The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
JULY/AUGUST 1997 ECONOMIC PERSPECTIVES A review from the Federal Reserve Bank of Chicago Reversal of fortune: Understanding the Midwest recovery Accounting for the federal government's cost of funds FEDERAL RESERVE BANK OF CHICAGO Contents Reversal of fortune: Understanding the Midwest recovery............................................................ 2 William A. Testa, Thomas H. Klier, and Richard H. Mattoon Has the Midwest been good or lucky in its recent recovery? In this article, the authors assess the internal and external factors that have contributed to the revival of midwestern economic fortunes over the last decade. Accounting for the federal government's cost of funds........................................................................... 18 George J. Hal, and Thomas J. Sargent This article describes and defends the authors’ corrections to the federal government’s flawed measure of its cost of funds. Further, it examines how the maturity structure of the debt influences the way inflation risk and interest rate risk are shared by the government and its creditors. ECONOMIC PERSPECTIVES July/August 1997, Volume XXI, Issue 4 President ECONOMIC PERSPECTIVES is published by the Research Department of the Federal Reserve Bank of Chicago. The views expressed are the authors’ and do not necessarily reflect the views of the management of the Federal Reserve Bank. Single-copy subscriptions are available free of charge. Please send requests for single- and multiple-copy subscriptions, back issues, and address changes to the Public Information Center, Federal Reserve Bank of Chicago, P.O. Box 834, Chicago, Illinois 60690-0834, telephone (312) 322-5111 or fax (312) 322-5515. Michael H. Moskow Senior Vice President and Director of Research William C. Hunter Research Department Financial Studies Douglas Evanoff, Assistant Vice President Macroeconomic Policy Charles Evans, Assistant Vice President Microeconomic Policy Daniel Sullivan, Assistant Vice President Regional Programs William A. Testa, Assistant Vice President Editor Helen O’D. Koshy Production Rita Molloy, Kathryn Moran, Yvonne Peeples, Roger Thryselius, Nancy Wellman ECONOMIC PERSPECTIVES is also available on the World Wide Web at http:// www.frbchi.org. Articles may be reprinted provided the source is credited and the Public Information Center is sent a copy of the published material. ISSN 0164-0682 Reversal of fortune: Understanding the Midwest recovery William A. Testa, Thomas H. Klier, and Richard H. Mattoon The Midwest economy has received considerable attention in recent years as it has shed its image as the Rust Belt and reemerged as a strong regional competitor both on the national and international stage. This reversal of fortune has surprised some analysts, and explanations of the region’s resurgent strength have often been more anecdotal than empirical. In this article, we take a more systematic approach to analyzing the elements that have contributed to the region’s recovery since the mid-1980s.1 Specifically, we describe the contribution of external and internal factors to the economic revival of the Midwest and identify the challenges and opportunities the region now faces. Charting the turnaround A region’s economy can be represented by many diverse measures. Unemployment rates are perhaps the most widely recognized indicators of both economic progress and participation of the region’s population in the economy. Looking at the aggregate unemployment rate for the Midwest versus the nation from the 1980s to date, two remarkable features can be seen.2 First, from an average annual rate that exceeded the nation’s by 3 percentage points in 1983, the Midwest’s unemployment rate had fallen to a full percentage point below the nation’s by 1996 (see figure 1). The same year marked the fifth consecutive year that the rate remained below the nation’s. A second feature of the labor market reflected by the unemployment rate is the behavior of the 2 Midwest economy during the most recent (1990–91) recession. In prior recessions, the highly cyclical nature of the Midwest economy, combined with the region’s eroding share of national production, resulted in a more rapid rise in Midwest unemployment relative to the nation. In contrast, during 1990–91, the underlying secular strength of the region’s economy allowed its labor market to continue to gain on its national counterpart and, ultimately, to experience a more fully employed work force. Despite the Midwest’s tight labor markets, its growth of employment and population are not, in general, exceeding the nation’s. The Midwest turnaround has been characterized by a convergence in the pace of employment growth with that of the nation. Job growth in the early 1990s was especially strong relative to the nation, but has now probably eased to a pace that is on par with the nation. Because the region’s work force is approximately at full capacity, any further supranational employment growth cannot reasonably be expected unless population growth increases sharply. Although recent employment gains in the Midwest have been accompanied by population William A. Testa is assistant vice president and team leader of regional programs and Thomas H. Klier and Richard H. Mattoon are senior economists in the Economic Research Department of the Federal Reserve Bank of Chicago. This article is drawn from the Bank’s “Assessing the Midwest Economy” study that began in the fall of 1995. The authors wish to acknowledge the contributions of the numerous participants in this study in helping to frame the issues presented in this article. ECONOMIC PERSPECTIVES Heightened work force participation appears to be reviving Unemployment rates the incomes of Midwest residents. percent Per capita income relative to the 13 nation had dipped sharply from a superior position in the late 1970s to an inferior position by the early 1980s. However, the 10 region’s relative position began to improve in 1991 and slightly Midwest exceeded the national average in 1994 and 1995, the latest year 7 available (see figure 2). Median U.S. household income, measured in constant purchasing power, 4 largely parallels this pattern. 1978 ’81 ’84 ’87 ’90 ’93 ’96 After dipping to parity with the Note: Unless otherwise noted in tables and figures, Midwest refers to the states of the Seventh Federal Reserve District—IL, IN, IA, MI, and WI. nation from 1980 to 1983, MidSource: U.S. Department of Labor, Bureau of Labor Statistics. west income continued on par with the nation through 1994, and is now showing preliminary growth, reflecting a turnaround from a net outsigns of strength relative to the nation. flow in the 1980s to a net inflow in the 1990s, Midwest income continues to flow from the Midwest continues to lag other regions the region’s traditional industries. The Mid(especially most of the Sun Belt) in terms of west remains markedly more concentrated population growth. than the nation in its mainstay industries— This combination of strong employment durable goods manufacturing and agriculture growth and lagging population accounts for (see figure 3). Examining industry composithe marked improvement in the Midwest’s tion at a finer level of detail does little to alter labor force participation relative to the rest this conclusion. of the nation. From a deficit position during the It is not surprising, then, to find that the 1980s, the region’s ratio of employed, aged 16 revival in Midwest job and production growth and above, to population (at 0.65) has surpassed has been led by manufacturing and agriculture. the U.S. average (0.63). The Midwest lost 2.5 percentage points in its share of the nation’s manufacturing employment from 1977 to FIGURE 2 1983 (going from roughly 19.5 Per capita income percent to 17 percent). It has since index, U.S.=100 regained 2 percentage points and 108 the rate of gain has accelerated in the 1990s. Manufacturing industries such as autos and steel have reconcentrated in the Midwest. 104 For example, the region had 31 Midwest auto plants in 1996, compared with 27 in 1979. Although nine Midwest auto plants closed be100 tween 1979 and 1996, 13 new plants opened. Rural areas have benefited from both the rising manufac96 1971 ’75 ’79 ’83 ’87 ’91 ’95 turing tide and the recovery of Source: U.S. Department of Commerce, Bureau of Economic Analysis. production agriculture. Over the past 15 years, rural growth in FIGURE 1 FEDERAL RESERVE BANK OF CHICAGO 3 in-migration. Midwestern metro areas experienced a modest outPersonal income by industry, Midwest/U.S. migration from 1990 to 1994, location quotients even as employment continued 0.0 0.5 1.0 1.5 2.0 to grow. Some workers may be Agriculture choosing to reside in adjacent 1994 1977 Construction counties, while commuting to jobs in metro areas. Manufacturing Metro areas continue to be Nondurable manufacturing magnets for jobs, but the nature Durable manufacturing of jobs is changing: Many Transportation midwestern metro areas are and public utilities successfully transforming from Wholesale trade manufacturing centers to service Retail trade centers (see table 1).6 Important Finance, insurance, and real estate industries of the service and Personal information economy of the services Business U.S. average 1990s include producer services, services such as management consulting, Government advertising, accounting, and Note: A quotient of 1.0 indicates the same relative importance of an industry business and legal, as well as for the Midwest as for the U.S. as a whole. trade, travel, and financial serSource: U.S. Department of Commerce, Bureau of Economic Analysis. vices. Midwest metro areas have been very successful in attracting these industries and, in some cases, developing manufacturing jobs has outpaced metropolitan a service industry niche, such as sports-oriented growth. Meanwhile, the recovery of the farm travel centers (Indianapolis), convention tourism sector from the debt overhang and sagging mar(Appleton–Oshkosh), health services and insurkets of the late 1970s and early 1980s has lifted ance (Peoria), air freight/air maintenance cenfarmland prices. Grain prices are high by historters (Indianapolis), financial services/insurance ical standards and demand from developing (Des Moines), automotive R&D (Detroit), and countries has buoyed world markets for grains, convention–business meeting centers (Chicago). meat products, and some processed foods. Industrial restructuring has differed by size Emergence of these goods-producing of metro area, with large metro areas tending industries in rural areas—especially manufacto transform to a greater degree away from turing—has translated into an improvement in manufacturing toward business and financial population growth and a turnaround from net services.7 Manufacturing losses in large metro out-migration to net in-migration.3 In Midwest rural counties, population declined by 2.2 percent areas, especially core counties, have been sharp. from 1980 to 1990, with 70 percent of counties Smaller metro areas have tended to lose out experiencing declines in absolute terms. From on some business-oriented services such as 1990 to 1994, rural Midwest counties recorded financial service industries, while picking up population gains of 2.4 percent, with 74 percent the slack, in general, as preferred manufacturreporting gains during this period. 