countries would be expected to reduce their CO2 emissions in future years as their incomes rise. As shown in Figure 1, the U.S. Department of Energy has projected that CO2 emissions from the consumption of fossil fuels in the industrialized countries will have increased about 30 percent from 1990 to 2010. Therefore, compliance with the Kyoto accord would represent a sizable reduction in the use of fossil fuels from what could otherwise be expected. My analysis indicates that the developing countries would consume nearly 12 percent of the amount of fossil fuels the industrialized countries must conserve to comply with the Kyoto accord. The net effect Global Warming Policy: Some Economic Implications Stephen P. A. Brown Senior Economist and Assistant Vice President Federal Reserve Bank of Dallas C ompliance with the Kyoto accord would represent a sizable reduction in the use of fossil fuels. Figure 1 Annual Carbon Emissions, 1990–2015 Billions of metric tons 12 10 World 8 Effect of the Kyoto accord 6 OECD countries 4 Effect of the Kyoto accord 2 0 1990 1995 2000 2005 2010 2015 SOURCES: Energy Information Administration (1997); author’s calculations. 26 What Is Global Warming? Global warming is a scientific theory or scenario in which increased levels of atmospheric CO2 are linked to generally rising temperatures around the world. To better understand global warming, start with the greenhouse effect. Sunlight heats the earth, but the earth would be far cooler if not for the presence of water vapor and greenhouse gases in the atmosphere. These gases let sunlight through to warm the earth but trap some of the heat escaping back into space in the form of infrared radiation. In this manner, the gases act like the glass walls and ceiling of a greenhouse. Increasing the level of greenhouse gases in the atmosphere is like using thicker glass in the greenhouse: less heat escapes. Many scientists believe the CO2 released by human activities is intensifying the greenhouse effect and contributing to an increase in the earth’s overall temperature. This general increase in the earth’s temperature is commonly known as global warming. Many scientists and other people are concerned about global warming’s potential effect on the environment. Among the predicted consequences are increased rainfall, melting polar ice caps, rising ocean levels, increased flooding, and widespread crop failure. The evidence of global warming is inconclusive, but most scientists who study the issue think that it is occurring to some degree. Nevertheless, there is considerable uncertainty about the magnitude of the change, whether CO2 is a contributing factor, and the environmental consequences. would be to slow the growth of global CO2 emissions from the projected 45 percent to 30 percent between 1990 and 2010.2 Some worry that compliance with the Kyoto accord would impose drastic costs on the industrialized countries with little proven benefit. Others worry the Kyoto targets are too modest to prevent costly environmental problems. These concerns raise several questions. What is the rationale for government intervention in markets to reduce CO2 emissions? By how much does economic analysis suggest that the United States reduce its CO2 emissions, and how do President Clinton’s proposal and the Kyoto accord compare with what is optimal? The former question can be answered with simple economic theory. The latter question can be answered by combining estimates of the economic benefits of reducing CO2 emissions with the opportunity costs of doing so. the debates that occurred at the UN conference on global warming. The representatives of each country jockeyed for advantage and criticized each other for not doing enough. At the international level, debate has been exacerbated by the fact that reducing energy consumption on a global scale yields gains in the terms of trade for energy-importing countries. CALL FOR GOVERNMENT POLICY Although nature contributes its own CO2 to the atmosphere, policy is more easily directed to the CO2 contributed by human activity. The principal way people contribute to atmospheric CO2 is through the consumption of carbon-based fuels. These fuels include petroleum products, natural gas, coal, and wood. Jointly, the first three are often identified as “fossil fuels.” Deforestation is another way people contribute to increased levels of atmospheric CO2. Scientists estimate that the world’s forests remove about one-third of the current CO2 emissions from fossil fuels. Thus, large reductions in the world’s forests could significantly affect the atmospheric levels of CO2. Nevertheless, recent changes in the earth’s forests have had little