The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
SEPTEMBER/OCTOBER 1991 ECONOMIC PERSPECTIVES A review from the Federal Reserve Bank of Chicago Energy dependence and e fficie n cy B alan cin g a ct: Tax structure in the Seven th D istrict FEDERAL RESERVE BANK OF CHICAGO Contents Energy dependence and e ffic ie n c y ........................................................................................................2 Ja ck L. Hervey What ever happened to solar power and yurts? The author discusses developments in energy since the oil crisis of the early 1970s. B alan cin g a ct: Tax structure in the Seventh D istric t...................................................................................... 22 Richard H. M atto o n Should states tax snacks or cap taxes? As discussed in this article, raising revenues in a recession can be tricky business. ECONOMIC PERSPECTIVES Karl A. Scheld, Senior Vice President and Director of Research Editorial direction Carolyn McMullen, editor, David R. Allardice, regional studies, Herbert Baer, financial structure and regulation, Steven Strongin, monetary policy, Anne Weaver, administration Production Nancy Ahlstrom, typesetting coordinator, Rita Molloy, Yvonne Peeples, typesetters, Kathleen Solotroff, graphics coordinator Roger Thryselius, Thomas O’Connell, Lynn Busby-Ward, John Dixon, graphics Kathryn Moran, assistant editor SEPTEMBER/OCTOBER 1991 Volume XV, Issue 5 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 Public Information Center, Federal Reserve Bank of Chicago, P.O. Box 834, Chicago, Illinois 60690-0834, or telephone (312) 322-5111. Articles may be reprinted provided source is credited and The Public Information Center is provided with a copy of the published material. ISSN 0164-0682 Energy dependence and efficien cy Jack L. Hervey World oil market fragility was demonstrated again last year following the August 2 inva sion of Kuwait by Iraq and the subsequent United Nations imposed embargo on oil shipments from Iraq and Kuwait. Concern about the potential cut off of Middle East oil brought back memories of long lines at gas stations, dim lighting in offices, lower thermostats on furnaces and reduced use of air conditioners. In public de bate, issues of energy dependence, efficiency, and conservation were again in vogue. Devel opments in the Middle East from August 1990 through February 1991 serve as an effective reminder of the importance of energy in general and petroleum in particular to the industrial economies. The intent of this article is two fold. In order to set the stage, it first reviews recent developments in the oil markets. Next, it sur veys concepts of energy and oil-use dependence as well as energy and oil-use efficiency. It proposes measures of dependence and efficien cy and uses these measures to compare devel opments in the world’s six heaviest consumers of energy—Canada, France, West Germany, Japan, the United Kingdom, and the United States—during the period 1970 to 1988. R ecent d ev elo p m en ts Following the U.N. embargo on oil ship ments from Iraq and Kuwait, world crude oil production declined 6 percent in August, a reduction of 3.5 million barrels per day. How ever, initial fears of an oil shortage proved 2 unfounded. Within days, other major oil ex porters pledged to increase production to offset the 4.5 million barrels per day of lost Iraq and Kuwait oil. World crude oil production and prices are shown in Figure 1. Nevertheless, oil prices soared, bringing on the fourth major price shock to world oil mar kets in less than two decades (counting the rapid decline in oil prices in 1986). Uncertain ty in the markets intensified as multinational forces opposing the Iraqi move began to deploy in Saudi Arabia and the threat of military con frontation grew. Spot prices for crude oil peaked at over $40 per barrel in early October, up from $19 per barrel prior to the invasion of Kuwait. By late November, oil production by the 11 remaining OPEC members, in particular Saudi Arabia, exceeded that of all 13 members of OPEC prior to the invasion, thus wiping out the Iraqi-Kuwaiti export deficit. World production returned to pre-August levels of 60-61 million barrels per day. In addition, world oil demand was slipping because of weakening economic conditions in oil importing countries and oil consumers’ response to higher oil prices. The author is a senior econom ist at the Federal Reserve Bank of Chicago. The author w ould like to express his appreciation to Kenneth Kuttner, Prakash Loungani, James Moser, and Steve Strongin for their helpful com m ents during the project and on earlier drafts of the paper. The views expressed herein, however, remain those of the author and do not necessarily reflect the views of the com m entators or the Federal Reserve Bank of Chicago. ECONOMIC PERSPECTIVES By mid-December, spot prices for crude had declined to well below $30 per barrel. Analysts were beginning to suggest that crude oil prices would once again drop to levels well below $20 per barrel during 1991, assuming a favorable resolution of the Persian Gulf situa tion. With the initiation of the allied air offen sive against Iraq on January 16, 1991, oil prices surged again, to more than $30 per barrel. However, it soon became apparent that oil production in the Gulf states would not be ap preciably affected by the war, nor would ship ping lanes in the area be disrupted. Oil prices promptly declined to the $20-$25 per barrel range. The initiation of the land war by Coalition Forces on February 24 and its rapid conclusion contributed to a further easing of market ten sions. Real economic factors once again domi nated the market. Oil prices dropped into the $ 18-$20 per barrel range and day-to-day price variation decreased markedly. The lesson learned from Kuwait Soaring oil prices once again raised the issue of the economic dependance of the world’s economies on energy, in general, and on petroleum in particular. This issue has been ignored, if not forgotten, by all but a few ana lysts and policy makers during the last half of the 1980s, a period during which nominal crude oil prices declined from around $30 per barrel to $10 per barrel before increasing again to settle in the mid-to-high teens. FEDERAL RESERVE BANK OF CHICAGO Not surprisingly, public interest in energy conservation and a national energy policy subsided as soon as oil prices declined into the low twenties and high teens. Indeed, the oil price most often faced by the consumers, motor gasoline, was lower, when adjusted for infla tion, at mid-year 1991 than prior to the 1973-1974 price shock. Public officials’ disinclina tion towards serious energy con servation measures is illustrated by Congressional response to an Administration proposal of a 12 cent per gallon gas tax as part of the 1990 tax bill. The Congress enacted a 5 cent per gallon in crease in the gasoline tax, an increase motivat ed more by the desire to increase revenue than to promote conservation.1The Iraq and Kuwait episode illustrates the fact that the stability of the world oil market is fragile. That fragility grows out of at least two factors: 1) the depen dence of the world’s economies on petroleum as an energy source and 2) the mismatch be tween petroleum producers and consumers.2 In the United States, in particular, discus sion about energy and oil usage focused on how dependent the economy is on oil and how inef ficient the U.S. economy is in its use of energy and oil. This raises the issue of what it means to say that a country is dependent on oil, or on energy in general. In fact, there are a number of different ways to measure how dependent a country is on energy in general, or on a particu lar energy source such as oil. In this article I discuss two measures of energy dependence— total requirements and per capita require ments—and compare the dependence of the six heaviest users of energy (Canada, France, West Germany, Japan, the United Kingdom, and the United States) according to these measures. Investigation into the sources of dependence requires a discussion of energy efficiency. Again, there are different ways to measure a country’s energy efficiency. Four measures of efficiency are presented and compared for the six countries according to each measure of efficiency. Finally, in order to explain some of the differences in efficiency across countries, efficiency by economic sector is examined. 