4 ing locations. As a result, the Midwest’s ecoDespite the dramatic swings in rural fornomic recovery, as reflected in relatively low tunes, the Midwest’s and nation’s population unemployment rates, has been pervasive across continues to shift to metropolitan (metro) areas, metro areas. much as it has done throughout this century. Timing and depth By 1994, metro areas’ share of population was From 1947 to 1987, the Midwest’s (defined approaching 80 percent in the nation and 76 in this case as Illinois, Indiana, Michigan, Ohio, percent in the Midwest. However, the pace and Wisconsin) share of national manufacturing of the shift appears to be slowing considerably declined from 30 percent to 22.1 percent.8 To a in the 1990s, harkening back to the rural–urban large extent, this reflected a natural process of turnaround of the 1970s.5 In metro areas, the population deconcentration within the continental natural population increase continues to dominate FIGURE 3 4 ECONOMIC PERSPECTIVES After a languid recovery in 1983, the Midwest economy Personal income derived from labor and proprietor showed some vigor in 1984, earnings (indexes of concentration) supported by strengthening auto Midwest/U.S. demand. However, in 1985 and 1969 1977 1985 1994 1986, the dollar value of overall export sales from the region Manufacturing remained flat despite depreciation Large MSAs 1.31 1.35 1.29 1.30 of the value of the dollar against Core counties 1.28 1.34 1.22 1.20 currencies of trading partners.10 Medium MSAs 1.52 1.63 1.70 1.70 Small MSAs 1.41 1.48 1.55 1.68 Domestic markets for capital Nonmetro 1.10 1.19 1.34 1.62 goods, other than computer-related purchases from coastal regions, FIRE Large MSAs 1.00 1.09 1.14 1.15 also continued to disappoint. Core counties 1.12 1.22 1.37 1.32 Beginning in 1987, the MidMedium MSAs 0.70 0.71 0.70 0.78 west’s capital goods sectors began Small MSAs 0.71 0.73 0.71 0.72 to recover late in the expansion Nonmetro 0.60 0.62 0.51 0.47 and exports began to grow. In the Business services agricultural sector, farm equipment Large MSAs 1.06 1.07 1.16 1.13 purchases and exports began to Core counties 1.13 1.02 1.01 0.97 show some life and balance sheets Medium MSAs 0.52 0.58 0.63 0.75 Small MSAs 0.43 0.51 0.54 0.61 began to strengthen. Nonmetro 0.49 0.53 0.44 0.40 Many observers believed that the shakeout of the early 1980s Note: An index value of 1 signals the same importance of an industry for the Midwest as for the U.S. That is, an index value of 2.0 could be was so severe that it destroyed a obtained from a 40 percent share of manufacturing in Midwest metro large portion of the most inefficient areas divided by a 20 percent share of manufacturing in U.S. metro areas. Source: U.S. Department Of Commerce, Bureau of Economic Analysis, and antiquated physical capital Regional Economic Information System. stock. Accordingly, the renewed strength of the Midwest in the second half of the decade was interpreted as an inevitable bounce-back in U.S. For example, the Southeast developed production and productivity, albeit from a much manufacturing industries as its work force was lower baseline level.11 Even as other regions released from agriculture, and improvements in technology, infrastructure, and transportation such as the Southwest and New England began opened up previously isolated areas in many to experience economic setbacks, few believed parts of the country. that these setbacks would continue for very Recessionary periods were particularly long. Although the high-tech industry, the difficult for the Midwest because of its condarling of the decade, was toppling in New centration in capital goods and consumer England, along with defense-related industries durables, which were most vulnerable to a there and elsewhere, many believed that a falloff in demand.9 Furthermore, the region’s bounce-back in high technology was only a technology and physical stock of capital matter of time, and that the Midwest was at tended to be of earlier vintage (and often a disadvantage because the fastest growing lower efficiency) than in other regions of the sectors were almost nonexistent in the region.12 U.S. and abroad. Consequently, when demand It also took some time before the extent of overslackened, it was more cost-effective to conbuilding in real estate in other regions could be tinue remaining production at newer (lower fathomed. The coastal regions and parts of the cost) plants elsewhere. During the 1979–83 Southwest struggled through the overbuilding period, as the nation passed though two recesand savings and loan debacles to a significantly sions in quick succession, the Midwest lost greater extent than the conservative and still over one-fifth of its manufacturing work force “shell-shocked” Midwest. Through the distorted at the same time that the rural agricultural lens of these events, fundamental changes undereconomy experienced its worst times since the lying a sustained turnaround of the Midwest Great Depression. were difficult to distinguish. TABLE 1 FEDERAL RESERVE BANK OF CHICAGO 5 Today, it is evident that the signs of strength in the Midwest economy of the mid- to late 1980s were more than the anticipated snap-back from the restructuring of the early 1980s. Much of the adjustment had taken place by 1985 and the transitory shocks in other regions have significantly dissipated. Yet, the pace of economic growth in the Midwest remains strong and capacity utilization remains high. The Midwest economy has been changing from within during the past ten to 15 years, and these changes have been supported by favorable external conditions and trends. External conditions External factors in the Midwest’s economic turnaround include technological and organizational changes in the automotive industry, which have favored its reconcentration in the midsection of the nation; the geographic pattern of federal defense spending; declining real energy prices, important both as an input to the region’s industries and as a determinant of demand for its products; and, from the mid1980s until recently, a declining dollar, which improved the international competitiveness of the region’s companies. Changing geography of the auto industry U.S. auto assembly plants have tended to reconcentrate in the Midwest over the 1980s and 1990s. Auto supplier plants had tended to disperse over the three decades to 1990, but this trend appears to be reversing during the 1990s (see figure 4) as more technologically advanced and innovative automotive parts and services providers continue to locate in the Midwest. The reconfiguration of auto assembly, the continued preference of supplier R&D operations to locate close to Detroit, plus evidence of spatial clustering of tier 1 supplier plants around their assembly plant customers suggest a strengthening of agglomeration effects in the auto industry. As discussed in Rubenstein (1996), the reconcentration of auto assembly has resulted from broad changes in the industry’s product mix that, consistent with neoclassical 6 location theory, have changed the economics of plant location in favor of the midsection of the country. The costs of distributing the final product to the customer have always been important in deciding the location of auto assembly plants. Henry Ford opened far-flung branch assembly plants to produce identical Model T cars closer to the population centers outside the Midwest; it was cheaper to ship parts to branch assembly plants than to ship finished automobiles across the country from a centrally located assembly plant. Soon, General Motors and Chrysler emulated that strategy. However, by the 1960s the proliferation of car and truck models meant that location strategy was no longer optimal. The number of different car and truck models sold in the U.S. increased eightfold, from 30 in 1955 to 241 in 1995, while sales only doubled from about eight million units to about 16 million in 1995. With reduced output per individual model, the entire output would best be produced at one plant. Consequently, the geographic argument for an interior location became compelling; that way the company could minimize the cost of distributing the output to a national market. As a result, during the past 16 years auto producers have opened assembly plants in the interior, especially FIGURE 4 Auto supplier plants by region and start-up date percent 80 60 Midwest 40 Southeast 20 West Northeast 0 prior to 1950 ’50-59 ’60-69 ’70-79 ’80-89 ’90-95 Note: The Midwest refers to IL, IN, MI, OH, and WI. The Southeast refers to AL, AR, FL, GA, KY, LA, MS, NC, SC, TN, TX, VA, and WVA. The Northeast refers to CT, DE, ME, MD, MA, NH, NJ, NY, PA, RI, and VT. The West refers to all other states. Source: James Rubenstein, “The evolving geography of production—Is manufacturing activity moving out of the Midwest? Evidence from the auto industry,” Assessing the Midwest Economy Working Paper Series , No. SP-3, Federal Reserve Bank of Chicago, 1996. ECONOMIC PERSPECTIVES along the I-65/I-75 corridor, and closed coastal plants. While freight costs can account for the reconcentration of auto production in the Midwest, variables such as the local labor climate, access to highways, and general costs of doing business influence the selection of particular communities or sites. Federal spending patterns FIGURE 5 Per capita federal expenditures percent of U.S. per capita levels 110 U.S. 100 IA 90 IL Historically, the Midwest has not fared well relative to other 80 regions in terms of receiving money WI IN from Washington. Measured as a MI percentage of U.S. per capita lev70 1985 ’87 ’89 ’91 ’93 ’95 els, federal expenditures were beSource: Kerry Sutten, “Federal spending in the Northeast and Midwest: low the national average in each of Fiscal 1995,” Northeast–Midwest Institute, Washington, DC, June 1996. the five states of the Seventh Federal Reserve District from 1985 to 1995, with the exception of Iowa in 1988 (see figure 5). First, because of the small and highway infrastructure funds) than many concentration of defense-related industries, other regions. According to data compiled by federal procurement spending in the region is the Northeast–Midwest Institute, in fiscal 1994 particularly weak. For example, in fiscal 1995, federal transfers on average represented 27.7 Illinois ranked 47th and Indiana and Wisconsin percent of total state budgets in the U.S.14 tied for 46th on per capita military procurement Federal transfers comprised 25.8 percent of expenditures. In addition, the relatively small the Illinois budget, 28.4 percent of the Indiana number of military bases in the region keeps budget, 25.7 percent of the Iowa budget, 25.4 military spending on wages and salaries signifpercent of the Michigan budget, and 24.4 percent icantly below the U.S. average. Spending by of the Wisconsin budget. the federal government for grants, federal salaries Energy and wages, and direct payments in the Midwest is Delivered prices of all major fuels have generally below average on a per capita basis. declined in the Midwest since the early to Over the past ten years, however, regions mid-1980s (Bournakis, 1996). Despite recent that depended heavily on federal dollars have price run-ups, national real gasoline prices are been particularly affected by program cuts. currently lower than in 1967 and 25 percent Defense spending reductions and the difficulty lower than their peak in the latter half of the of converting defense industries to nondefense 1970s. At that time, high petroleum-based functions have damaged economies in Califorfuel prices exerted a significant drag on Midnia and New England. Figures on U.S. military west industry and hampered sales of domestic procurement spending from 1985 to 1996 and automakers. Recently, midwestern energy projected to 2002 demonstrate the spending prices have been edging down relative to boom in states with concentrations of defensenational energy prices (see figure 6). Why related industries in the 1980s ($80 billion in these prices have eased is not clear but may 1991) and the rapid decline in expenditure be due to external developments, such as levels in the 1990s (an estimated $40 billion in deregulation of the U.S. natural gas market 1998). A study of the Chicago economy sugsince the mid-1980s, or regional issues, such gests that even those industries in the Midwest as state–local tax and regulatory policies. that have traditionally done business with the federal defense establishment may convert to Exports civilian products relatively easily.13 Exports now account for 13 percent of The Midwest is also less reliant on federal U.S. gross domestic product, compared with 8 transfers (primarily Medicaid, social welfare, percent in 1987 and 5 percent in 1971. From FEDERAL RESERVE BANK OF CHICAGO 7 markets in Asia and South America.16 Ratio of Midwest to U.S. fuel prices Currency swings since the ratio dollar’s peak in early 1985 are 1.18 often cited in the popular press as having boosted midwestern 1.12 exports and shielded domestic Natural gas Distillate fuel markets from displacement by foreign imports. The Midwest’s 1.06 Coal share of Big Three auto production has increased since 1991. 