effect on atmospheric CO2 in comparison with the effects from fossil fuel consumption. At the pragmatic level, the call for government action on global warming comes from the concern that rising consumption of carbonbased fuels will increase levels of atmospheric CO2, which will lead to warmer temperatures that could harm the environment. At a more analytical level, however, when consumers act individually they lack an incentive to consider the global side effects resulting from their consumption of carbon-based fuels. Taken collectively, individual actions could be contributing to greater emissions of CO2 than are desirable from the perspective of economic efficiency. To some extent, the divergence between individual and global interests can be seen in FEDERAL RESERVE BANK OF DALLAS EVALUATING GLOBAL WARMING POLICY: A COST–BENEFIT APPROACH Economics offers two approaches for evaluating how much effort should be put forth to reduce global warming. One approach is to set an appropriate tax on carbon-based fuels. Another is to use cost–benefit analysis. The cost – benefit approach offers two advantages. It allows us to consider more directly the uncertainty about the benefits of reducing CO2 emissions. It also allows representation of the wide range of debate about global warming policy. Under this approach, a reduction in environmental harm is the benefit of reducing CO2 emissions. The foregone economic opportunities from using less fossil fuel are the cost. Benefits of Reducing CO2 Emissions The benefits of reducing CO2 emissions are the environmental damages that are avoided by preventing rising concentrations of atmospheric CO2 that intensify the greenhouse effect and boost global temperatures. Potential environmental damage from global warming includes a variety of effects, from the impact on 27 ECONOMIC REVIEW FOURTH QUARTER 1998 Table 1 Reference Case Quantities, Prices, and Elasticities United States Consumption Oil Natural gas Coal Other Production Oil Natural gas Coal Other Other OECD Consumption Oil Natural gas Coal Other Production Oil Natural gas Coal Other C/EE/FSU Consumption Oil Natural gas Coal Other Production Oil Natural gas Coal Other OPEC Consumption Oil Natural gas Coal Other Production Oil Natural gas Coal Other Other LDCs Consumption Oil Natural gas Coal Other Production Oil Natural gas Coal Other World reference prices Oil Natural gas Coal agriculture and forests to the costs of coping with more severe weather, flooding of coastal property, and increased disease. An emerging literature attempts to evaluate the economic costs associated with the potential environmental damage caused by CO2 emissions. (For examples, see Fankhauser 1994; Hope and Maul 1996; Nordhaus 1991a, 1991b, 1992, 1993; and Peck and Teisberg 1993a, 1993b.) Analysts and researchers working in the field must contend with several uncertainties: the extent to which these emissions will affect global warming, the environmental harm caused by global warming, and the economic costs of those environmental effects. The emerging literature provides estimates of the benefits of reducing CO2 emissions. This literature suggests the worldwide marginal benefit from a reduction in CO2 emissions occurring in 2010 plausibly ranges from $0 to $33.75 (in 1995 dollars) per barrel of oil equivalent. (In this case, barrel of oil equivalent refers to the CO2 emissions resulting from the consumption of a barrel of oil.) The literature also suggests that the most likely range for worldwide benefits is about $.92 to $6.61 per barrel of oil equivalent, as is shown in Figure 2. The mean estimate is $2.86 per barrel of oil equivalent.3 Given the considerable uncertainty about the benefits of reducing CO2 emissions, some analysts have suggested that making the reductions is similar to insurance in one important respect: the costs of reducing emissions are fairly well known, but the benefits cannot be known with any certainty. Price elasticity of fuel at left with respect to price of Quantity (1015 Btu) Oil Natural gas Coal Other 42.5 29.2 22.8 14.2 –.72 .25 .12 .05 .25 –.72 .63 .05 .03 .10 –.96 .10 .06 .06 .06 –.50 17.9 24.9 25.2 14.2 .51 58.8 34.2 18.1 29.1 –.72 .25 .12 .05 25.7 24.2 18.3 29.1 .43 17.8 33.2 13.1 6.5 –.225 .075 .04 .02 20.4 43.3 14.3 6.5 .30 11.3 5.6 .3 .4 –.72 .25 .12 .05 72.6 5.6 .3 .4 * 64.5 26.8 68.5 17.0 –.45 .15 .08 .04 58.3 31.0 64.7 17.0 .43 .51 1.86 1.00 .25 –.72 .63 .05 .03 .10 –.96 .10 .10 .10 .10 –.50 .43 1.86 1.00 .075 –.225 .20 .04 .01 .04 –.31 .04 .05 .05 .05 –.25 .30 1.24 1.00 .25 –.72 .63 .10 .00 .00 –.96 .10 .01 .01 .10 –.50 Estimating the Costs of Reducing CO2 Emissions Completing the cost – benefit analysis requires