3 The next section deals with a discussion of dependence on energy in general, and on the major energy sources in particular. An o v e rv ie w o f en e rg y d ep en den ce The primary energy sources available to an economy are: 1) coal, 2) petroleum and petro leum products, 3) natural gas, 4) nuclear ener gy, 5) hydro-electric, geothermal, and solar energy (H-G-S), 6) solid fuels other than coal (for example, wood, peat, and incinerated gar bage), 7) electricity (normally electricity is a derived or secondary energy form, however, some countries import electricity, in which case it then becomes a primary energy source to those countries), and 8) heat derived from pub lic combined heat and power plants.3 In the aggregate the last three categories are of minor importance, or, as in the case for the other solid fuels category, the data are inadequate, there fore these categories are not considered in the analysis. An economy’s overall dependence on energy or on a particular energy source, for example, petroleum, can be measured in sever al ways. This article specifically examines total energy utilized from all sources by the economy; energy requirements relative to popu lation (for example, per capita energy require ments from all sources or per capita energy requirements derived from oil); and energy use by source (for example, oil) as a proportion of an economy’s total energy requirements. Energy consumption varies widely across economies. Table 1 presents total primary energy requirements (TPER) for the years 1970 TABLE 1 Total primary energy requirements (Millions of tons of oil equivalent) Year U.S. Japan Germany U.K. 1970 1,579 268 237 208 155 1972 1,695 299 249 213 170 1974 1,721 335 258 212 179 1976 1,778 328 262 205 180 1978 1,900 347 272 209 191 1980 1,826 355 274 201 198 France 1982 1,707 337 252 193 187 1984 1,782 364 262 192 195 1986 1,793 369 270 205 204 1988 1,928 399 274 209 209 4 through 1988 for the world’s six heaviest ener gy users: Canada, France, West Germany, Ja pan, the United Kingdom, and the United States (see note 3 for an explanation of the source and makeup of these data, and a definition of total primary energy requirements). According to Table 1, as of 1988, the U.S. was by far the largest user of energy. With total energy re quirements of 1,928 million tons of oil equiva lent (Mtoe—see note 3 for a definition of Mtoe) U.S. requirements were nearly five times that utilized by the next largest user (Japan) and nine times that of the smallest users, France and the U.K. At the same time, the rate of increase for the U.S. compared favorably with that of the other 5 countries. Indeed, in 1988, U.S. TPER were only 1.5 percent above 1978 lev els.4 Only in the U.K. and Germany was the growth in TPER lower. TPER falls short as a measure for compar ing the relative energy dependence of different countries because it does not take into account factors which determine energy dependence, such as population, the size of the economy, the geographical size of the country, the type of goods produced, and the preferences and con sumption habits of the population. For exam ple, a country with a large population may have greater total energy requirements than a country with a small population without necessarily being more energy dependent, if per capita energy requirements are the same in both coun tries. A geographically large country that relies heavily on automotive transportation may be more energy dependent than a small country where the automobile is a lesser factor. Issues of economic size, population, geographic size, industrial composition and con sumptions habits are addressed in more detail in the balance of the Canada article. Here I turn to a consider ation of per capita energy re 154 quirements for the six countries. 171 Per capita measures of total 184 primary energy requirements are 199 presented in Table 2 and Figure 211 2. They contrast sharply with the 223 total requirements data in Table 213 1. U.S. per capita requirements 224 for all energy remain high in 233 comparison with those of Japan 250 and the European countries but are substantially lower than for ECONOMIC PERSPECTIVES TABLE 2 Total primary energy requirements per capita (Millions of tons of oil equivalent/million population) Year U.S. Japan Germany U.K. France 1970 7.7 2.6 3.9 3.7 3.1 1972 8.1 2.8 4.0 3.8 3.3 1974 8.1 3.0 4.2 3.8 3.4 1976 8.2 2.9 4.3 3.7 3.4 1978 8.5 3.0 4.4 3.7 3.6 1980 8.0 3.0 4.5 3.6 3.7 1982 7.3 2.8 4.1 3.4 3.4 1984 7.5 3.0 4.3 3.4 3.6 1986 7.4 3.0 4.4 3.6 3.7 1988 7.8 3.3 4.5 3.7 3.7 Canada. Most telling, changes over time as derived from the data in Table 2 indicate that the increase in per capita energy dependence in the U.S. during the 1970-1988 period compared favorably with the other countries. In 1988, U.S. per capita dependence for all energy increased less than 2 percent as compared with 29 percent for Canada, 23 percent for Japan, 20 percent for France, and 14 percent for Germany. Per capita requirements declined 2 percent in the U.K. Even more favorable is the per capita energy requirement performance of the U.S. during the last ten years of the period. By this measure the U.S. joined the U.K. in reducing its per capita overall energy dependence while energy depen dence elsewhere continued to rise. FEDERAL RESERVE BANK OF CHICAGO Canada I now turn to a discussion of energy dependence by energy source. Petroleum Before discussing the specif ic measures of oil dependence it is important to distinguish be 7.9 tween an economy’s overall 8.2 dependence on petroleum and an 8.6 economy’s import dependence on 9.0 petroleum. For example, a coun 9.3 try may use little oil and produce 8.6 none domestically, consequently 9.0 importing all oil used. Such a 9.2 country would have a low total 9.6 dependence on petroleum but a high import dependence. Alterna tively, an oil-rich country that has a high level of oil utilization may be a net exporter of petro leum. This country would have a high total dependence but no dependence on imports. The press and policy makers are often interest ed in import dependence because it has impor tant implications for a country’s national secu rity and its international balance of payments.5 However, for the reasons just given, import dependence should not be confused with overall dependence. In this article I restrict attention to overall dependence and do not address the issue of import dependence. As a group, the six countries increased their TPER supplied by oil from 1,270 Mtoe in 1970 to 1,377 Mtoe in 1988, an increase of 8 percent (refer to note 3 for definitions of TPER and Mtoe). The totals are de rived from Table 3. Significant ly, however, the absolute level of oil requirements in 1988 was down 11 percent from the aver age 1,540 Mtoe requirement during the peak period of 19781980. Further evidence indicat ing that dependence on oil de clined is found in data presented in Tables 4 and 5. Here we see that in 1988 two measures of oil dependence, proportion of TPER supplied by oil and per capita oil requirements, respectively, were well below levels recorded during the high consumption period of the late 1970s. 7.2 5 throughout most of the 19 year period, oil accounted for a sub Primary energy requirements supplied by petroleum stantially smaller proportion of (Millions of tons of oil equivalent) TPER in the United States than Year U.S. Japan Germany U.K. France Canada elsewhere—except Canada and to a lesser extent the U.K. At the 184 1970 691 128 94 101 72 same time, however, the decline 1972 771 219 140 111 114 78 in oil’s proportional contribution 1974 244 774 134 105 118 83 to total energy requirements was 1976 824 232 139 92 116 89 markedly smaller in the United 904 1978 255 142 117 95 89 States than elsewhere. 