1.00 Thanks to the reconcentration of domestic automakers and the Motor gasoline 0.94 presence of Japanese automakers, Electricity the region’s share of domestic 0.88 car production has climbed from 1983 ’85 ’87 ’89 ’91 ’93 45 percent in 1981 to more than Source: Athanasios Bournakis, “Energy and environmental issues for the Midwest economy,” Assessing the Midwest Economy Working Paper Series, 56 percent.17 However, the No. SP-5, Federal Reserve Bank of Chicago, 1996. drop-off in the dollar’s value was completed by 1987. Over the past nine years, aggregate trade-weighted dollar indexes suggest that the 1992 to 1995, exports from the Midwest grew currency-influenced terms of trade have reeven more dramatically than exports from the mained mostly flat, even as exports have conU.S. as a whole (see figure 7). Exports from tinued to climb. Moreover, recent research by the region’s telecommunications, farm machinHervey and Strauss (1996) suggests that the ery, construction machinery and equipment, dollar has appreciated rather than depreciated machine tools, and specialized capital goods against currencies of nations to which the Midsectors, as well as the agriculture sector, have west exports.18 grown rapidly during the past ten years to meet growing demand from developing markets Changing how we do business worldwide.15 The lion’s share of future trade (internal adjustments) expansion is expected to derive, not from trade The Midwest’s constancy in line of busiwith developed nations, but from emerging ness and evidence of productivity gains driven by internal private and public sector actions suggest that internal FIGURE 7 factors have also been important sources of regional revival. In Total exports particular, midwestern industry index, 1990=100 has adopted new technologies 160 and modes of business operation, and the region’s relative cost position has improved. The pubMidwest 140 lic sector has facilitated regional competitiveness by prudent taxaU.S. tion and spending policies, by focusing spending on value120 producing services and public infrastructure, and, more recently, by adopting innovative delivery of public services. So too, the 100 region’s “institutional capital”— 1990 ’91 ’92 ’93 ’94 ’95 public and private organizations, Source: Massachusetts Institute for Social and Economic Research. including universities, research FIGURE 6 8 ECONOMIC PERSPECTIVES centers, and business and civic organizations— proved responsive in the face of economic crisis. setting is the U.S. auto industry. Successful auto assembly operations have been transplanted to the U.S. environment by companies such Technology and organization as Toyota, Honda, and Mitsubishi. In some There is substantial evidence that the Midcases, existing assembly plants, such as GM’s west has changed the way it does business—its NUMMI venture with Toyota in California, organization, mode of operation, and technology. have been transformed through organization As discussed in Klier (1996), implementation and technology alone. of best manufacturing practices, notably lean The extent to which this experience is manufacturing technologies, has helped revitalize characteristic of manufacturing in general Midwest manufacturing. Lean manufacturing, was addressed in two large-scale studies.19 which gained widespread attention in the early Both Statistics Canada (Baldwin, Diverty, and 1980s, combines aspects of both craft and mass Sabourin, 1988) and the U.S. Census Bureau production, ranging from teamwork on the (1988 and 1994) administered surveys of manshop floor, to emphasis on low inventory and ufacturing technologies to measure the extent flexible production equipment, to close relaand type of advanced manufacturing technolotionships with suppliers. The most familiar gies used in their respective country’s manufacturing plants. Both surveys found that the application of TABLE 2 advanced manufacturing technoloApplication of some advanced technologies gies was widespread across plants and industries, typically with mulFMC/ CAD/ Interco. FMS CAE network tiple technologies applied per (----percent of plants using----) establishment (see table 2). These results indicate that advanced Plant employment manufacturing techniques are 20–99 7.6 49.5 12.0 reshaping manufacturing on a 100–499 21.4 76.4 28.4 broad scale. In the Midwest, more 500+ 40.4 87.2 47.1 concentrated in manufacturing Age of plant than any other region, these techLess than 5 years 13.4 63.5 15.0 nological advances have tended 5–15 13.3 62.0 18.0 16–30 13.4 64.4 20.5 to boost the economy. Greater than 30 15.2 63.1 22.0 At the same time, the region’s industries are outside those (mostly Major industrial groups Fabricated metal products 9.5 46.5 16.7 defense) sectors that require both Industrial machinery a change in product mix and a and equipment 11.8 64.1 15.4 transformation in technology. Electronic and other Regions specializing in declining electric equipment 17.0 64.2 21.9 industries, such as defense-oriented Transportation equipment 15.5 53.9 23.4 manufacturing, must change not Instruments and related only how business is conducted products 14.2 65.5 15.3 but the entire product mix. To Notes: The table reports information on three of the 17 advanced date, for several regions that commanufacturing technologies surveyed. They are defined as follows: pete with the Midwest, the barriers Flexible manufacturing cells and systems (FMC/FMS): two or more of changing both “how” and machines with automated material handling capabilities controlled by computers or programmable controllers, capable of single/multiple “what” have been too high to bring path acceptance of raw material and single/multiple path delivery of finished product. about the resurgent experience of Computer-aided design and engineering (CAD/CAE): use of computers the Midwest. for drawing and designing parts or products and for analysis and testing of designed parts or products. Intercompany computer network (Interco. network): use of network technology to link subcontractors, suppliers, and/or customers with the plant. Source: U.S. Department Of Commerce, Bureau of the Census, Current Industrial Reports: Manufacturing Technology: Prevalence and Plans for Use, 1994, tables 4D and 4E. FEDERAL RESERVE BANK OF CHICAGO Costs of business operation The neoclassical view in economics suggests that firm location is significantly driven by the search for low costs of operation. 9 FIGURE 8 Real per worker earnings thousands of 1994 dollars 29 Midwest 28 U.S. 27 26 25 1980 ’82 ’84 ’86 ’88 ’90 ’92 Source: U.S. Department of Commerce, Bureau of Economic Analysis. Labor costs commonly comprise the largest share of operating costs to business enterprises. This implies that capital investment flows toward regions with low wage costs and that job openings grow in tandem with capital investment. In many instances, labor does not migrate, as might be expected, toward high-wage areas, because job openings are absent due to rigid wages and, perhaps, institutional features such as unionization.20 As a result, economies with low wage costs can experience economic growth of capital and labor. Evidence from the past ten to 15 years is consistent with this theory in partly explaining the Midwest turnaround. The Midwest has long been reputed as a high-wage locale, especially for manufacturing. But over the past ten to 15 years, workers in the Midwest have apparently eased their wage demands relative to those of their national counterparts. Real per worker earnings approached national levels from 1980–82 and continued to converge throughout the 1980s (see figure 8).21 While these figures are merely suggestive of labor costs, changes in the level of hourly wages of workers in the manufacturing sector point in the same direction (see table 3).22 Adjusting for differences in industry mix, Midwest manufacturing wages eased from 17 percent above national levels in the early 1980s to a 13 percent ’94 premium in the 1990s. As mentioned earlier, energy prices in the region have also eased relative to national prices, including prices of coal and natural gas, which the region consumes in greater proportion than the nation.23 The region has also taken measures toward greater energy conservation and efficiency; at the same time, industry composition has shifted away from energy-intensive sectors. Today, the Midwest consumes much less energy relative to gross state product than 20 years ago (see figure 9). Thus, Brown and Yücel (1995) suggest the region would experience dramatically milder responses to potential oil price shocks (less than half of 1980 levels). The public sector Some analysts suggest that the Midwest has assisted its own revival through judicious fiscal policies. The region followed a conservative fiscal path characterized by minimal increases in levels of taxation (even during the 1990–91 recession) and conservative spending policies. At the same time, public spending was generally above TABLE 3 national levels in areas that are Index of relative wages in manufacturing: seen as contributing to economic Midwest versus U.S. growth, such as education and 1979 1983 1989 1993 1995 highway expenditures, while spending was below national Illinois 1.09 1.10 1.07 1.03 1.02 levels in areas less associated Indiana 1.16 1.14 1.12 1.12 1.13 with economic growth, such as Iowa 1.16 1.14 1.03 1.04 1.03 government administration, corMichigan 1.30 1.32 1.29 1.31 1.32 rections/prisons, and welfare exWisconsin 1.09 1.11 1.03 1.04 1.03 penditures (see figure 10). Midwest 1.17 1.17 1.13 1.13 1.13 It is hard to say to what extent Source: U.S. Department of Labor, Bureau of Labor Statistics. this behavior contributed to the 10 ECONOMIC PERSPECTIVES to be influential to regional growth. In her 1995 book World Midwest energy consumption and gross state product Class, for example, Rosabeth quadrillion Btu trillions of 1992 dollars Moss Kantor suggests the places 20 1.00 that succeed in the new global economy often do so because Gross state product they have created and supported (right scale) organizations, their so-called 15 0.75 institutional capital.26 These organizations are often found in the