estimates of the costs of reducing CO2 emissions through conservation of fossil fuels. Following several previous studies, marginal cost estimates are obtained by using a welfaretheoretic framework built on top of a simulation model of world energy markets.4 Many analysts use U.S. Department of Energy (DOE) projections as a reference standard for analysis, and the simulation model is calibrated to reproduce the DOE’s 1997 projections for world energy market conditions in 2010, as shown in Table 1. The data in the table represent one of many possible world energy outlooks for 2010.5 Table 1 also summarizes representative estimates of the long-run supply and demand responses to prices for the major regional areas in the analysis. These estimates are derived from a variety of sources. The oil price elasticities of supply and demand are based on an Energy .40 1.65 1.00 .15 –.45 .40 .08 .02 .10 –.61 .08 .08 .08 .08 –.50 .43 1.86 1.00 $/106 Btu † $3.519 $1.9553 $ .7924 $/standard unit † $20.41 per barrel $2.01 per Mcf $16.919 per short ton * OPEC adjusts its production to maintain a constant share of the oil market. † Prices are in 1995 dollars. 28 Figure 2 Estimated Costs and Benefits of CO2 Abatement 1995 dollars per barrel of oil equivalent 50 40 Marginal cost—world 30 20 10 High estimate Marginal benefit 0 Extreme precautionary approach Risk-neutral approach Reduce emissions to 1990 level Reduce emissions 3% below 1990 level Low estimate Reduce emissions 7% below 1990 level –10 0 38 77 115 154 192 230 269 307 346 384 422 461 499 538 576 Millions of metric tons of carbon Modeling Forum study (1991) that compares ten major world oil market models.6 Elasticities for other fuels are calibrated to estimates adapted from Bohi (1981) and Brown and Yücel (1995). Following Brown and Huntington (1998), the responses for the region that encompasses China, Eastern Europe, and the former Soviet Union (C/EE/FSU) are judgmental. The production and consumption decisions in these countries are more likely to be influenced by the forces of economic transition than by changes in world energy prices. In fact, if the supply and demand responses for the C/EE/FSU were made comparable to responses for other country groups, the conservation scenarios considered here would reduce world energy prices enough that the model would predict these economies would import significant quantities of energy. The author considers such a result untenable and therefore assumes a smaller response to price than for other countries. To the extent that these countries are more sensitive in their response to price, the estimated costs of achieving various world conservation targets without cooperation from these countries would be larger than reported here, but the thrust of the current analysis would be unchanged. The response of oil producers within OPEC is highly uncertain. To date, formal modeling of OPEC decisions has been far from reliable. OPEC appears to operate like an imperfect cartel at some times but not at others. The OPEC FEDERAL RESERVE BANK OF DALLAS countries appear to be about as uncomfortable with a rapidly increasing market share (such as accompanied the relatively low prices of the 1960s) as they are with a rapidly decreasing market share (such as occurred in the aftermath of the price hikes of the late 1970s and early 1980s). The analysis presented here assumes that OPEC acts to maintain a constant market share.7 Cost estimates are obtained by computing the welfare costs of policies under which the United States works in concert with other countries in the Organization for Economic Cooperation and Development (OECD) to reduce global CO2 emissions through fossil fuel conservation. The modeling framework allows world energy prices to adjust to the conservation of fossil energy to restore a balance between supply and demand conditions in each market. Analytically, carbon taxes are used to reduce the consumption of fossil fuels in the two country groups. The tax approach assumes that conservation measures are applied across all end uses. Values from these simulations are used with equations in the appendix to construct marginal cost curves for U.S. abatement of CO2 emissions. This methodology follows the welfare-theoretic approach previously employed by Brown and Huntington (1994a, 1994b, 1998) and Felder and Rutherford (1993). The resulting cost curves take into account a number of factors, including the direct welfare costs of U.S. conservation efforts, transfers of wealth 29 ECONOMIC REVIEW FOURTH QUARTER 1998 Figure 3 Estimated Costs and Benefits of CO2 Abatement 1995 dollars per barrel of oil equivalent 50 40 Marginal cost—world 30 20 