1980 792 235 131 82 111 89 As shown in Table 4, in 1970 1982 708 208 111 77 93 77 the proportion of total energy requirements supplied by petro 1984 724 214 111 88 86 72 leum ranged from a low of 44 207 1986 751 118 78 86 73 percent in the United States to a 791 1988 226 116 79 86 79 high of 69 percent in Japan. Eighteen years later the propor The reduction in dependence on oil by the tion of total energy requirements provided by oil was substantially reduced, ranging from a six countries combined, as compared with the late 1970s, is all the more interesting in the face low of 32 percent in Canada to a high of 57 of the increase in total primary energy require percent in Japan. ments by these economies over the same period Indeed, the absolute dependence on petro (see Table 1). Total energy requirements in the leum, that is, the total amount of oil utilized by six countries combined increased 23 percent the economy, declined for three of the countries between 1970 and 1988 but only 4 percent between 1970 and 1988: France, Germany, and during the latter half of the period—between the U.K. (see Table 3). In all six countries the 1978-1980 and 1988. TPERs supplied by oil were lower in 1988 than In an examination of the individual country during the peak oil-use years of the late 1970s; data presented in Table 4, two patterns stand ranging from down 13 percent for the U.S. to out. First, in each of the six countries, petro down 37 percent for France. leum accounts for a major but declining propor This reduction in the proportion of TPER tion of total energy requirements. Second, supplied by oil and/or the reduction in the abso lute contribution of oil to TPER occurred in the face of a contin TABLE 4 ued expansion in overall energy Proportion of total energy met by petroleum requirements in these econo (Percent) mies—with the exception of the Year U.S. Japan Germany U.K. France Canada United Kingdom, where TPER from all sources remained stable 1970 43.8 68.8 54.1 48.8 60.5 46.9 (see Table 1). 1972 73.2 45.5 56.2 52.1 66.7 45.7 During the 1970-1988 peri 1974 45.0 72.7 51.8 49.6 65.6 45.0 od, less than half of U.S. energy 1976 46.3 70.6 53.0 45.0 64.4 44.6 requirements (ranging from 48 1978 47.6 73.4 52.4 45.5 61.1 42.3 percent in 1978 down to 41 per 1980 43.4 66.1 47.9 40.9 56.0 39.9 cent in 1988) were derived from 1982 61.7 41.5 44.3 39.9 49.6 36.1 oil, while in Japan well over half 1984 (ranging from 73 percent in 1972 40.6 58.9 42.5 45.9' 44.3 32.3 to 57 percent in 1988) of energy 1986 41.9 56.2 43.8 37.7 42.1 31.4 requirements were supplied by oil 1988 41.0 56.6 42.5 38.1 41.2 31.5 (see Table 4).6 *The downward trend was interrupted by increased oil use during a During much of the 1970s, prolonged coal miners' strike in 1984. the proportion of total energy requirements supplied by oil in 6 TABLE 3 ECONOMIC PERSPECTIVES In sum, the per capita mea sure of oil dependence gives a Primary energy requirements supplied by somewhat different picture than petroleum per capita does the measure of oil require (Million of tons of oil equivalent/million population) ments as a proportion of total Canada Year Japan Germany U.K. France U.S. energy requirements. The pro portional measure suggests that 3.4 1.8 1.9 3.4 1.8 2.1 1970 the dependence on oil relative to 2.0 2.2 3.6 1972 3.7 2.0 2.3 all energy sources is compara 2.2 3.7 2.2 2.2 1.9 1974 3.6 tively low for Canada, France, 2.2 3.9 2.1 2.3 1.6 1976 3.8 and the U.K. The relative oil 1.7 2.2 3.8 4.1 2.2 2.3 1978 dependence of the U.S., accord 3.7 2.1 2.1 1.5 1980 3.5 2.0 ing to this measure, is in the middle of the six countries, and is 1.4 1.7 3.1 1982 3.1 1.8 1.8 especially high for Japan. How 1.6 1.6 2.9 1984 1.8 1.8 3.1 ever, the per capita measure 1.4 1.6 2.9 1986 1.7 1.9 3.1 indicates that the U.S. and Cana 1.4 1.9 1.5 3.0 1988 3.2 1.8 da experience a comparatively high level of dependence on petroleum; substantially lower dependence Canada, France, Germany, and the U.K. gener levels are recorded in the other four countries. ally fell within the range circumscribed by the As discussed in more detail later, these high U.S. and Japan. During the late 1980s, howev dependence levels for the U.S. and Canada are er, the relative degree of reliance on oil as an in part linked to their dependence on transporta energy source by these countries declined so tion and the related large geographical size of that their use of oil as a proportion of total the countries. There is a common thread energy requirements became nearly equal to or, through both measures across countries, howev in some cases, less than that of the United er. Dependence on oil, especially since the late States. 1970s, has declined. Thus, while the United States compared The above discussion implies several con favorably in oil usage as a proportion of total clusions for the issue of petroleum dependence. energy requirements at the outset of the period, First, the response of the U.S. economy to the its economy did not progress as rapidly toward oil price shocks in the post 1973 period appears the replacement of oil with other sources of weaker than elsewhere. During the 1970-1973 energy within the overall energy requirements period the U.S. economy relied proportionately composite as did the others. The per capita measure of oil dependence presents a rather different perspective. Per capita oil requirements, as shown in Table 5 and Figure 3, split the six countries into two packs. Ac cording to this measure, Canada and the United States appear as high oil dependent economies, as they did for total per capita ener gy requirements. Per capita oil requirements in the U.S. and Canada in 1988 (3.2 Mtoe/million population and 3.0 Mtoe/ MP) were double that of France and the U.K. and were more than 50 percent larger than the per capita measures of 1.8 and 1.9 Mtoe/MP in Japan and Germany. TABLE 5 FEDERAL RESERVE BANK OF CHICAGO 7 less on petroleum to meet its total energy needs than did any of the other five countries. In the aggregate, U.S. oil use accounted for 44 percent of U.S. energy requirements, well below an average (weighted by TPER by country) of 52 percent for the other five. By 1988, oil’s share of total U.S. energy requirements had declined, but by only 3 percentage points, to 41 percent. The weighted average oil share of total energy requirements for the other five countries de clined 8 percentage points, but at 44 percent remained above the U.S. figure, primarily as a result of the influence of Japan’s continued heavy relative dependence on oil. However, in an absolute sense, that is, in terms of quantity of energy consumed, the U.S. and Canadian economies are heavily dependent on energy in total and on petroleum in particular. Per capita requirements for these two countries are consis tently well above those for the other countries. The reduction in their dependence levels has been substantial, but they have a long way to go to attain levels comparable with the other coun tries. Indeed, the geographical size and the related dependence on transportation of the U.S. and Canada could effectively set lower limits on their dependence levels that are well above those of the other four countries. O th e r energ y sources movement away from oil utilization toward the alternatives. However, within this general pattern, there were substantial differences be tween countries and between energy forms. Coal During much of the 1970-1988 period, coal ranked second to oil as an energy source in the three European countries and Japan. Coal utilization for the six countries in total, like that for oil, increased during 1970-1988 while at the same time coal’s relative importance as an energy source declined. However, the aggre gated figures mask important individual coun try diversions from the overall trend. Total energy requirements derived from coal declined in France, Germany, and the U.K. during the 1970-1988 period. As coal is prima rily a power source for the generation of elec tricity, it is not surprising that the decline in coal energy requirements appears to parallel the increased use in these economies of natural gas and nuclear power. Coal remained an impor tant energy source in Germany and the U.K., accounting for around 30 percent of their total energy requirements in 1988. Along with the absolute decline in coal use in France, Germa ny, and the U.K., the relative importance of coal as an energy source also declined (see Table 6). In the U.S., Japan, and Canada, coal use increased progressively during the 1970-1988 period. In Canada, coal was relatively less Apart from petroleum there are two other major hydro-carbon energy categories (coal and natural gas) and two nonhydro carbon energy categories (nuclear and an agglomeration of hydroTABLE 6 electric-geothermal-solar) that Proportion of total energy met by coal constitute the remainder of pri (Percent) mary energy sources for these Year Japan U.S. Germany U.K. France Canada economies. Given some reason able adjustment period and favor 1970 18.5 37.9 23.0 42.7 23.9 11.1 able relative prices, these energy 1972 17.1 18.3 32.7 33.3 17.3 9.6 sources are potential substitutes 1974 18.1 18.4 32.8 32.1 16.5 8.1 for petroleum products in numer 1976 19.2 17.2 29.7 34.2 16.7 8.4 ous industrial and power genera tion uses. The marginal cost of 1978 18.8 13.4 27.4 32.3 15.8 8.5 these other primary energy sourc 1980 20.6 16.8 30.3 34.3 16.6 9.5 es rose less rapidly than for petro 1982 21.7 19.1 33.2 32.5 16.0 10.8 leum during the decade following 1984 23.1 19.1 31.6 24.4* 13.2 11.9 the initial 1973-1974 OPEC oil 1986 23.2 18.7 28.7 31.9 10.1 10.3 price shock. (In the U.S. this was 1988 27.1 31.7 23.5 18.5 9.2 11.0 partially due to government regu *The sharp drop in share was due primarily to a prolonged coal miners' lation.) Consequently, it is not strike during 1984. surprising that over the period examined there was a relative 8 ECONOMIC PERSPECTIVES important as an energy source TABLE 7 than in the European or Japanese Proportion of total energy met by natural gas economies. During the 19 year (Percent) span, coal utilization increased Year Japan Germany UK. France Canada U.S. apace with the increase in the economy’s total energy require 4.9 1970 31.6 1.1 5.0 5.3 19.0 ments. Thus, in Canada, coal 1972 8.2 30.8 1.2 11.0 6.8 20.5 maintained a rather stable though 7.9 1974 12.2 14.2 20.1 29.0 2.0 comparatively low level share of 9.4 19.4 1976 25.8 3.0 13.5 16.3 total energy requirements— 24.2 15.2 17.7 9.9 19.3 1978 4.5 ranging between 9 percent and 11 1980 26.1 6.1 16.1 20.1 10.9 19.3 percent of the total. 