not-for-profit sector and 10 0.50 include public–private partnerEnergy consumption (left scale) ships and councils, nonprofit organizations of business leaders, public–private development 5 0.25 1971 ’73 ’75 ’77 ’79 ’81 ’83 ’85 ’87 ’89 ’91 ’93 councils, foundations, trade assoSource: See figure 6 ciations, chambers of commerce, research centers at local universities, and research institutes. The Midwest has fostered a Midwest’s economic revival, since the relationrich endowment of organizations that form its ship between public spending/taxation and ecoinstitutional capital stock, contributing to a varinomic growth has not been definitively demonety of regional economic development efforts. strated. Beneficial effects between state–local In the case of state and local economic develgovernment fiscal health and private sector opment planning, for example, communityeconomic growth run in both directions, thereby based organizations and local business associamaking it difficult to discern cause from effect. tions provide important information to publicThe Midwest’s current prosperity is evident sector decisionmakers on which efforts and in its state and local sector, as midwestern state programs work best and often promote soluand local governments have, in general, rebuilt tions that fall outside of narrow political their budget balances and improved their fiscal boundaries.27 For example, several multistate position. While the national average state fund efforts have addressed environmental challengbalance (as a percentage of state expenditures) es in the Midwest. One such effort is aimed at was slightly more than 5 percent in 1996, Indiunderstanding the atmospheric science and ana recorded a fund balance of 20 percent; fashioning compliance solutions to the ozoneIowa, 15 percent; and Michigan, 13 percent. related ambient air quality standards of the The recessionary period of the early 1980s Clean Air Act Amendments.28 Another arises reduced the region’s fiscal capacity and induced from the Great Lakes Water Quality Initiative, a states to strain their capacity to fund public basin-wide approach to reducing toxic contamispending. However, Midwest states are seen nation of the Great Lakes system.29 Proactive to have begun easing the strain on their fiscal development initiatives at a multistate level capacity by the mid-1980s. 24 The fiscal expehave been no less common, including tourism rience of District states followed the same and export promotion and efforts to broaden break with the past that has characterized the skill standards and certification.30 economic performance of the region. Unlike FIGURE 9 previous recessions which had usually forced dramatic tax increases in the region, the national recession of 1990–91 had a relatively shallow impact.25 Institutional capital Nonprofit institutions and organizations that engage in economic growth and development policies and programs are often believed FEDERAL RESERVE BANK OF CHICAGO What does the future hold for the Midwest? This article has identified developments that have restored the region’s luster. With ten to 15 years hindsight it is evident that, while business cycle timing and external factors have been very important, profound changes have taken place in the way midwestern businesses and governments compete and conduct their 11 affairs. Although it is impossible to discern the relative contribution of these forces with a great deal of precision, it is clear that the Midwest is partly responsible for its own recovery. The challenges and opportunities facing the region can be discussed in terms of two important characteristics of this economic recovery: the resurgent strength of the region’s FIGURE 10 State and local public sector spending per $1,000 personal income, 1993 Total state/local spending Total education expenditures dollars 250 dollars 100 200 75 150 50 100 25 50 0 IL IN IA MI WI Midwest Rest of U.S. Corrections spending 0 IL dollars 40 6 30 4 20 2 10 IL IN IA MI WI Midwest Rest of U.S. Administration spending dollars 12 0 6 6 3 3 IN WI Midwest Rest of U.S. IA IL IN IA MI WI Midwest Rest of U.S. IA MI WI Midwest Rest of U.S. MI WI Midwest Rest of U.S. dollars 12 9 IL MI Interest on debt 9 0 IA Public welfare expenditures dollars 8 0 IN MI WI Midwest Rest of U.S. Total state/local debt 0 IL IN Highway expenditures dollars 240 dollars 25 20 180 15 120 10 60 0 5 IL IN IA MI WI Midwest Rest of U.S. 0 IL IN IA Source: U.S. Department of Commerce, Bureau of the Census, “State and local government finance estimates, by state,” Internet at www.census.gov/govs/www/index.html, March 17, 1997. 12 ECONOMIC PERSPECTIVES mainstay industries and the increase in labor force participation. First, production agriculture and manufacturing exhibited startling resilience during this period by improving productivity, regaining market share, and aggressively targeting export markets. In this regard, there is reason for optimism. Following the shocks of the 1980s, the region’s firms will most likely stay on their guard with regard to changing technology and the benchmarks of global competition. So too, much of the future growth in export markets is expected to derive from developing countries, whose needs for the region’s products—capital goods and agricultural products—are expected to continue to grow. Second, the region’s labor force faces an important challenge. Will work force numbers and skills be adequate to sustain the growth that the region has experienced in the 1990s? In achieving recent growth, the region has drawn from both new work force entrants and adults who were unemployed or underemployed. However, this pool of skilled workers is now showing signs of strain. Currently tight labor markets suggest the need for midwestern workers who no longer are seen as disproportionately expensive relative to other regions. Labor supply may become further strained in the years ahead, because the work force in the Midwest is reported to be older than in the nation, so loss of workers through retirement is likely to be relatively high in the region.31 Insofar as the region’s two industry mainstays—agriculture and manufacturing—continue to shed labor, a lower level of growth in the work force may be needed. If so, natural population growth, upgrading of skills, and work force innovations to bring more people into the labor force may suffice. For example, job networking of workers from the inner city to the suburbs and moving welfare recipients into the workplace may ease labor market pressures. Furthermore, some workers may decide to defer retirement, especially if the rewards of working become more attractive.32 In-migration of population could provide another “release valve” to labor market pressures. In most parts of the Midwest, the cost of living does not present a barrier to in-migration. The median home price in the Midwest is the lowest of any of the four regions defined by the National Association of Realtors; the association’s index of housing affordability is also more favorable here than in other regions. However, significant in-migration to the region has not occurred to date, although out-migration has been stemmed. Moreover, regions from which the Midwest might expect to draw workers, such as the West Coast and the Northeast, are experiencing labor market tightening. An alternative approach to easing labor shortages is the upgrading of skills of the existing and emerging work force. This is a preferred approach, because higher skills tend to be rewarded in the form of higher wages and income for Midwest residents. For this to come about, workers must act to acquire needed skills or credentials and policymakers must act to create publicly assisted training, educational, and job-assistance programs or increase the effectiveness of existing programs. Young adults continue to enter the work force, but there is concern that some training/ educational programs have fallen into disrepair or were abandoned during the early 1980s, when one in five manufacturing jobs evaporated in the region and growth in other sectors stagnated. As a result, a renewed push is needed to reestablish school-to-career and other programs in selected skill areas. In the absence of such initiatives, the region’s residents could miss opportunities for better and higher paying jobs and the region’s businesses and property owners could miss significant income-generating opportunities. NOTES 1 For a more detailed summary of this work, see Testa, Klier, and Mattoon (1997). Unless specified otherwise, Midwest refers to the states of the Seventh Federal Reserve District—Illinois, Indiana, Iowa, Michigan, and Wisconsin. 2 See Allardice and Bergman (1996). 3 See Johnson (1996). FEDERAL RESERVE BANK OF CHICAGO 4 Surprisingly, net migration into the rural Midwest exceeded population gains derived from natural increase, that is, births minus deaths. Throughout this century, population gains in rural counties have generally been realized through natural increase concurrent with net out-migration of young adults. By the 1990s, the resulting aging of the population, coupled with in-migration, resulted in a 13 notable reversal; in-migration gains were leading those achieved by natural increase in rural counties of the nation and the Midwest. from a favorable expansion and pattern of expansion in foreign markets and from improving productivity. 19 5 See Johnson, op. cit. 6 See Groshen and Robertson (1993). 7 For further industry-specific analysis, see Testa (1992). Other authors note the further spatial division of labor by size of metro area according to industries or facilities characterized by routinized or “back office” operations and those engaged in “command and control,” operations such as corporate headquarters or highly specialized business and legal services. For further discussion, see Atkinson (1996) and Federal Reserve Bank of Chicago (1996a). 8 Kim (1996). 9 Howland (1984). Baldwin, Diverty, and Sabourin (1994); U.S. Department of Commerce, Bureau of the Census (1994). 20 A popular exposition of the neoclassical theme with regard to the recent Midwest experience can be found in Swonk (1996). For a discussion of the possible effects of one such institutional feature, that is, right-to-work laws, see Holmes (1995); also Kendix (1990). 21 Such figures are merely suggestive and not definitive, assuming, for example, the shift between part-time and full-time workers across regions does not distort the findings, and that the findings are not similarly distorted by regional differences in labor force growth. 22 10 Languishing total dollar volume of exports is somewhat expected in this instance because rising physical quantities of exports may be slight or insufficient to make up for lower dollar prices per physical unit, that is, the “J-curve” effect. Of course, other forces, especially changing economic growth of export destinations, also determine export sales. The evidence is also consistent with falling wages having resulted from a shrinking economy (and shrinking labor demand). The early period strongly suggests the falling wages were caused by loss of manufacturing and attendant high-paying jobs and by excess supplies of willing workers. It is unclear, as yet, whether lower wages have helped revive investment and employment in the region. 23 See Bournakis (1996). 24 11 A more farsighted view of the region is attributed to Annable (1985) and Swonk (1991). 12 For example, Browne (1983). 