Free lunch Marginal cost—U.S. 10 High estimate Marginal benefit 0 Reduce emissions to 1990 level Reduce emissions 3% below 1990 level Low estimate Reduce emissions 7% below 1990 level –10 0 38 77 115 154 192 230 269 307 346 384 422 461 499 538 576 Millions of metric tons of carbon between countries, the effect lower energy prices would have in stimulating energy consumption in nonparticipating countries, and the economic cost of OPEC cartelization.8 oil equivalent, substantially more than the mean estimate of marginal benefit of $2.86 per barrel of oil equivalent or the likely upper bound estimate of $6.61 per barrel of oil equivalent. Compliance with the Kyoto accord would seem to imply 478 metric tons of CO2 abatement in 2010 from fossil fuel conservation, at a marginal cost of nearly $25 per barrel of oil equivalent. U.S. officials expect to use offsets and other credits, however, to reduce the burden from conservation of fossil fuels to 3 percent below 1990 levels. That implies 424 metric tons of CO2 abatement from the conservation of fossil fuels, at a marginal cost of just under $20 per barrel of oil equivalent. Costs of Reducing CO2 Emissions The first marginal cost curve, which is labeled “Marginal cost—world,” is shown in Figure 2. This curve represents the marginal costs to the world of the U.S. fossil fuel conservation necessary to reduce CO2 emissions. Economic well-being is maximized at the level of CO2 abatement where the marginal cost equals the marginal benefit. The risk-neutral approach would be to equate marginal cost to the mean estimate of marginal benefit. As shown in Figure 2, estimated marginal cost equals the mean estimate of marginal benefit at 65 metric tons of CO2 abatement. An extreme precautionary approach to avoiding the risk of global warming suggests using the upper bound estimate of the likely range as the measure of marginal benefit. As shown in Figure 2, marginal cost equals this upper bound at just under 200 metric tons of CO2 abatement. In comparison, President Clinton’s proposal to reduce U.S. emissions to 1990 levels by 2010 implies 384 metric tons of CO2 abatement, a figure substantially higher than what is optimal under either measure of marginal benefit. At this level of CO2 abatement, the marginal cost is more than $16 per barrel of Improved Terms of Trade Because the United States imports oil and natural gas, it can improve its terms of trade by conserving those fuels. The improved terms of trade mean that the United States incurs lower costs for its actions to reduce CO2 emissions than the world incurs from the U.S. actions. As shown by the curve labeled “Marginal cost— U.S.” in Figure 3, the lower marginal cost implies that a greater reduction in CO2 emissions is optimal. Such a conclusion is flawed, however. Optimality cannot be found by equating the marginal benefit to the world to the marginal cost to the United States. Sound analysis requires consistency in defining the incidence of costs and benefits. 30 When regulation is used instead of broad market incentives, such as taxes, the lowest cost methods of energy conservation can be ignored. Interference in free trade is another way policies to reduce CO2 emissions can have hidden costs. Broad programs of energy conservation permit energy-importing countries to improve their terms of trade with energyexporting countries—a fact that has not been lost on OPEC. The countries that are the most dependent on imported energy have been the most aggressive in promoting global cooperation to reduce CO2 emissions. More self-sufficient countries, such as the United States, have been more reluctant to participate. Within the United States, energy conservation has a decided tilt toward the conservation of oil, the fuel for which we are most import-dependent. Rent-seeking behavior is another way in which policy can have hidden costs. Changes in policy create winners and losers. Both groups have an incentive to expend real resources to achieve their objectives by influencing the political process, which can add sizable costs to policy. Using an approach first suggested by Tullock (1967), the author estimates the potential hidden costs of policy. As shown in Figure 4, these costs can be sizable. With the hidden costs incorporated, the marginal cost of reducing U.S. CO2 emissions is more than $30 per barrel of oil equivalent at zero abatement and more than $80 at full compliance with the Kyoto Free Lunch: There Is No Such Thing A number of energy analysts argue that the United States can cut its energy consumption by 25 percent and achieve a cost savings at the same time. As shown by the curve labeled “Free lunch” in Figure 3, President