6.7 14.9 1982 25.2 21.1 11.3 19.9 Coal’s role as an energy source in the United States 1984 23.7 22.7 12.1 20.2 8.9 15.6 moved counter to the trend else 11.9 1986 21.7 9.6 15.2 23.0 20.1 where. Indeed, coal was the only 11.4 22.2 9.4 16.2 22.2 20.8 1988 major hydrocarbon-based fuel to record an increased proportional made more plentiful in Europe by the opening of contribution to U.S. energy requirements during natural gas pipelines from the U.S.S.R. and the the period—increasing from 19 percent to 24 development of economically viable ocean going percent of total energy requirements. natural gas tankers during the 1970s. During the On a per capita basis, energy requirements same period, natural gas use doubled in Canada, supplied by coal increased during 1970-1988 in approximately keeping pace with total energy Canada and the United States, remained stable requirements. As a result, only marginal gains in in Japan, and declined elsewhere. As an abso the relative contribution of natural gas to total lute measure, U.S. per capita dependence on energy requirements occurred in Canada. In the coal remains well above that of any of the other U.S., natural gas use declined between 1970 and countries— 1.8 Mtoe/MP as compared with 1.2 1988, with a consequent sharp drop in the rela Mtoe/MP in Germany, the second largest per tive contribution of this energy form to total capita dependent user of coal. energy requirements. In sum, by 1988, coal still retained its Per capita dependence on natural gas in position as the second largest energy source in creased in all countries but the U.S., where it Germany, Japan, and the U.K., and became the declined by 33 percent between 1970 and 1988. second largest source of energy for the U.S. Still, by comparison, the U.S. was highly depen Natural gas dent on natural gas. Only in 1984 did Canada With the exception of the United States surpass the U.S. as a country more per capita and Canada, natural gas was a distinctly minor dependent on natural gas. In 1988, Canada’s factor in the overall energy package during the dependence level on natural gas stood at 2 Mtoe/ early 1970s. This probably was due, in large MP. The United States followed with a depen part, to the lack of known indigenous supplies dence level of 1.7 Mtoe/MP. Of the six coun and the lack of adequate transport facilities. As tries, Japan recorded the lowest dependence level shown in Table 7, prior to the oil crisis of 1973with per capita requirements of 0.3 Mtoe. 1974, natural gas accounted for 30 percent and Nuclear power 20 percent of total energy requirements in the Nuclear energy is the only primary energy U.S. and Canada, respectively. source to record a common pattern across coun Elsewhere, the natural gas contribution to tries over the time frame examined. Table 8 total energy requirements of the respective shows the proportion of total energy require economies ranged from 1 percent in Japan to 5 ments met by nuclear energy. In each country, percent in France and Germany. Increases in nuclear power recorded multiple gains during the the importance of natural gas-derived energy 1970-1988 period, in terms of its total energy from the mid-1970s to mid-1980s were substan equivalents as well as in its proportion of total tial, in total volume as well as proportional energy requirements. terms. This was possible because supplies were FEDERAL RESERVE BANK OF CHICAGO 9 other energy forms. Within this category, hydroelectric energy Proportion of total energy met by nuclear energy was among the earliest energy (Percent) forms harnessed. Despite new Japan Germany U.K. France Canada Year U.S. technologies utilized to extract geothermal and solar power, 0.4 0.8 0.1 0.6 2.8 1970 0.3 these energy sources have not yet 0.9 1972 0.8 0.8 3.1 1.9 0.8 made a widespread impact. For 1.4 1.1 3.5 1.8 1.8 1974 1.6 example, as of 1988, geothermal 2.1 3.9 2.0 2.0 2.5 2.5 1976 power is estimated to have ac 3.3 3.4 4.2 3.0 4.0 3.6 1978 counted for less than 3 percent of 4.1 5.7 6.9 3.8 3.3 3.6 1980 U.S. total energy requirements.7 As shown in Table 9, of the 7.4 5.1 13.0 4.0 1982 5.6 3.9 six countries examined, only 5.2 4.4 7.9 21.9 1984 9.0 6.3 two—Canada and France—re 6.4 27.8 6.8 10.2 9.9 1986 5.5 corded appreciable gains in the 7.4 29.5 10.0 11.8 6.8 1988 6.5 absolute level of energy derived from H-G-S. Only Canada, which in fact relies heavily on hydroelectric The most dramatic of the increases was in power, recorded an appreciable increase in the France, where the nuclear power contribution to share of its total primary energy supplied by total energy requirements rose from less than 1 this source—from 23 percent of the total in percent in 1970 to 30 percent in 1988. The gain 1970 to 27 percent in 1988. in the nuclear share of total energy requirements Canada’s per capita dependence on H-G-S in the other countries was less dramatic but none totaled 2.6 Mtoe/MP in 1988, 27 percent of its theless substantial. Except for the U.K., where total energy requirements. Per capita depen nuclear power accounted for nearly 3 percent of dence on H-G-S energy by the other countries total energy requirements in 1970, nuclear power was well below that of Canada. France ranked generally accounted for less than 1 percent of second with a dependence level of 0.3 Mtoe/ total energy requirements in 1970. Apart from MP, about 9 percent of its total energy require France, by 1988 nuclear power’s contribution to ments. Dependence levels for the remaining total energy requirements ranged from less than four countries were at 0.2 Mtoe/MP or lower. 7 percent in the U.S. to 12 percent in Germany. However, in Japan H-G-S energy accounted for On a per capita basis, France also appears to about 6 percent of total energy. be relatively dependent on nuclear energy. In 1988 it recorded the highest per capita dependence on nuclear TABLE 9 power, at 1.1 Mtoe/MP. Canada, Proportion of total energy met by at 0.7 Mtoe/MP, and Germany, at hydroelectric-geothermal-solar energy 0.5 Mtoe/MP ranked well behind (Percent) France in both per capita and U.K. Canada Year Japan Germany France U.S. relative dependence. The U.S., the U.K., and Japan round out the 1.7 0.6 8.2 23.0 7.3 1970 3.6 list in terms of their dependence 23.7 3.7 7.2 1.2 0.5 6.5 1972 on nuclear power. TABLE 8 1974 Hydroelectric-geothermal-solar (H-G-S) power The proportion of energy requirements met by H-G-S pow er is shown for each country in Table 9. A country’s utilization of H-G-S energy is more heavily dependent on the natural resource base of the country than are the 10 4.0 6.2 1.5 0.5 7.1 25.6 1976 3.6 6.6 1.2 0.6 6.1 23.9 1978 3.4 5.3 1.5 0.6 8.1 25.0 1980 3.5 6.4 1.5 0.6 8.0 25.1 1982 4.2 6.2 1.7 0.7 8.6 27.1 1984 4.2 5.3 1.6 0.7 7.8 28.6 29.8 27.4 1986 3.8 5.3 1.5 0.8 7.2 1988 2.7 5.4 1.7 0.7 8.4 ECONOMIC PERSPECTIVES In summary, energy consumption and energy dependence vary widely across coun tries and by source of energy. Among the ma jor industrial countries, the U.S., along with Canada, recorded levels of dependence on total energy that are comparatively high, as mea sured by per capita requirements. U.S. per capita requirements are on the order of twice those in Western Europe and Japan. At the same time, however, the rate of growth in U.S. per capita energy requirements was generally lower than elsewhere. Dependence on oil in per capita terms for the U.S. and Canada also stands out. In Mtoe per million population, the U.S. and Canada’s oil dependence are considerably higher than the next most dependent country, Germany. To some degree this is likely