13 Research completed by Philip Israilevich on defenserelated businesses in the metro Chicago economy found that less than 1 percent of the total output of goods and services in the Chicago economy was related to military procurement expenditures in 1987 (during the height of the military spending boom). Furthermore, electrical machinery, business services, food, and control instruments were the industry groups that accounted for 73 percent of the procurement funds that the metro economy received. These industry groups can serve civilian as well as defense markets without the difficult transition associated with prime contractors, such as ship builders, plane manufacturers, or weapons and munitions firms. See Israilevich and Weiss (1992). 14 Sutten (1996). Fiscal capacity measures of a state are constructed by comparing a state’s per capita tax base to the nation’s, aggregated across all commonly used tax bases, for example, sales, income, and property value. See U.S. Advisory Commission on Intergovernmental Relations (1989). 25 See Mattoon and Testa (1992). 26 See Kantor (1995). The hypothesis that long-developed regions will necessarily have an advantage in sustaining growth during a period of adversity or shock remains contentious. For an opposing hypothesis, see Kendix, op. cit. For a balanced and wide-ranging discussion of the role of such institutions in economic development, see Bonser (1995). For an in-depth discussion of state and local government development initiatives and concepts in the 1980s, see Eisinger (1988). 27 For example, see Ameritrust Corporation (1994), Wisconsin Strategic Development Commission (1985), Iowa Business Council and Federal Reserve Bank of Chicago (1987), and the Commercial Club of Chicago (1984). 15 See Aguilar and Singer (1995) and David Walters’ comments in Federal Reserve Bank of Chicago (1996d). 16 Walters, ibid. 28 To understand and facilitate compliance with urban ozone regulations, the Lake Michigan Air Directors Consortium has been studying regionwide atmospheric chemistry; see Gerritson (1993). 17 The share of production has slipped during 1996. Some analysts partly attribute this slippage to the climbing value of the dollar versus the yen. See Meredith (1997). 29 The Council of Great Lakes Governors has been active in shaping the new environmental guidance for protecting the basin’s water quality. See DRI/McGraw Hill (1993). 18 Hervey and Strauss (1996) constructed foreign currency measures against the dollar that are specific to the Midwest’s export composition. They found that the Midwest’s export success has run counter to deleterious trends in the exchange currencies of the region’s major export destinations. It is likely that the region’s export success derives 14 30 For a review of such efforts, see McNulty (1991). 31 McAlinden, Smith, and Cole (1995). 32 Judy and D’Amico (1997). ECONOMIC PERSPECTIVES REFERENCES Aguilar, Linda M., and Mike A. Singer, “Big emerging markets and U.S. trade,” Economic Perspectives, Federal Reserve Bank of Chicago, Vol. 19, No. 4, July/August 1995, pp. 2–15. Allardice, David R., and William Bergman, “The state of the region,” presentation at the workshop, The Midwest Economy: Structure and Performance, Federal Reserve Bank of Chicago, February 13, 1996. Commercial Club of Chicago, Make No Little Plans: Jobs for Metropolitan Chicago, Chicago: Commercial Club of Chicago, 1984. DRI/McGraw Hill, “The Great Lakes Water Quality Initiative: Cost-effective measures to enhance environmental quality and regional competitiveness,” report prepared for the Council of Great Lakes Governors, Chicago, 1993. Ameritrust Corporation, “Choosing a future for Mid-America: Strategies for revitalizing the midwestern economy,” Detroit, MI, report, 1994. Eberts, Randall, “Highway infrastructure: Policy issues for regions,” Assessing the Midwest Economy Working Paper Series, No. SL-2, Federal Reserve Bank of Chicago, 1996. Annable, James E., “The Midwest economic miracle,” First National Bank of Chicago, internal research paper, Autumn 1985. Eisinger, Peter K., The Rise of the Entrepreneurial State, Madison: University of Wisconsin Press, 1988. Atkinson, Robert, “Technology and the future of metropolitan economies,” Assessing the Midwest Economy Working Paper Series, No. MA-4, Federal Reserve Bank of Chicago, 1996. Federal Reserve Bank of Chicago, “Midwestern metropolitan areas: Performance and policy,” Assessing the Midwest Economy Workshop Summary, No. 1, Federal Reserve Bank of Chicago, 1996a. Baldwin, John, Brent Diverty, and David Sabourin, “Technology use and industrial transformation: Empirical perspectives,” paper presented at the conference, Technology, Information and Public Policy, held by Statistics Canada at Queen’s University, November 1994. Bonser, Charles F., ed., Proceedings: The Role of NGOs in Economic Development, Bloomington, IN: Indiana University, Institute for Development Strategies, 1995. Bournakis, Athanasios, “Energy and environmental issues for the Midwest economy,” Assessing the Midwest Economy Working Paper Series, No. SP-5, Federal Reserve Bank of Chicago, 1996. , “The changing rural economy of the Midwest,” Assessing the Midwest Economy Workshop Summary, No. 3, Federal Reserve Bank of Chicago, 1996b. , “Work force developments: Issues for the Midwest economy,” Assessing the Midwest Economy Workshop Summary, No. 4, Federal Reserve Bank of Chicago, 1996c. , “Global linkages to the Midwest economy,” Assessing the Midwest Economy Workshop Summary, No. 6, Federal Reserve Bank of Chicago, 1996d. Brown, P.A., and Mine Yücel, “Energy prices and state economic performance,” Economic Review, Federal Reserve Bank of Dallas, No. 2, 1995, pp. 13–21. Gerritson, Stephen, “The status of the modeling of ozone formation and geographic movement in the Midwest,” in Cost Effective Control of Urban Smog, W. Testa, R. Kosobud, and D. Hanson (eds.), Chicago: Federal Reserve Bank of Chicago, 1993, pp. 32–42. Browne, Lynn E., “Can high tech save the Great Lakes states?,” New England Economic Review, Federal Reserve Bank of Boston, November/December 1983, pp.19–33. Groshen, Erica, and Laura Robertson, “Are the Great Lakes cities becoming service centers?,” Economic Commentary, Federal Reserve Bank of Cleveland, June 1, 1993. FEDERAL RESERVE BANK OF CHICAGO 15 Helms, Jay, “The effect of state and local taxes on economic growth: A time series–cross section approach,” Review of Economics and Statistics, Vol. 67, 1985, pp. 574–582. Hervey, Jack, and William Strauss, “A regional export-weighted dollar: A different way of looking at exchange rate changes,” Assessing the Midwest Economy Working Paper Series, No. GL-2, Federal Reserve Bank of Chicago, 1996. Holmes, Thomas J., “The effects of state policies on the location of industry: Evidence from state borders,” Federal Reserve Bank of Minneapolis, working paper, No. 205, 1995. Howland, Marie, “Age of capital and regional business cycles,” Growth and Change, April 1984, pp. 29–37. Israilevich, Philip, and David Weiss, “The effects of defense cuts on the Chicago economy,” Chicago Fed Letter, Federal Reserve Bank of Chicago, No. 53, January 1992. Iowa Business Council and the Federal Reserve Bank of Chicago, The Iowa Economy: Dimensions of Change, Chicago: Federal Reserve Bank of Chicago, 1987. Johnson, Kenneth M., “Recent nonmetropolitan demographic trends in the Midwest,” Assessing the Midwest Economy Working Paper Series, No. RE-1, Federal Reserve Bank of Chicago, 1996. Judy, Richard W., and Carol D’Amico, Workforce 2020: Work and Workers in the 21st Century, Indianapolis, IN: Hudson Institute, 1997. Kantor, Rosabeth Moss, World Class, New York: Simon and Schuster, 1995. Kendix, Michael, “Institutional rigidities as a barrier to growth: A regional perspective,” Federal Reserve Bank of Chicago, working paper, No. WP-1990-6, 1990. Kim, Sukkoo, “Changing structure of U.S. regions: A historical perspective,” Assessing the Midwest Economy Working Paper Series, No. SP-1, Federal Reserve Bank of Chicago, 1996. 16 Klier, Thomas H., “Structural change and technology in the manufacturing sector,” Assessing the Midwest Economy Working Paper Series, No. SP-7. Federal Reserve Bank of Chicago, 1996. Mattoon, Richard H., and William A. Testa, “State and local governments’ reaction to recession,” Economic Perspectives, Federal Reserve Bank of Chicago, Vol. 16, No. 2, March/ April 1992, pp. 19–27. McAlinden, Sean P., Brett C. Smith, and David E. Cole, Driving America’s Renaissance, Ann Arbor, MI: University of Michigan, Office for the Study of Automotive Transportation, 1995. McNulty, Timothy, “Joint initiatives in the states and provinces,” in The Great Lakes Economy Looking North and South, William A. Testa (ed.), Chicago: Federal Reserve Bank of Chicago, 1991. Meredith, Robyn, “Ford sales increase 2.8% while Toyota soars 55%,” The New York Times, February 5, 1997, p. C4. Oakland, William H., and William A. Testa, “State–local business taxation and the benefits principle,” Economic Perspectives, Federal Reserve Bank of Chicago, Vol. 20, No. 1, January/February 1996, pp. 2–17. Plaut, T.R., and J.E. Pluta, “Business climate, taxes and expenditures, and state industrial growth in the U.S.,” Southern Economic Journal, Vol. 50, 1983, pp. 99–119. Rubenstein, James M., “The evolving geography of production—Is manufacturing activity moving out of the Midwest? Evidence from the auto industry,” Assessing the Midwest Economy Working Paper Series, No. SP-3, Federal Reserve Bank of Chicago, 1996. Sutten, Kerry, “Fiscal devolution: The impact on state budgets,” Northeast–Midwest Institute Review, Washington, DC: Northeast–Midwest Institute, June 1996. Swonk, Diane, “State winners and losers,” Economic Backgrounder, First National Bank of Chicago, May 1991. ECONOMIC PERSPECTIVES , “The Great Lakes economy revisited,” Assessing the Midwest Economy Working Paper Series, No. SP-2, Federal Reserve Bank of Chicago, 1996. U.S. Advisory Commission on Intergovernmental Relations, Measuring State Fiscal Capacity and Effort 1987, Washington, DC: ACIR, 1989. Testa, William A., “Producer services: Trends and prospects for the Seventh District,” Economic Perspectives, Federal Reserve Bank of Chicago, Vol. 16, No. 3, May/June 1992, pp.19–28. U.S. Department of Commerce, Bureau of the Census, Current Industrial Reports: Manufacturing Technology, Washington, DC, 1988. Testa, William A., Thomas H. Klier, and Richard H. Mattoon, Assessing the Midwest Economy: Report of Findings, Federal Reserve Bank of Chicago, April 1997. FEDERAL RESERVE BANK OF CHICAGO , Current Industrial Reports: Manufacturing Technology: Prevalence and Plans for Use, Washington, DC, 1994. Wisconsin Strategic Development Commission, Final Report, Madison, WI, 1985. 17 Accounting for the federal government’s cost of funds George J. Hall and Thomas J. Sargent The press routinely reports extraordinarily large government deficits, mainly consisting of interest costs, for countries experiencing high rates of inflation.1 For example, the New York Times reported that in 1993 Brazil’s government deficit was 30 percent of the country’s gross domestic product (GDP).2 Most of this deficit was accounted for by interest costs. In the late 1980s, less dramatic but still large government interest costs (around 12 percent of GDP) were reported for Italy. These large ratios are computed by dividing a government’s nominal interest payments by nominal GDP. Financial specialists know such figures to be substantial overstatements because they fail to account for the real capital losses that government creditors experience during high inflation.3 Every year, an equally flawed ratio is reported in the federal budget of the United States. Figure 1 reports these official interest expenses as a percent of federal outlays over the period 1960 to 1995.4 The figure displays the well-known 1980s growth in interest payments as a fraction of outlays, a hallmark of Reaganomics. Figure 2 displays our corrected estimates of federal interest expenses as a fraction of federal outlays. Compared with the official numbers, the true figures are much more variable, and lower on average. It is timely to note that section 7 of the recently proposed balanced budget amendment explicitly includes the official interest payments on the federal debt as expenditures. 