Clinton’s target for reducing CO2 emissions would be very close to optimal if the free-lunch cost curve represented reality. In fact, some analysts who consider this cost curve accurate and also favor the extreme precautionary approach to reducing CO2 emissions have criticized the president’s target as too conservative. As shown in Figure 3, achieving the target of reducing CO2 emissions from fossil fuels 3 percent below 1990 levels would have a marginal cost that is above the mean estimate of marginal benefit but below the likely upper bound. Analysts who believe in the free lunch use conceptually flawed studies to support their claims. What these analysts are unable to explain is why a free market would leave such cost savings on the table. Instead, they offer vague explanations of market barriers, including inappropriate lifestyle choices, and demand government regulation to reduce what they see as wasteful use of energy (see Brown 1996). Hidden Costs of Policy Economic policy often carries with it hidden costs that are not captured by traditional welfare-theoretic measures. Regulatory inefficiency is one way in which costs can escalate. Figure 4 Estimated Costs and Benefits of CO2 Abatement 1995 dollars per barrel of oil equivalent 140 120 100 Marginal cost, including hidden cost 80 60 40 Marginal cost—world 20 Marginal benefit 0 0 38 77 115 154 192 230 269 307 346 384 422 461 499 Millions of metric tons of carbon FEDERAL RESERVE BANK OF DALLAS 31 ECONOMIC REVIEW FOURTH QUARTER 1998 538 576 Aggregate Economic Consequences for the United States In terms of foreFigure B.1 gone economic oppor- U.S. CO Abatement and GDP Losses 2 tunities, the energy Percentage of GDP loss conservation associ0 ated with CO2 abatement can be seen as –1 equivalent to increasing the price of –2 energy, which carries with it the expectation –3 Low estimate of slower economic High estimate growth. Most econo–4 mists agree that GDP is a poor way to mea–5 Reduce sure economic wellReduce emissions emissions to 1990 level being, particularly 3% below Reduce –6 1990 level emissions when evaluating envi7% below 1990 level ronmental policy. –7 Nevertheless, slower 0 38 77 115 154 192 230 269 307 346 384 422 461 499 538 576 economic growth can Millions of metric tons of carbon result in significant political pressure on fiscal and monetary authorities to offset the slowdown. Yielding to that pressure could lead to higher inflation. As shown in Figure B.1, the effect on aggregate economic activity depends on the amount of CO2 abatement. The amount of energy conservation required to reduce 2010 CO2 emissions to the 1990 level would imply that U.S. GDP would be 2.7 to 3.7 percent lower in 2010 than it would otherwise be.* Assuming the United States could use offsets and credits, compliance with the Kyoto accord would imply that U.S. GDP would be 3 to 4.3 percent lower in 2010. These estimates imply that if the United States embarked on a ten-year program to achieve compliance with the Kyoto accord, U.S. GDP growth would be 0.3 to 0.4 percent lower per year. income. Cost–benefit analysis suggests that reducing U.S. emissions of CO2 to comply with the Kyoto accord or to reach the more modest target proposed by President Clinton represents too much insurance. Analysis for the other OECD countries yields a similar result. It is not surprising, therefore, that little headway has been made in ratifying the accord. NOTES 1 2 3 * GDP loss estimates are obtained through elasticities that relate energy prices to aggregate economic activity. The elasticities were chosen to represent the range of estimates from a number of prominent economic studies. accord. If the potential for hidden costs were taken seriously, the cost of reducing CO2 emissions would greatly outweigh its benefits. 4 CONCLUSIONS 5 Global warming is a theory partially supported by the facts. More evidence is needed to definitively conclude that the rising atmospheric level of CO2 resulting from the use of carbonbased fuels is causing global warming. Nevertheless, most scientists who study the issue think that the use of fossil fuels is contributing at least some of the global warming that appears to be occurring. Despite this seeming consensus, there is considerable uncertainty about both the magnitude and the environmental consequences of global warming. Given these uncertainties, reducing CO2 emissions is like insurance against global warming and its possible environmental consequences. Under most current proposals, the developed nations would buy all or most of the insurance. Developing nations would be asked to contribute once they attain higher levels of 6 7 8 32 The author would