due to the large geo graphical area of these two countries and the importance transportation plays in their respec tive economic activity (this issue is discussed in more detail below). It is interesting to note that Japan, an economy that recorded the lowest per capita dependence on total energy of the six countries, was the only economy of the six that in 1988 recorded a per capita dependence on oil equal to 1970 levels (still only 1.8 Mtoe/MP) although by this measure its oil dependence had declined from the higher levels in the late 1970s. Only Canada and France moved signifi cantly away from hydrocarbon energy forms during the period examined. Both developed a strong reliance on nuclear and H-G-S energy forms while the other four countries remained heavily dependent on the various forms of hydrocarbon energy. Energy e ffic ie n c y Energy is a ubiquitous factor-input in any industrial/consumption oriented economy. An understanding of how well or how efficiently energy is utilized in the output of any economy is a key variable in examining the energy envi ronment. Efficiency in the utilization of energy inputs, and differentials in energy efficiency across countries, may explain in part why one economy is more dependent on energy, or on certain forms of energy, than is another econo my. It should also be expected to be a signifi cant factor contributing to the overall and rela tive productivity of the economies. The concept of efficiency is based on the relationship between the physical inputs in FEDERAL RESERVE BANK OF CHICAGO production and the resulting level of physical product. Measures of energy-use efficiency are easily enough derived where there are well defined inputs and outputs. Unfortunately, the physical product (output) of an economy is not so neatly defined. The closest such output measure is in the form of gross national product (GNP) or gross domestic product (GDP) adjust ed for inflation to give real GNP:GDP.8 Thus, given measures of aggregate energy inputs (total primary energy requirements) and eco nomic output (real GDP) a technical efficiency measure can be defined as: (1) E = GDP / TPER ; where E. t is the technical efficiency level for country i, at time t\ GDP t is gross domestic product valued in billions of home currency at constant prices for country i, at time t; and TPER is total primary energy requirements in millions of ton oil equivalents (Mtoe) for coun try / at time t. Technical efficiency is adequate for a within-country measure of “home country” efficiency, but is clearly meaningless for analy sis of relative changes in cross-country efficien cy or in an analysis of relative levels of effi ciency across countries. One problem is that cross-country currency exchange value is not taken into account in measures of technical efficiency. For example, the $2.3 billion per million tons oil equivalent (Mtoe) technical energy efficiency for the U.S. in 1988 cannot be meaningfully compared with the DM7.2 billion/Mtoe level for Germany. Energy effi ciency, in particular, output, must be measured in common units in order to compare countries. However, finding a common base to use in computing energy efficiency measures raises a new set of problems. This article examines four efficiency mea sures, the results of which are described below. Technical efficiency level (TEL), or home country efficiency, as described above, is a measure of the relationship between an econo my’s output (in price adjusted GDP, valued in terms of the home country currency) relative to the economy’s energy input (all energy forms are converted to oil equivalents). The technical efficiency ratio (TER) re moves the units of measure (that is, value of home country GDP per quantity of oil equiva- 11 lents used) from the technical efficiency level measure, and thus allows cross-country com parisons in terms of rates of change from some common base period. This ratio is derived from the index of a country’s GDP divided by the index of its energy inputs. The base period is defined in the home country GDP index and the energy inputs index. Except where other wise noted these indexes are set equal to 100 for the period 1970-1972. This measure suffers from the standard problems associated with indexes. In particular it is devoid of informa tion about the level of efficiency (that is, the value of GDP output relative to the quantity of energy input) across countries. Further, it de pends critically upon the countries’ relative energy efficiency positions in the base year. The ratio reflects comparative developments in energy efficiency across countries as compared to their relative positions as of the base period. However, despite its shortcomings, this mea sure arguably provides the most meaningful basis for cross-country comparisons of energy efficiency. Observed energy efficiency is obtained by converting TEL to a U.S. dollar base using annual average market exchange rates. I refer to this efficiency measure as “observed” be cause it is based on an observed exchange rate. The observed rate is the market exchange rate that is typically, though not necessarily appro priately, used to convert various countries’ outputs to a common currency base as a means to facilitate cross-country comparisons. Purchasing power parity (PPP) energy efficiency is calculated by converting the TEL to a U.S. dollar base using PPP exchange ratebased estimates of GDP. Because PPP rates are based on the relative real purchasing power of currency units, a relationship that changes slowly, period-to-period movement in PPP exchange rates is far more constrained than that of market exchange rates. Consequently, PPP energy efficiency should be less volatile than that for observed energy efficiency. The last two measures of energy efficiency represent an attempt to express levels of energy efficiency across countries. Such measures would be useful, for example, in the analysis of cross-country productivity. This article pre sents efficiency measures based on these for mulations but the severity of the limitations inherent in such measures merits more than 12 cursory examination. To that issue I now turn. One of the problems with the observed efficiency measure is that the wide fluctuation in exchange rates during the past 20 years has exerted a profound influence on the measure of efficiency, an influence that does not reflect changes in the relative output or the relative welfare across the economies. For example, during the period examined, the annual average for the German mark/U.S. dollar exchange rate ranged from 3.65 DM/$ in 1970 to 1.76 DM/$ in 1988. During 1988, the DM/$ exchange rate ranged from a low of 1.57 DM/$ to a high of 1.90 DM/$. If no change whatsoever had oc curred in relative technical efficiency between the U.S. and Germany, the change in exchange rates would have implied a reduction in the level of U.S. observed energy efficiency by more than 70 percent between 1970 and 1988. Prima facie, this is not a plausible conclusion. A decrease in the energy efficiency level of that magnitude implies a concurrent deterioration in U.S. welfare relative to Germany. There is no evidence that such a shift in relative welfare occurred. Another possible approach is to use an exchange rate that provides a ratio of exchange between two currencies such that a specified value of either currency would purchase the same bundle of goods in either country. Econo mists refer to this construction as the purchas ing power parity (PPP) exchange rate. The OECD estimates that in 1988 the GDP purchasing power parity exchange rate between the German mark and the U.S. dollar was equivalent to 2.44 DM/$, as compared with the annual average DM 1.76/$ market rate. From an economic perspective, a PPP exchange rate would appear to be the theoretically correct rate to use when converting economic output mea sures of foreign countries to common dollar base. Economists agree that market rate devia tions from PPP should be expected because PPP is a long-run concept, while the market rate is short-term. Still, the last time the DM/$ ex change rate approached 2.44 DM/$ was in January 1986, when the rapidly depreciating dollar “passed through” on its way down from the exchange rate highs reached during the first quarter of 1985. While it may be argued that the PPP rate is the theoretically correct rate over the long term, the market rate has seldom been even remotely in line with PPP rates. Market decisions are not ECONOMIC PERSPECTIVES based on PPP based exchange rates. A firm’s management does not look at the international