5 We are not necessarily suggesting that the framers of the amendment are unaware that this measurement is flawed from an purely economic standpoint. The current measure tends to overstate 18 interest payments more the higher the inflation rate is. By including the official measure of interest costs, the amendment’s framers may intend to add incentives to lower both inflation and expenditures. This article describes and defends our corrections to the official series. After showing how to do the accounting correctly, we calculate how the interest costs of the government would have been affected had it used a different debt-management strategy. We simulate the consequences of particular versions of shorts only and longs only debt-management policies, two classic policies that have been advocated. A flawed measure of the government’s cost of funds When investors compute the real return on an equity or debt investment, they take into account dividend and coupon payments, the change in price of the stock or bond, and the effect of inflation on the general price level. So should the government in accounting for its interest costs to the public. In each time period, the government repays its debtholders in two ways: explicitly in the form of coupon payments and principal repayments, and implicitly in the form of real capital gains on outstanding debt stemming from the diminished term to maturity of the debt, interest rate changes, and inflation. To measure the government’s cost of funds, one must account for George J. Hall is an economist at the Federal Reserve Bank of Chicago. Thomas J. Sargent is a senior fellow at the Hoover Institution at Stanford University, the David Rockefeller Professor of Economics and Social Sciences at the University of Chicago, and a consultant to the Federal Reserve Bank of Chicago. ECONOMIC PERSPECTIVES FIGURE 1 Net interest costs as a share of government spending percent 20 10 0 -10 1960 ’65 ’70 ’75 ’80 ’85 ’90 ’95 Source: U.S. Department of Commerce, Bureau of Economic Analysis, National Income and Product Accounts, various years. the capital gains and losses on all outstanding Treasury securities. The federal government reports an incorrect measure of its cost of funds. It records an imperfect measure of its explicit interest costs and ignores its implicit interest costs. The government computes its cost of funds by forming the sum of current coupons on long-term FIGURE 2 Total cost of funds and interest costs as a share of government spending coupon bonds and the appreciation on shortterm discount bonds.6 This measure of the government’s cost of funds, shown in figure 1, is a remarkably smooth series, and it is always positive. The following example illustrates how the government’s methodology mismeasures its cost of funds. Consider two bonds that would raise the same value for the government at time t = 0, assuming no uncertainty and a constant real interest rate, r. One is a pure discount (zero-coupon) bond with ten periods to maturity, paying off P0 at time 10; the second is a coupon bond with coupon c, paying off P1 at time 10. From the net one-period interest rate we can compute the discount factor, 1/(1 + r). The value of the pure discount bond p0(t) satisfies p0(t) = (1 + r)–1p0(t + 1), for t = 0, 1, . . . , 9, where p0(10) = P0. Evidently, for the pure discount bond, interest accrues through the gradual appreciation in the value of the bond from p0(0) = (1 + r)–10 P0, at time 0 to p0(10) = P0 at time 10. The rate of appreciation equals the gross interest rate: 1+r = . p0(t) The value p1(t) of the coupon bond satisfies p1(t) = c + (1 + r)–1 p1(t + 1), t = 1, . . . , 9 and p1(0) = (1 + r)–1p1(1), where p1(10) = P1. The interest rate satisfies 1) 1 + r = percent 20 p0(t + 1) c p1(t) + p1(t + 1) p1(t) for t = 1, . . . , 9. 10 Interest cost 0 Cost of funds -10 1960 ’65 ’70 ’75 ’80 ’85 ’90 Note: The cost of funds was calculated using equation 5. Sources: Authors’ calculations and U.S. Department of Commerce, Bureau of Economic Analysis, National Income and Product Accounts , various years. FEDERAL RESERVE BANK OF CHICAGO ’95 For coupon bonds of finite maturity with a principal payment at the end (really a last big coupon), interest payments (that is, the left-hand side of equation 1) include more than the coupon. Hence, it is not appropriate to measure the interest costs associated with coupon bonds by simply adding up the coupons due this period. Indeed, coupon payments do not represent pure interest in an economic sense; they are partly a repayment of principal. Furthermore, part of the return to investors, and of the cost to the issuer, is in the form of capital gains or losses on bonds as time passes. Any bond with a large final 19 payment is partly a pure discount bond with a significant portion of its return coming in the form of capital gains or losses over time. Our example indicates some but not all of the corrections that we want to make in the government’s accounting for its interest costs. The other adjustments have to do with the treatment of inflation, the time variation in interest rates, and the existence at any moment of a variety of bonds with various coupon schedules and maturities. Next we expand our example to incorporate all of these features and show how to do the accounting properly. Doing the accounting right We can build a system for properly counting the government’s real interest costs by carefully rearranging the government’s period by period budget constraint. We manipulate the budget constraint algebraically to isolate explicit and implicit interest expenses. Explicit interest expenses are the real capital gains on one-period discount bonds; implicit interest expenses are the capital gains to the public from holding longer term bonds. First we need to convert nominal yields to maturity on government debt into prices of claims on future dollars in terms of current goods. The modern theory of the term structure of interest rates prices a coupon bond in three steps: 1) viewing the coupon bond as a bundle of pure discount bonds; 2) unbundling it into the constituent pure discount bonds and valuing these components; and 3) adding up the values of the components to attain the value of the bundle. The theory thus strips the coupons from the bond, and prices the bond as though it is a weighted sum of pure discount bonds of maturities 1, 2, . . . , j. (The market and the government have followed theory: pre-stripped zero-coupon bonds, or STRIPS, themselves are available in the market.) Let sjt be the number of dollars at time t + j that the government has promised to deliver, as of time t. To compute sjt from historical data, we have to add up all of the dollar principalplus-coupon payments that the government has promised to deliver at date t + j as of date t. Let ajt be the number of time t goods that it takes to buy a dollar in time t + j. We work with a real (inflation-adjusted) price, ajt, denominated in units of time goods (so-called dollars of constant purchasing power) and not time t dollars, because we want to keep track of the 20 government accounts in real (in goods) terms. We can calculate the prices ajt from 2) ajt = νt (1 + ρjt ) j , where νt is the value of currency (the reciprocal of the price level, measured in goods per dollar), and ρjt is the yield to maturity on a j-period pure discount bond.7 Equation 2 tells how to convert the yield to maturity ρjt on a j-period nominal pure discount bond into the real price of a promise, sold at time t, to one dollar at time t + j. Let deft be the government’s real net-ofinterest budget deficit, measured in units of time t goods. We can write the government’s time t budget constraint as: n 3) a s Σ j=1 jt jt n = a Σ j=1 s j–1,t j,t–1 + deft , where it is understood that a0,tm ≡ νt and n denotes the longest years to maturity for bonds.8 The left-hand side of equation 3 is the real value of the interest bearing debt at the end of period t, determined by multiplying the number of time t + j dollars that the government has sold in the form of j period pure discount bonds, sjt, by their price in terms of time t goods, ajt, and then summing this product (or value) over all such outstanding bonds, j = 1, . . . , n. The right side of equation 3 is the sum of the current net-ofinterest real deficit, deft, and the value of the outstanding debt that the government owes at the beginning of the period, which in turn is simply the value this period of the outstanding promises to deliver future dollars, sj ,t–1, that the government issued last period. Equation 3 can be rearranged to take the form n 4) a s Σ j=1 jt jt n = (a Σ j=1 j–1,t – aj,t–1)sj,t–1 + n a Σ j=1 s j,t–1 j,t–1 + deft . These two forms of the budget constraint are algebraically equivalent. We have remarked how equation 3 expresses the real value of total debt with which the government n leaves a period t, Σ j=1 ajt sjt as the sum of the real value of obligations with which it enters ECONOMIC PERSPECTIVES FIGURE 3 Capital gains on zero-coupon bonds unit cost 0.15 0.10 7-year 0.05 1-year -0.00 -0.05 14-year -0.10 1960 ’65 ’70 ’75 ’80 ’85 ’90 ’95 Note: Capital gains are calculated per unit on a year t + j – 1 dollar, as of date t, measured in time t dollars. The authors converted the capital gains to current dollars by multiplying aj – 1, t – aj,t – 1 by the time t price level, vt–1. the period, Σj=1 aj–1, t sj,t–1, and the government’s net-of-interest deficit, deft. Equation 4 breaks the first term on the right side of equation 3 into an interest component and a previous value component. Again, the left-hand side of the budget constraint in equation 4 is the real value of government debt that the government has outstanding at the end of period t. The first term on the right-hand side of the budget constraint in equation 4 represents interest on the government debt, and can be decomposed as n occur during periods of rising inflation or rising interest rates. Second, the capital gains and losses of bonds of different maturities move together. So the government could not have eliminated the inflation and interest rate risk inherent in its portfolio by manipulating the maturity structure of the debt. Third, the longer the maturity of the bond, the greater the volatility of the capital gains. Increasing (or decreasing) the average maturity of the outstanding debt increases (or decreases) the government’s and the public’s exposure to inflation and interest rate risks. In figure 4 we report our breakdown of the total interest costs on the marketable federal debt between explicit and implicit real interest costs. In general, the explicit interest costs were relatively small and relatively constant from 1960 to 1995. In contrast, the implicit interest costs were substantial, variable, and often negative. Since the real value of the outstanding debt was growing over this period, sjt was growing over time. So the per dollar capital gains are being multiplied by increasingly large numbers. Thus the implicit interest cost became more volatile throughout the sample period. We compute the total interest costs born by the federal government by simply adding up the explicit and implicit interest costs. Total interest costs as a percent of government outlays are plotted in figure 2. The explicit, implicit, and total interest costs, as well as the total debt outstanding, in millions of 1983 dollars are reported in table 1. In contrast to the Treasury’s n 5) (νt – a1,t–1)s1,t–1 + Σ (aj–1,t – aj,t–1)sj,t–1. j=2 The first term in equation 5 is explicit interest and the second term is implicit interest or the capital gain to the public on its claims on the government. Thus, the term νt – a1, t–1 is the per dollar real capital gain accruing to one-period discount bonds issued at time t – 1. The term aj–1, t – aj, t–1 is the change in the price in terms of goods between t – 1 and t of a claim to one dollar in time t – 1 + j; multiplying this change in price by the dollar value of time t – 1 + j claims outstanding, sj,t–1 at time t – 1, and summing over j gives the capital gain to the public. These capital gains are not trivial. In figure 3 we plot the per dollar nominal capital gains, n Σj=1(aj–1,t – aj,t–1), for one-year, seven-year, and 14-year zero-coupon bonds. There are three things to note. First, capital losses can be quite large, and they occur frequently. These losses FEDERAL RESERVE BANK OF CHICAGO FIGURE 4 Real interest costs percent 20 10 Explicit costs 0 Implicit costs -10 1960 ’65 ’70 ’75 ’80 ’85 ’90 ’95 Sources: See figure 2. 