like to thank Frank Berger, Sterling Burnett, Mike Canes, Roger Hemminghaus, Hill Huntington, Don Norman, Steve Prowse, Cece Smith, Ron Sutherland, Lori Taylor, Mine Yücel, and Carlos Zarazaga for helpful comments and discussions without implicating them in the conclusions. For a comprehensive treatment of emissions other than CO2, see Hall (1990, 1992). The Department of Energy projection that world CO2 emissions from fossil fuels would increase about 45 percent from 1990 to 2010 owes greatly to expectations of accelerating growth and industrialization in developing areas of the world. These estimates of damage are adapted from Brown and Huntington (1998). Previous analysis suggests a flat marginal damage curve. Summarizing the previous literature, Peck and Teisberg (1992) explain that marginal damage costs are essentially unaffected by the emissions levels in any given decade. This conclusion rests on the finding that temperature change depends on gas concentration, which is not greatly affected by emissions levels in any given decade. I follow this characterization by assuming horizontal damage curves that depict a constant level of benefits for any level of CO2 abatement. The author developed the analytical framework with Hillard G. Huntington of Stanford University. The projected energy-demand conditions depend on a variety of assumptions about economic growth and the extent of energy-saving technological change in the absence of price change. The supply conditions other than OPEC oil production incorporate assumptions about the resource base, engineering constraints on developing resources, and producer-country taxes and policies. In these projections, OPEC members satisfy the excess demand but adjust the next period’s price in response to market tightness. See Huntington (1992, 1993). Griffin (1985) and Dahl and Yücel (1991) provide empirical estimates of OPEC behavior that are broadly consistent with this view. My analysis abstracts from a number of considerations featured in other studies of energy conservation. Hoel (1991a) and Newberry (1992) consider the effects of other taxes and redistributive policies. Bohm (1993); Brown and Huntington (1994b); Eyckmans, Proost, and Schokkaert (1993); Hoel (1991b and 1994); Manne and Rutherford (1994); Welsch (1995); and Whalley and Wigle (1991) consider alternative policies for distributing conservation goals across countries and gains from cooperation. Felder and Rutherford (1993) and Pezzey (1992) allow for different types of goods. “Unilateral CO2 Reductions and Carbon Leakage: The Consequences for International Trade in Oil and Basic Materials,” Journal of Environmental Economics and Management 25 (September): 162 – 76. Griffin, James M. (1985), “OPEC Behavior: A Test of Alternative Hypotheses,” American Economic Review 75 (December): 954 – 63. REFERENCES Bohi, Douglas (1981), Analyzing Demand Behavior: A Study of Energy Elasticities (Baltimore, Md.: Johns Hopkins University Press for Resources for the Future). Hall, Darwin C. (1990), “Preliminary Estimates of Cumulative Private and External Costs of Energy,” Contemporary Policy Issues 8 (July): 283 – 307. Bohm, Peter (1993), “Incomplete International Cooperation to Reduce CO2 Emissions: Alternative Policies,” Journal of Environmental Economics and Management 24 (May): 258 – 71. ——— (1992), “Social Cost of CO2 Abatement from Energy Efficiency and Solar Power in the United States,” Environmental and Resource Economics 2 (5): 491– 512. Brown, Stephen P. A. (1996), “Directions for U.S. Energy Conservation and Independence,” Business Economics 31 (October): 25 – 30. Hoel, Michael (1991a), “Efficient International Agreements for Reducing Emissions of CO2,” The Energy Journal 12 (2): 93 –107. Brown, Stephen P. A., and Hillard G. Huntington (1994a), “The Economic Cost of U.S. Oil Conservation,” Contemporary Economic Policy 12 (July): 42 – 53. ——— (1991b), “Global Environmental Problems: The Effects of Unilateral Actions Taken by One Country,” Journal of Environmental Economics and Management 20 (January): 55 – 70. ——— (1994b), “LDC Cooperation in World Oil Conservation,” The Energy Journal, Special issue, 310 – 28. ——— (1994), “Efficient Climate Policy in the Presence of Free Riders,” Journal of Environmental Economics and Management 27 (November): 259 – 74. ——— (1998), “Some Implications of Increased Cooperation in World Oil Conservation,” Federal Reserve Bank of Dallas Economic Review, Second Quarter, 2 – 9. Hope, Chris, and Phillip Maul (1996), “Valuing the Impact of CO2 Emissions,” Energy Policy 24 (March): 211–19. Brown, Stephen P. A., and Mine K. Yücel (1995), “Energy Prices and State Economic Performance,” Federal