com petitive ability of its firm in terms of PPP exchange rates. It seems a reasonable question then wheth er efficiency measures based on PPP conversions are any more economically meaningful than observed efficiency.9 Clearly, if levels of energy efficiency are a vital consider ation, the analyst is faced with an unpleasant choice of tools. As will be seen shortly, over the last decade, market exchange rate movements have overwhelmed technical efficiency changes. While some discussion of effi ciency levels follows it is emphasized that it is not the intent of this paper to focus undue atten tion on such measures. Their inclusion is in tended primarily to be illustrative of the diffi culty in developing an economically meaning ful measure of petroleum efficiency levels. Results As shown in Table 10 and Figure 4, techni cal efficiency improved substantially in each of the six countries during the 1970-1988 period. Not surprisingly, most of the improvement occurred during the 1980s as changes in eco nomic structures, prompted by the 1973-1974 and 1979-1980 oil price shocks, filtered through the economies. By 1988, technical efficiency gains in the six countries ranged from a low of 14 percent in Canada, relative to its 1970-1972 average, to a high of 39 percent in the U.K. Performance of the U.S. economy compared favorably with respect to the remaining coun tries; its technical efficiency ratio rose 31 per cent from its 1970-1972 average—a more rapid gain than in France and Germany but slower than in Japan. Table 10 shows that observed efficiency varied broadly across countries and illustrates the dramatic influence of movements in ex change rates. As can be seen in Figure 5, the impact of the dollar appreciation during the 1980-1985 period and the subsequent deprecia tion during 1985-1987 is clearly outlined in the data. Canada is the exception, where exchange rate movements were less pronounced. TABLE 10 Change in energy efficiency (Efficiency levels in billions of dollar/Mtoe) U.S. Japan Germany U.K. France Canada Technical efficiency Percent change (1970 to 1988) 31 38 24 39 20 14 Observed efficiency 1970 (level) 1.70 1.73 1.54 3.10 3.61 1.71 1988 (level) 2.31 7.05 4.13 3.43 4.11 1.71 31 141 99 10 13 0 1970 (level) 1.70 2.42 1.71 4.53 4.21 1.67 1988 (level) 2.31 4.35 2.97 3.18 3.27 1.67 31 59 55 -35 -25 0 percent change PPP efficiency percent change FEDERAL RESERVE BANK OF CHICAGO 13 Table 10).1 The level of energy 0 efficiency for the U.S. remained well below that of the other coun tries (except Canada) during most of the period. The rate of gain in U.S. energy efficiency based on PPP compares favorably with Canada, France, and the U.K. and lagged behind that of Germany and Japan, though not so severely as when the common valuation measure of output was based on the prevailing exchange rates. Indeed, the most dramatic devel opment coming out of the PPP based data is the deterioration in energy efficiency levels in France and the U.K. Observed efficiency levels in four of the countries (Canada, Germany, Japan, and the U.S.) began the period in relatively close prox imity—$1.5 to $1.75 billion GDP/Mtoe. On the other hand, the French and the U.K. econo mies recorded substantially higher observed efficiency levels in 1970; $3-$3.5 billion GDP/Mtoe. By 1988, observed efficiency levels for the six countries diverged broadly, ranging from $1.7 billion/Mtoe for Canada and $2.3 billion/Mtoe for the U.S. to $7.1 bil lion/Mtoe for Japan. As expected, PPP efficiency levels show change over time that is considerably less dra matic than the fluctuations recorded in the observed efficiency measure (see Figure 6 and 14 Energy e ffic ie n c y by eco n o m ic s ecto r Here to fore the discussion has focused on energy utilization by whole economies. Differ ent sectors of an economy might be expected to be more or less dependent on energy and more or less efficient in their use of energy. Large efficiency gains in certain sectors might be expected to positively influence the competi tiveness of those sectors relative to other sec tors, or relative to similar sectors in other coun tries. The limited data available suggest, not surprisingly, that efficiency differentials exist across sectors of an economy as well as across countries. This article focuses upon two different types of comparison. First, it looks at two broad economic sectors defined as “industrial” and “nonindustri al.” The second comparison examines a sector classification defined by “all transportation” and “nontransportation.” The industrial/nonindustrial sector analysis examines three countries—Germany, Japan, and the United States—for the period 1974-1988. The industrial sector includes a broad aggregation of manufacturing, construction, and mining and quarrying. The non industrial sector includes all other sectors of the economy. This particular sector and country breakdown is used to facilitate the construction of efficiency ECONOMIC PERSPECTIVES measures, as GDP (output) and energy consump tion (input) data are available by these aggregate economic sectors. Efficiency measures are calculated as the ratio of the dollar value of gross domestic prod uct generated by industrial and nonindustrial sectors to total energy consumption by these two sectors. As in the earlier discussion, com parisons are based on measures of technical efficiency levels (TELs) and dollar output mea sures using prevailing exchange rates and PPP exchange rates, which yield measures of ob served energy efficiency and PPP energy effi ciency, respectively. Within countries, GDP output by sector, valued in the home currency relative to units of energy consumed by the sector, indicate an interesting diversity in efficiency trends. It was expected that TELs (the technical efficiency level) would be higher in the industrial sector than in the nonindustrial sector. This pattern did indeed emerge, but did not hold universally. In the United States the nonindustrial sector recorded a TEL of $1.8 billion/Mtoe, slightly higher but probably not significantly different from the $1.7 billion/Mtoe recorded for the in dustrial sector. By 1987, the TELs for both sec tors were identical at $2.5 billion/Mtoe. In Germany, the industrial sector began the period with a TEL of DM6.0 billion/Mtoe, well below the DM7.3 billion/Mtoe for the non industrial sector. However, rapid efficiency gains in industry during the late 1970s and early 1980s pushed Germany’s industrial TEL to DM9.2 billion/Mtoe, substan tially higher than the DM7.8 billion/Mtoe TEL for the nonin dustrial sector. Among the most interesting developments were the relative sector levels and changes in tech nical efficiency derived from the Japanese data. In 1974, in dustry’s TEL stood at Y493 billion/Mtoe, far lower than the nonindustrial Y 1,106 billion/ Mtoe. By 1987, nonindustrial TEL still remained above that of industry, but industry’s dramatic increase in energy efficiency brought its TEL to Y 1,092 bil lion/Mtoe compared with Y1,266 billion/Mtoe for the nonindustrial sector. FEDERAL RESERVE BANK OF CHICAGO One result which met the expectation of higher industrial TELs in each of the three economies was the more rapid rates of gain in energy efficiency in the industrial sector as compared with the nonindustrial sector. The most rapid TEL gains in industry occurred in Japan, up 80 percent, and Germany, up 42 percent, as compared with a gain of 36 percent for U.S. industry. Higher energy costs could be expected to stimulate increased efficiency in that sector, relative to the nonindustrial sector. The industrial sectors in Japan and Germany account for a somewhat larger proportion of GDP, 42 and 36 percent respectively, in 1974 as well as in 1987, than in the United States where industry’s proportion dropped from 33 percent in 1974 to 30 percent in 1987. It re mains puzzling, however, as to why the indus trial TELs in Germany and especially Japan compared poorly with nonindustrial TELs in the early 1970s. Another interesting development is most easily seen in a graph of the technical efficien cy ratio (TER). Recall that this ratio is the index of the home country GDP divided by the index of energy input. In this case the base period is 1974, the first year of the data set. TERs for the industrial and nonindustrial sec tors in the U.S., Japan, and Germany for the period 1974-1988 are shown in Figures 7 and 8. Figure 7 shows that in the nonindustrial sector the U.S. performed relatively better than the economies of either Japan or Germany. U.S. technical efficiency in this sector rose 31 per- 15 cent