21 TABLE 1 Federal government’s interest costs in millions of 1983 dollars Year Explicit Implicit Total 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 9,358 4,886 5,484 3,916 7,536 5,809 4,523 5,370 3,440 2,027 6,780 5,174 2,501 –9,354 –13,212 –3 3,596 –4,796 –5,321 –6,629 –2,081 10,887 27,383 18,009 23,430 24,336 31,339 7,496 12,637 19,623 6,981 20,768 7,037 5,884 6,882 0 3,863 10,142 394 7,593 –2,738 3,325 –7,843 –3,733 –7,477 13,596 8,640 –1 –7,662 –9,568 539 10,573 –10,077 –16,504 –17,580 –17,979 –3,618 57,556 8,730 37,918 87,049 101,356 –20,884 19,428 73,032 12,851 105,116 40,695 90,648 –73,186 9,358 8,748 5,626 4,310 15,129 3,071 7,848 –2,472 –293 –5,449 20,376 13,813 2,500 –17,016 –22,780 536 4,169 –14,874 –21,825 –24,208 –20,060 7,269 84,939 26,739 61,349 111,385 132,696 –13,388 32,066 92,655 19,832 125,884 47,732 96,532 –66,304 1995 28,537 173,862 202,399 Evaluating alternative portfolio strategies Assuming that postwar U.S. interest rates had remained unchanged, how would the government’s interest expenses have been affected if it had followed a different debtmanagement policy? If we restrict the government to issuing from its historically observed menu of instruments, this question can be answered by mechanical calculations. We compose alternative hypothetical portfolio strategies, and track the net costs the government would have incurred at historically realized interest rates.10 Below we describe how these costs can be calculated, and perform some of these calculations.11 Given historical time series data on {ajt , sjt , deft , νt}Tt=t 0 we can use equations 4 and 5 to account for interest payments on the government debt. Given {ajt , deft, νt}Tt=t 0, we can evaluate the effects on the government budget of portfolio strategies {sjt}Tt=t 0 other than the historical one. These alternative portfolio strategies must be constructed to respect the government budget constraint in equation 3. The alternative strategies are: 1. Shorts only: Set sjt = 0 for j > 1, ∀t. 2. Tens only: Set sjt = 0 for j ≠ 10, ∀t. 3. Longs only: Set sjt = 0 for j < n, ∀t, where n is the longest bond priced by the McCulloch and Kwon (1993) dataset. The first and third policies represent the poles of proposed debt-management policies.12 For an economy with only nominal interest bearing debt, the class of feasible financing rules is akt skt calculations plotted in figure 1, our computed costs of funds are quite volatile.9 These costs were negative during periods of large capital losses in the Treasury bond market (for example, the high inflation episodes of the 1970s and the dramatic fall in bond prices in 1994). Ultimately how volatile the federal government’s interest costs are depends on how the government shares inflation risk and interest rate risk with the public. The Treasury and the Federal Reserve can alter this risk-sharing arrangement between the government and the public by manipulating the maturity structure of the outstanding debt. To illustrate this, we run three counterfactual portfolio strategies. 6) 22 ECONOMIC PERSPECTIVES n deft + Σ aj–1,t sj,t–1 = fk t , j=1 and n 7) Σf k=1 kt = 1. In words, fkt is the fraction of the outstanding debt at time t that is due at time t + k. Restrictions in equations 6 and 7 are algebraic implications of the government budget n constraint in equation 3. Let Σ j=1 aj–1,t sj,t–1 ≡ Vt be the value of interest bearing government debt at the beginning of period t. Given a policy fkt, k = 1, . . . n, together with observed interest rates, equation 5 can be solved for skt, k = 1, 2, . . . n: skt = fkt (a ) (def + V ). t t kt This equation can be solved recursively to build up records skt and decompositions of interest cost under alternative hypothetical debt-management rules. The first two policies fall into the set of simple rules that are time invariant, that is, fit = fj ∀ t . For the bills only policy, f1t = f1 = 1 and fkt = fk = 0 for all k ≠ 1. For the tens only policy f10t = f10 = 1 and fkt = fk = 0 for all k ≠ 10. Our third policy, longs only, is a time varying policy, which depends on the maturity of the longest bond outstanding each year during our sample period. For each of these policies, the entire debt is purchased and resold to make sure all the debt is held in either one-year bills, ten-year zero-coupon bonds, or n-year zero-coupon bonds (depending on the experiment). For any feasible specifications of fjt , we can evaluate the implicit and explicit interest costs of financing a stream of government deficits. Our three policies will have quite different effects, largely through the behavior of the value of currency, νt. December observation of each year to create the annual pt series. The yield to maturity series, ρjt, is constructed by point sampling end-of-month data from McCulloch and Kwon (1993) and Bliss (1996) containing the zero-coupon yield curve implicit in U.S. Treasury coupon bond prices. Hence ρjt is the yield to maturity on a j-period pure discount bond as of December 31 of year t. We calculate the prices, ajt, using equation 2: ajt = νt , (1 + ρjt )j where j = 1, 2, . . . , 30 and t = 1960, . . . , 1995. In constructing our counterfactual debtmanagement figures, we use equations 4 through 7 with the appropriate time-invariant fit. We imputed the real net-of-interest deficit series, deft, from the government budget constraint, equation 3, using the actual ajt and sjt series. The results Below, we discuss some of the properties of the actual sjt and fjt series, review the historical paths for inflation and the term structure of interest rates, and report the results of our experiments. Figure 5 shows the average maturity for our calculated sjt series. Its variations generally match those of the average maturity of the federal debt series reported by the Treasury, though the levels differ. We believe there are The data The sjt series are computed using data from the CRSP Government Bonds Files. For each Treasury note and bond outstanding, CRSP reports the maturity date, the coupon rate, and the face value held by the public. The original source for these data is table PDO-1 of the Treasury Bulletin. Since neither the Treasury Bulletin nor CRSP reports the face value of Treasury bills held by the public, these data are backed out of table FD-5 of the Treasury Bulletin. All data are as of December 31 of each year. The value of the currency, νt, is computed by: 100 νt = p , t where pt is the price level. The price level is the monthly series CPI—all items, from the Bureau of Labor Statistics. The base for the CPI series is 1982–84 = 100. We sample the FEDERAL RESERVE BANK OF CHICAGO FIGURE 5 Average maturity of constructed pure discount bonds years 5 4 3 2 1 1960 ’65 ’70 ’75 ’80 ’85 ’90 ’95 23 FIGURE 6 FIGURE 7 Maturities of federal debt due in j years percent 100 1975 Inflation rate percent 15 1965 75 10 50 5 1995 25 0 5 10 15 20 25 30 0 1960 ’65 ’70 ’75 ’80 ’85 ’90 ’95 Sources: U.S. Department of Labor, Bureau of Labor Statistics and Federal Reserve Board of Governors. some problems the Treasury’s series (for example, it confines itself to marketable securities only after 1975) and prefer our methodology of reducing each bond to a zero-coupon basis by allocating coupons to the year in which they fall due. The average maturity falls from the mid1960s to 1975 and rises steadily for about thirteen years before leveling off at four years for the last seven years of the sample. The steady fall in the average maturity during the late 1960s and early 1970s is partly the consequence of federal legislation, repealed in 1975, which prevented the Treasury from issuing long-term instruments paying interest above a threshold rate that market rates were then exceeding. As we shall see, by causing the Treasury to shorten the average maturity of its debt during the high inflation years of the 1970s, this law prevented the government from fully benefiting from the negative implicit real interest it managed to pay through inflation.13 Figure 6 plots the percentage of the federal debt due within j years for 1965, 1975, and 1995. This figure was constructed by taking a cumulative sum of the observed f jt series for each of the three years. Throughout the period we studied, the federal debt was heavily weighted toward securities with maturities of one year or less. In 1995, almost 40 percent of the government’s portfolio was due within one year. Only a tiny fraction of the federal debt is financed with long-term bonds. 