Reserve Bank of Dallas Economic Review, Second Quarter, 13 – 21. Huntington, Hillard G. (1992), “Inferred Demand and Supply Elasticities from a Comparison of World Oil Models,” in International Energy Economics, ed. Thomas Sterner (London: Chapman and Hall), 239 – 61. Dahl, Carol, and Mine K. Yücel (1991), “Testing Alternative Hypotheses of Oil Producer Behavior,” The Energy Journal 12 (4): 117– 38. ——— (1993), “OECD Oil Demand: Estimated Response Surfaces for Nine World Oil Models,” Energy Economics 15 (January): 49 – 66. Energy Information Administration, U.S. Department of Energy (1997), 1997 International Energy Outlook (Washington, D.C.: Government Printing Office). Manne, Alan S., and Thomas F. Rutherford (1994), “International Trade in Oil, Gas and Carbon Emission Rights: An Intertemporal General Equilibrium Model,” The Energy Journal 15 (1): 57–76. Energy Modeling Forum (1991), International Oil Supplies and Demands, EMF Report 11 (Stanford, Ca.: Stanford University). Newberry, David M. (1992), “Should Carbon Taxes Be Additional to Other Transport Fuel Taxes?” The Energy Journal 13 (2): 49 – 60. Eyckmans, Johan, Stef Proost, and Erik Schokkaert (1993), “Efficiency and Distribution in Greenhouse Negotiations,” Kyklos 46 (3): 363 – 97. Nordhaus, William D. (1991a), “A Sketch of the Economics of the Greenhouse Effect,” American Economic Review 81 (May, Papers and Proceedings): 920 – 37. Fankhauser, Samuel (1994), “The Social Costs of Greenhouse Emissions: An Expected Value Approach,” The Energy Journal 15 (2): 157– 84. ——— (1991b), “To Slow or Not to Slow: The Economics of Global Warming,” Economic Journal 101 (July): 920 – 37. Felder, Stefan, and Thomas F. Rutherford (1993), FEDERAL RESERVE BANK OF DALLAS 33 ECONOMIC REVIEW FOURTH QUARTER 1998 ——— (1992), “The DICE Model: Background and Structure of a Dynamic Integrated Climate Economy Model of the Economics of Global Warming,” Cowles Foundation Discussion Paper no. 1009 (New Haven, Connecticut, February). Pezzey, John (1992), “Analysis of Unilateral CO2 Control in the European Community and OECD,” The Energy Journal 13 (3): 159 –71. Tullock, Gordon (1967), “The Welfare Costs of Tariffs, Monopolies, and Theft,” Western Economic Journal (June): 224 – 32. ——— (1993), “Optimal Greenhouse Gas Reductions and Tax Policy in the ‘DICE’ Model,” American Economic Review (May, Papers and Proceedings): 313 –17. Welsch, Heinz (1995), “Incentives for Forty-Five Countries to Join Various Forms of Carbon Reduction Agreements,” Resource and Energy Economics 17 (November): 213 – 37. Peck, Stephen C., and Thomas J. Teisberg (1992), “CETA: A Model for Carbon Emissions Trajectory Assessment,” The Energy Journal 13 (1): 55 –77. Whalley, John, and Randall Wigle (1991), “Cutting CO2 Emissions: The Effects of Alternative Policy Approaches,” The Energy Journal 12 (1): 109 – 24. ——— (1993a), “CO2 Emissions Control: Comparing Policy Instruments,” Energy Policy 21 (March): 222– 30. ——— (1993b), “Global Warming Uncertainties and the Value of Information: An Analysis Using CETA,” Resource and Energy Economics 15 (March): 71– 97. Appendix Some Analytics of CO2 Abatement A welfare-theoretic approach can be used to derive formulas for the marginal cost of CO2 abatement achieved through the conservation of carbonbased energy. For any country (or country grouping), the economic welfare obtained from the market for a particular source of energy is the sum of consumer and producer surpluses: (A.1) Wij = and no significant international trade in nonfossil energy, summing over the marginal effects of the emissions reduction policy on each fossil energy source yields the marginal cost of compliance for country i : (A.2) QDij ∫ PDij (Q j )dQ j − PijQDij MCi = n ∑ (PDij − PWj ) j =1 ∂QCij ∂Ei + QMij ∂PWj . ∂Ei In the above equation, Wij denotes the economic welfare country i obtains from the market for energy source j, QDij the quantity of primary energy j demanded in country i, PDij country i ’s demand price for energy source j (the market’s marginal valuation of consumption excluding externalities) at each quantity (Qj ), Pij the market price of energy source j in country i, QSij the quantity of energy j produced in country i, and PSij the domestic supply price of energy source j in country i (marginal cost of its oil production excluding externalities) at each quantity (Qj ). In the above equation, MCi denotes the gross marginal cost of reducing CO2 emissions through the conservation of carbon energy sources, PWj the world price of energy source j, QCij the quantity of energy source j that is conserved (where ∂QCij = – ∂QDij ), QMij country i ’s imports of energy source j, and Ei is the