between 1974 and 1987 while nonindustri al energy efficiency in Japan and Germany rose 13 percent and 7 percent, respectively. As shown in Figure 8, the TERs for the industrial sector indicates that in this case the U.S. did not fair so well relative to gains in Japan and Germany. The TER for U.S. indus try rose 35 percent compared with 31 percent for the nonindustrial sector. The ratio’s gain in Germany was somewhat greater (42 percent) and in Japan substantially greater (80 percent). The United States’ overall TER, buoyed up by gains in the nonindustrial sector, showed a gain (up 32 percent) about midway between Japan (47 percent) and Germany (19 percent). Measures of observed and PPP efficiency levels for the industrial and nonindustrial sec tors were subject to the same dominating influ ences of exchange rate movements noted earli er. Throughout, the PPP based and observed efficiency levels for industrial and for nonindustrial sectors in Japan and Germany were well above efficiency levels in the United States, indicating prima facie that the U.S. lost ground in energy efficiency in the industrial sector as well as in the nonindus trial sector. However, two-thirds of the gain in observed energy efficiency in Japan’s industrial category and nine-tenths of the gain in its nonindustrial category, for example, were due to changes in market exchange rates. Ex change rate movements had a similar effect on Germany’s observed efficiency measures for industrial and nonindustrial sec tors. Once again, exchange rate movements dominate the data. As noted earlier, a country’s physical size and the dispersion of its population may affect energy efficiency. In particular, one would expect a country with a comparatively high level of economic activity in a small geograph ic area to have a comparatively high level of energy efficiency, due to lower energy expend ed on transportation. That is, countries with greater population densities should be more energy efficient, other things being equal. As shown in Table 11, the U.S. and Canada have substantially smaller population density per square mile than the other countries in the sam ple. This may explain in part why the U.S. and Canada fair relatively poorly on energy effi ciency measures compared to the other coun tries. In order to investigate this hypothesis, we examine the development of energy efficiency TABLE 11 Selected demographic characteristics (1988) U.S. Japan Germany UK. France Canada 3,615.1 144.0 96.0 94.2 212.8 3,851.8 246.3 122.6 61.4 57.1 55.9 26.0 68.1 851.5 639.5 606.0 262.5 6.7 Geographic area, (000) sq. mi. Population, millions Population density per sq. mi. 16 ECONOMIC PERSPECTIVES in the transportation vs. nontrans portation sectors over the 19701988 period. Two formulations of the data are examined: Both are modifi cations of previously discussed measures—technical efficiency and the per capita utilization measure used above in the energy dependence section. This analy sis relies on OECD data that facilitate a breakdown of energy consumption by source of energy into total transportation and nontransportation sectors." Because transportation relies primarily on oil, rather than all energy, oil is used as the energy source mea sure. The reader should be aware that the data limitations cited in note 11 mean that the technical efficiency measures constructed here are not comparable with those constructed in earlier sec tions of the paper. The TERs for the transportation/nontransportation sectors shown in Figures 9 and 10 and Tables 12 and 13, respectively, suggest some interesting relation ships. First, gains in oil efficien cy in the transportation sector were well below those for nontransportation. This is not sur prising because the opportunity for the substitution of alternative energy sources is greater for nontransportation than for trans portation. Figure 9 also indicates that these ratios tended to remain closely bundled until the 1979-1980 oil price shock, after which the ratios for the U.S., Canada, and to a lesser degree Japan, broke from the pack. A somewhat surprising result was that three of the six countries (France, Germany, and the U.K.) recorded TERs for transportation that declined and one country (Japan) recorded a transportation efficiency ratio that increased only modestly. During the 1970-1988 period, percentage changes in the transportation TERs ranged from a decline of 13 percent for Germa ny to an increase of 27 percent for Canada. The gain in the U.S. TER was 20 percent. FEDERAL RESERVE BANK OF CHICAGO In those countries where gains were record ed the data suggest that it took the transporta tion sector some time to adjust to the initial shock of higher oil prices in 1973-1974 and the subsequent shock in 1979-1980. The major gains occurred post-1980. The pattern of change in the TERs for the nontransportation sector was markedly different from that of the transportation sector (see Fig ure 10). All six of the countries recorded sub stantial gains in oil efficiency in nontransportation—gains ranged from 64 percent in the U.S. to 117 percent in the U.K. One might expect that the heterogeneous nature of this sector, with its greater diversity of potential energy 17 TABLE 12 Technical energy efficiency ratio for the transportation sector (1970-1972 average=l) Year U.S. Japan Germany U.K. France Canada 1970 1.02 1.03 1.04 1.02 1.04 0.98 1972 0.99 1.00 0.98 0.98 0.97 1.00 1974 1.01 0.95 1.02 1.02 0.95 0.98 1976 0.97 0.92 0.99 1.03 0.88 0.99 1978 1.01 0.91 0.91 1.00 0.90 1.01 1980 1.12 0.94 0.92 0.97 0.91 0.98 1982 1.12 1.02 0.92 0.98 0.92 1.11 1984 1.18 1.04 0.92 0.96 0.91 1.22 1986 1.21 1.07 0.91 0.95 0.91 1.30 1988 1.22 1.06 0.88 0.93 0.89 1.30 TABLE 13 Technical energy efficiency ratio for the nontransportation sector (1970-1972 average=l) Year U.S. Japan Germany U.K. France Canada 1970 0.98 1.04 1.01 1972 1.00 1.02 0.99 1.17 1.01 0.96 1.20 0.99 1974 1.11 1.00 1.08 1.37 1.11 1.02 1.06 1976 1.11 1.13 1.11 1.56 1.24 1.28 1978 1.12 1.24 1.16 1.68 1.29 1.32 1980 1.28 1.56 1.39 2.16 1.48 1.45 1982 1.49 1.92 1.70 2.42 1.92 1.77 1984 1.70 1.92 1.76 2.76 2.10 2.11 1986 1.81 2.07 1.70 2.94 2.27 2.28 1988 1.89 2.14 1.96 3.22 2.51 2.42 sources and substitutability, contributed to the progressive improvement in the efficiency ratio throughout the period. The second approach to examining the geographical size/transportation issue looks directly at the geographical size component of the economies. The per capita consumption measures used in the energy dependence dis cussion earlier is modified to incorporate coun try size. The modification is accomplished by constructing a standard population density series for each country across the 1970-1988 period. This results in two series per country— oil consumption relative to density for the transportation sector and oil consumption rela 18 tive to density for the nontranspor tation sector. The data indicate that low population density does appear to go hand-in-hand with high oil consumption. Both the U.S. and Canada recorded much higher consumption to density figures than did the other countries (see Tables 14 and 15) These data also indicate that oil consumption in transportation, relative to popula tion density, increased through out the period in France, Germany, Japan, and the U.K. On the other hand, the U.S. and Canada record ed declines in consumption rela tive to density in transportation from the late 1970s, although the data showed an up-tick in 1988. In the nontransportation sector the geographical size of the U.S. and Canada also appear to domi nate the data. The data for the U.S. does indicate a decline in oil consumption relative to popula tion density, albeit from compara tively high levels. The U.K. and France also recorded reductions in oil consumption relative to popula tion density. In short, it would appear that geographic size does influence an economy’s level of oil use efficiency in the transportation sector, and also in the nontranspor tation sector. C onclusion Dependence on energy is a fact of life for the world’s economies. How dependent and on what energy forms that dependence relies is not universally alike. On the contrary, there appear to be substantial differences across countries in their level of aggregate dependence, the form of that dependence, and how they have responded to changes in the energy environment following the 1973-1974 oil shock. While oil continues to dominate the energy picture in each of the six countries, each of the countries has reduced its relative dependence on oil, at least from those periods of highest