24 Figure 7 plots the percentage change in the price level, the inflation rate, and is dominated by the high inflation rates of the 1970s. The spread between the ten-year bond rate and the one-year bond rate is plotted in figure 8. When this difference is positive, the yield curve is upward sloping. When it is negative, the yield curve is inverted. In general, the inflation rate and the slope of the term structure moved in opposite directions. The yield curve tended to FIGURE 8 Ten-year rate minus one-year rate basis points 400 200 0 -200 1960 ’65 ’70 ’75 ’80 ’85 ’90 ’95 Source: J. McCulloch and H. Kwon, “U.S. term structure data, 1947–1991,” Ohio State University, working paper, No. 93-6, 1993; and R. Bliss, “Testing term structure estimation methods,” Advances in Futures and Options Research, 1996. ECONOMIC PERSPECTIVES FIGURE 9 FIGURE 10 Interest as a share of total government expense, bills only Marketable debt as a share of gross domestic product, bills only percent 20 percent 45 10 30 Bills only interest Bills only marketable debt 0 15 Total marketable debt Total interest -10 1960 ’65 ’70 ’75 ’80 ’85 ’90 ’95 Sources: See figure 2. flatten or become downward sloping during periods of rising inflation. Figures 9 and 10 display the results of our first experiment: a bills only policy. Figure 9 shows that the realized total real interest (explicit plus implicit) would have been somewhat higher during the late 1960s and most of the 1970s under the bills only policy than under the actual policy followed. During the late 1960s the yield curve was inverted, so longterm rates were below short-term rates. But, more importantly, the high inflation of the 1970s substantially decreased the real value of the federal government’s outstanding obligations.14 However, this pattern reversed in the early 1980s; in the second half of the sample, as inflation fell and the yield curve became consistently upward sloping, the interest costs under the bills only policy would have been lower than under the actual policy. Figure 10 shows that under the bills only policy the real value of the marketable interestbearing debt would have been higher through the mid-1980s. But by the end of the period the real value of marketable interest-bearing debt would have been lower under the bills only policy than under the actual policy. Had the bills only policy been followed, the outstanding debt would have been 34 percent of GDP. In 1995, the actual debt to GDP ratio was 41 percent. FEDERAL RESERVE BANK OF CHICAGO 0 1960 ’65 ’70 ’75 ’80 ’85 ’90 ’95 Sources: See figure 2. Figures 11 and 12 show what would have been the total real interest costs and the real value of marketable debt had a policy been in place of leaving only ten-year bonds outstanding at the end of each year. Figure 11 shows that relative to the actual policy, real interest costs would have been much more variable year to year and would have been negative for many years, especially during the inflationary years of the 1970s. Note that as the size of the federal debt grew, following such a policy would have substantially increased the volatility of the federal government’s cost of funds. Figure 12 shows that under a tens only policy, real government debt would have been 53 percent of GDP in 1995. Figures 13 and 14 show the total real interest costs and the real value of the debt under a policy of issuing the longest available maturity (that is, the longest maturity that was actually priced in McCulloch and Kwon’s (1993) data set). We see more variable interest costs but less accumulation of debt under the longs only policy than under the tens only policy. Note that in the 1970s, due to the high inflation, the real value of the outstanding debt would have been substantially lower under the longs only than under the actual policy. However, by the end of the sample period, the ratio of the outstanding debt to GDP under the longs only policy would have been considerably higher than the ratio under the actual policy. 25 FIGURE 11 FIGURE 12 Interest as a share of total government expense, tens only Marketable debt as a share of gross domestic product, tens only percent 50 percent 60 25 40 Total interest 0 Total 20 Ten-year bond policy Tens only interest -25 1960 ’65 ’70 ’75 ’80 ’85 ’90 ’95 0 1960 Sources: See figure 2. ’65 ’70 ’75 ’80 ’85 ’90 ’95 Sources: See figure 2. These results indicate that debt-management policies weighted toward longer maturities would have led to lower interest costs and less accumulation of debt over the period from 1960 to 1980. After 1980, debt-management policies weighted toward shorter maturities would have generally lowered interest costs and led to less accumulation of debt. From figure 5 it is clear that the Treasury and Federal Reserve reduced the average maturity of outstanding debt throughout the 1960s and early 1970s; they then increased the average maturity during the late 1970s and throughout the 1980s. Our analysis indicates that to have minimized its borrowing costs, the government should have engaged in the opposite strategy. Of course, with hindsight we could have found the portfolio-share policy that would have minimized the government’s cost of funds. However, the purpose of these counterfactual FIGURE 13 FIGURE 14 Interest as a share of total government expense, long bonds only Marketable debt as a share of gross domestic product, long bonds only percent 60 percent 90 60 40 30 Total interest Total debt 0 20 -30 -60 1960 ’65 ’70 Sources: See figure 2. 26 Longs only debt Longs only interest ’75 ’80 ’85 ’90 ’95 0 1960 ’65 ’70 ’75 Sources: See figure 2. ECONOMIC PERSPECTIVES ’80 ’85 ’90 ’95 exercises is not to engage in “Monday morning quarterbacking” but to illustrate how the maturity structure of the debt affects how the government and bondholders share inflation and interest rate risk. Moreover, caution is in order in interpreting the results of an evaluation of a counterfactual debt-management policy. Interest rates are a random process, and the result of following a given strategy is too. What most drives the outcome of our counterfactual exercises is the outcome for inflation. In issuing nominal securities, the government is offering the public a risky instrument whose real return is sensitive to the rate of inflation over the life of the bond. Our results indicate that from 1960 to 1995, inflation came in high with sufficient frequency to let the government often pay negative real interest and sometimes substantially negative real interest. These high inflation rates make the longer maturity portfolio policies come in with lower interest costs during the 1970s. Clearly, the outcome would have been different had inflation come in much lower. Conclusion This article makes two points. First, the federal government reports a flawed measured of its own cost of funds. Second, the maturity structure of the debt influences the way inflation risk and interest rate risk are shared by the government and its creditors. The first point is not just nit-picking. By ignoring the effects of inflation and changes in interest rates on the value of the outstanding federal obligations, the official interest payment calculations make it difficult to evaluate the true cost of various proposals. For example, the introduction of index bonds will change how the government shares inflation risk with its creditors since the government can not induce capital losses on these bonds through inflation. How these bonds can be expected to influence to government’s cost of funds is beyond the scope of this paper; but it should be clear that the Treasury’s accounting methods are inappropriate for evaluating the costs of these new bonds. The second point is a word of caution regarding periodic calls for the Treasury to “painlessly pare billions from its interest bill by refinancing the government’s existing debt with bonds that mature more quickly.”15 While our counterfactual experiments demonstrate that shortening (or lengthening) the average maturity of the U.S. debt can at times save the Treasury billions of dollars, these savings depend on the future paths of interest rates and inflation—two series which are notoriously hard to predict. And if the government bets the wrong way, the mistake can be quite expensive. NOTES 1 This article extends estimates and arguments from Sargent (1993). 2 See the article by Nash (1993). 3 See Ljungqvist and Sargent (1997) chapter 8, exercise 1; and Blanchard and Sachs (1981). 8 See Sargent and Wallace (1981) for a discussion of this form of the government budget constraint, in particular for a defense of the use of pre-tax real yields on government debt and a net of interest government deficit. Sargent and Wallace use a ‘crowding out’ assumption to justify the use of pre-tax yields. 9 4 The series plotted is net interest paid by the federal government from the National Income and Product Accounts. The figure displays a series which is remarkably smooth and always positive. 5 The Treasury’s calculations include some assets (chiefly savings bonds and some securities issued to state and local governments) that are not included in our analysis. So these two graphs are not strictly comparable. Nevertheless, we expect that adding these assets to our analysis would not change the results in any meaningful way. See H.J. Resolution 1, 105th Congress, 1st Session. 10 6 The Department of the Treasury calculates the net interest as the sum of coupon payments, accrued interest on bills and zero-coupon bonds, and interest on nonmarketable debt. Under the assumption that historical interest rates would have been unaffected by the switch in debt policy, this accounting exercise involves no use of economic theory. To infer the government’s costs had it issued different assets, for example indexed bonds, we would need a theory about the price of pure discount indexed bonds. 7 We use the yield to maturity series for pure discount bonds constructed by McCulloch (1990) and McCulloch and Kwon (1993). These data were updated by Bliss (1996). FEDERAL RESERVE BANK OF CHICAGO 11 The standard theory of the term structure of interest rates assumes that interest rates on all maturities would be 27 unaffected by alterations in the maturity structure of the government’s debt. This assumption can be justified by appealing to the logic of the Modigliani-Miller theorem from finance. 13 This is the type of law that Missale and Blanchard (1994) suggest as a device that governments with a large ratio of debt to gross domestic product use to assuage investor inflation fears by reducing the government’s returns from inflation. 12 Actually, not quite the extreme poles. Milton Friedman (1948) advocated the policy that the government finance its deficits and surpluses only by issuing or retiring currency. This is a ‘shorts only’ policy in which only j = 0 maturity debt is issued. At the other end of the spectrum, the classic British policy was to issue only consoles, which are infinite maturity bonds, which amount to an infinite stream of pure discount bonds, one for each date in the future. 14 Recall that a Treasury bond is a promise to pay a certain number of nominal dollars at a future date. When inflation increases, the real value of those nominal dollars falls. 15 The quote is from Blinder (1992). Also see Forsyth (1993) and Passell (1993). REFERENCES Blanchard, Olivier Jean, and Jeffrey Sachs, “There is no significant budget deficit,” New York Times, March 6, 1981, p. 26. McCulloch, J. Huston, and Heon-Chul Kwon, “U.S. term structure data, 1947–1991,” Ohio State University, working paper, No. 93-6, 1993. Blinder, Alan, “Let’s start the new year by cutting the debt load,” Business Week, January 13, 1992, p. 18. Missale, Alessandro, and Olivier Jean Blanchard, “The debt burden and debt maturity,” American Economic Review, Vol. 84, No. 1, March 1994, pp. 309–319. Bliss, Robert, “Testing term structure estimation methods,” in Advances in Futures and Options Research, Vol. 8, Peter Ritchken (ed.), November 1996. Forsyth, Randall, “Don’t stop thinkin’ of how to borrow,” Barron’s, January 25, 1993, pp. 52–53. Friedman, Milton, “A monetary and fiscal framework for economic stability,” American Economic Review, Vol 48, June 1948, pp. 245– 264. Ljungqvist, Lars, and Thomas J. Sargent, “Recursive macroeconomic theory,” Stanford University, manuscript, Internet, available from http://riffle.stanford.edu, March 1997. Nash, Nathaniel C., “Brazil seeks to revive its economic miracle, lying in ruins,” New York Times, February 16, 1993, Section A, p. 9. Passell, Peter, “Economic scene,” New York Times, January 21, 1993, Section D. Sargent, Thomas J., “Fact or fiction: Shortening debt maturity lowers interest costs,” Catalyst Institute Research Project, manuscript, December 1993. Sargent, Thomas J., and Neil Wallace, “Some unpleasant monetarist arithmetic,” Federal Reserve Bank of Minneapolis, Quarterly Review, Vol. 5, Fall 1981, pp. 1–17. McCulloch, J. Huston, “U.S. term structure data, 1946–87,” in Handbook of Monetary Economics, Vol 1, Benjamin Friedman and Frank Hahn (eds.), Amsterdam: North-Holland, 1990, pp. 672–715. 28 ECONOMIC PERSPECTIVES