reduction in country i ’s emissions under the policy whose costs are being estimated. As Equation A.2 shows, the gross marginal cost of reducing emissions is the difference between the domestic and world prices of each carbon energy source (PDij – Pij) weighted by the shares of each fuel conserved for a one-unit reduction in CO2 emissions, minus (plus) the transfers obtained (lost) by reducing the price of imported (exported) carbon energy, noting that ∂PWj / ∂Ei is negative. Cost of Gross CO2 Abatement Cost of Net CO2 Abatement The most direct measure of the cost of reducing CO2 emissions is the welfare losses occurring in the energy markets that result from altering energy consumption to reduce emissions. Assuming no other distortions in domestic energy markets The net effect of the CO2 abatement actions taken by a country or group of countries is the quantity of their abatement minus the induced 0 + PijQSij − QSij ∫ PSij (Q j )dQ j . 0 (continued on next page) 34 Some Analytics of CO2 Abatement (continued ) change in CO2 emissions in the rest of the world. The change in CO2 emissions in nonparticipating countries depends on how their consumption of fossil energy is affected by a change in world energy prices and how the conservation actions in the participating countries affect the world oil price. Therefore, the relationship between a change in participant CO2 emissions and the net change in world CO2 emissions can be expressed as (A.3) World Perspective From the world perspective, the cost of reducing CO2 emissions through fossil energy conservation is the sum of costs borne by each country. From this perspective, net transfers cancel to zero. For every country or group of countries that obtains transfers from reduced prices for carbon energy, another country or group of countries yields an offsetting transfer, and Mij (∂PWj /∂Ei ) is exactly offset. Accounting for the offsetting transfers, as well as the distortion in world oil markets resulting from OPEC holding its oil production below free market levels, alters Equation A.4 to yield n ∂QDXj ∂PWj ∂EW = 1− ∑ E j . • ∂Ei ∂PWj ∂Ei j =1 In the above equation, EW denotes the amount by which world CO2 emissions are reduced, Ej the CO2 emissions associated with consuming one unit of carbon energy j, and QDXj the quantity of carbon energy j consumed by nonparticipating countries. Following Felder and Rutherford (1993) and Brown and Huntington (1994a, 1998), Equations A.2 and A.3 can be combined to express the marginal cost of the net world reduction in CO2 emissions for country (or country grouping) i. Specifically, multiplying the marginal cost of the gross reduction in CO2 emissions for country i by the net change in world CO2 emissions resulting from country i reducing its CO2 emissions yields (A.4) MCWi = n ∑ (PDij − PWj ) j =1 n j =1 • 1 − ∑ E j ∂QCij ∂Ei + QMi (A.5) −1 ∂EW . ∂Ei • In the above equation, MCWi denotes the net marginal cost to the world of country i ’s actions to reduce emissions, QCi 1 the amount of oil conserved, SO 1 OPEC’s share of world oil production, and CO 1 OPEC’s cost of oil production. ∂PWj ∂Ei Hidden Costs Tullock (1967) argues that economic agents have the incentive to expend real resources to influence economic policy up to the total value of the transfers that might result from the policy. In doing so they would dissipate the total value of the potential transfers as costs. Incorporating these hidden costs into the analysis alters Equation A.5 to yield ∂QDXj ∂PWj • . ∂PWj ∂Ei In the above equation, MCWi denotes the net marginal cost to country i of its actions to reduce world CO2 emissions. As Equation A.4 shows, the effect fossil energy conservation has on the cost of energy imports and on nonparticipant consumption of carbon energy are related through the effect conservation has on the world prices for these fuels. As cooperative conservation lowers the world prices of fossil energy, it reduces the cost of country i ’s energy imports and increases nonparticipant fossil energy consumption. If conservation has no effect on world energy prices, however, the energy-importing countries will not obtain terms-of-trade advantages, and fossil energy consumption will not be stimulated in nonparticipating countries. FEDERAL RESERVE BANK OF DALLAS ∂Q n MCWi = ∑ (PDij − PWj ) Cij ∂Ei = 1 j ∂QCi 1 SO1(PW 1 − CO1) + ∂Ei (A.6) ∂(PDij − PWj ) n MCHi = ∑ QDij ∂Ei j =1 ∂QCi 1 SO1(PW 1 − CO1) + ∂Ei −1 ∂EW . ∂Ei • In the above equation, MCHi denotes the net marginal cost to the world, inclusive of hidden costs, of country i ’s actions to reduce emissions. Note that QDij is the consumption of fuel j in country i. 35 ECONOMIC REVIEW FOURTH QUARTER 1998