depen dence in the late 1970s. But it is also the case that the alternative energy sources toward which ECONOMIC PERSPECTIVES example, nuclear and H-G-S ener gy provide 38 and 35 percent, Oil consumption by transportation sector respectively, of France’s and Cana relative to population density da’s energy requirements. France (Millions of tons of oil equivalent/population density) in particular has moved well away Year Japan Germany U.K. France Canada U.S. from dependence on hydrocarbon fuels toward nuclear power during 0.05 0.05 0.05 0.09 5.25 1970 6.11 the period examined. Not only did 1972 6.65 0.05 0.05 0.05 0.10 5.68 it maintain a comparatively low 1974 6.67 0.05 6.28 0.05 0.05 0.11 per capita total energy requirement 7.06 0.06 0.05 0.05 0.12 6.47 1976 but it also maintained a compara 7.34 6.57 1978 0.06 0.06 0.06 0.13 tively low reliance on oil and hy drocarbon fuels in general. In 6.34 0.07 0.07 1980 0.06 0.13 6.90 relative terms, Canada moved well 6.44 1982 0.07 0.06 0.06 5.87 0.13 away from hydrocarbon fuels as a 1984 6.62 0.07 0.07 0.14 5.77 0.06 general category, but because of its 0.07 0.07 0.14 1986 6.73 0.07 5.77 high per capita total energy re 7.07 0.07 1988 0.08 0.08 0.15 6.21 quirements, the highest of the six countries, its dependence on oil remained high. TABLE 15 An economy’s reliance on Oil consumption by nontransportation sector energy depends on numerous fac relative to population density tors. Central to how an economy (Millions of tons of oil equivalent/population density) responds to shocks in prices or the Year Japan U.S. Germany U.K. France Canada availability of its energy resources is how efficient the economy is in 1970 15.25 0.23 0.23 0.20 0.43 14.54 energy utilization. Standard tech 1972 15.54 0.24 0.25 15.54 0.20 0.45 nical efficiency measures indicate 1974 0.27 0.24 15.02 0.20 0.45 16.63 that each of the six economies have recorded substantial overall 1976 14.79 0.26 0.25 0.19 0.42 16.96 gains in technical efficiency. In 1978 15.03 0.25 0.25 0.19 0.45 17.39 the U.K. the efficiency ratio for 0.24 1980 14.30 0.17 0.44 0.25 17.56 GDP-to-energy input stood 39 12.71 1982 0.22 0.22 0.17 0.39 16.18 percent higher in 1988 than in 1984 0.24 13.03 0.23 0.16 0.40 16.60 1970. In the U.S., which ranked 1986 12.40 0.23 0.24 0.17 0.39 16.69 third in overall efficiency gains 1988 13.18 0.25 0.17 0.23 0.38 17.42 behind Japan, technical efficiency was up 31 percent. As one would expect, these these economies have shifted tend to be hydro gains are not uniform across sectors within an carbon fuels, specifically coal and natural gas. economy and the pattern of gains across sectors The U.S. increased its relative dependence on varies considerably between countries. Gains in coal. Germany, Japan, the U.K., and to a lesser technical energy efficiency in U.S. industry were degree France, increased their relative depen only modestly greater (up 35 percent between dence on natural gas. Indeed, as of 1988, hy 1974 and 1987) than for the nonindustrial sector drocarbon fuels continued to provide 85 percent (up 31 percent). In Japan and Germany, techni to as much as 92 percent of total fuel require cal energy efficiency gains in industry were ments in Japan, Germany, the U.S. and the dramatically larger (up 80 percent and 42 per U.K. (in 1970, hydrocarbons provided well over cent, respectively) than in the nonindustrial sec 90 percent of fuel requirements in each of these tor (up 13 percent and 7 percent, respectively). countries). This differential in technical efficiency gains Only in France and Canada do nonhydro could be expected to be a factor in maintaining carbon fuels constitute a conspicuous portion of or enhancing international competitiveness by their economies’ energy sources. In 1988, for TABLE 14 FEDERAL RESERVE RANK OF CHICAGO 19 reducing energy input costs, thus possibly helping to offset the adverse competitive impli cations, for Germany and Japan, of the dollar’s depreciation in foreign exchange markets. Several points stand out from this examina tion of energy dependence and efficiency: The major industrial economies continue to be heavily dependent on oil and other hydrocarbon fuels. Among those countries, Canada and France have made substantial strides in shifting their dependence to nonhydrocarbon fuels. In the aggregate, energy is more efficiently uti lized than it was prior to the 1973-1974 oil shock. The gains in energy efficiency in the U.S. have been spread rather evenly across the industrial and nonindustrial sectors of the econo my. The efficiency gains in Japan and Germany were primarily in the industrial sectors of the two economies. Much work remains to be done concerning the issues of energy dependence and efficiency. Because of the problems noted above concerning the measurement of efficiency levels in cross country analysis, there is need for further work concerning the measurement of efficiency, as well as further study of the impact of geographi cal size on energy utilization and the impact of prices, environmental concerns, and government polices on energy use and efficiency. FOOTNOTES Currently, the Congress is considering an omnibus energy production and conservation bill, the major focus of which is to promote increased domestic oil production (and reduced dependence on foreign oil) by relaxing drilling restrictions in Alaska and in offshore areas. Also under consideration are measures to decrease oil consumption through administrative auto mileage requirements and another token increase in the gasoline tax of 5 cents per gallon. 2One might assert that this fragility is also due to an appar ent lack of appreciation by policy makers of their econo mies’ dependence on energy, especially petroleum, for continued economic viability. This is exemplified by a lack of will in some countries, especially the U.S., to apply significant economic disincentives (e.g., gasoline taxes) to the consumption of energy and oil. The preference instead is for administrative distortions to the market place. 3Measures of energy utilization used in this study draw on the Organization for Economic Cooperation and Develop ment (OECD) definition of domestic “Total Primary Ener gy Requirement” (TPER) and “Total Final Consumption” (TFC). Where a common energy unit is required in the analysis the OECD’s common energy unit, “tons of oil equivalent,” usually measured in millions (Mtoe) is used. During any given period, TPER is defined as the sum of a country’s internal production of all energy resources, plus imports, less exports, less international marine bunkers, plus or minus inventory changes of these resources. This measure differs from TFC primarily in that TPER includes energy used in the transformation process, e.g., coal to electricity, and distribution losses as in the transmission of electricity. Energy forms also differ between TPER and TFC. Nuclear or solar energy contribute toward fulfilling a country’s energy requirements, TPER, but are not used directly in consumption, TFC. Nuclear or solar energy is consumed in the form of electrical energy, and thus does not appear directly in energy consumption. 20 The OECD defines a “ton of oil equivalent,” where ton refers to a metric ton (2,204.6 U.S. pounds), as equal to 107 kcal. of energy. All energy forms, be they petroleum, nuclear power used to generate electricity, or electricity consumption itself are converted to the common unit “t.o.e.” In this article units will be reported in millions of ton oil equivalent (Mtoe). Percentage changes throughout the article are reported on a logarithmic basis. 5Import dependence has implications for a country’s international balance-of-payments. The larger the oil import requirement needed to sustain the economy, the greater the real resources required to finance the importa tion of oil. Other things remaining the same, a lower standard of living results for the oil importing country than if the oil could be sourced at equal cost domestically. Import dependence is also a concern for national security. The greater a country’s dependence on imported oil the less independent it is from the political or economic whims of its foreign oil suppliers. It is clear that in the current political/economic environment, energy and petroleum security is vital. From a near-term perspective, it may be undesir able to be dependent on the political or economic whims of foreign oil producers, however, one must also be aware of longer-term security issues. Arguably, a nation’s energy sources would be more secure the less its dependence on foreign supplies. During 19891990, 40-45 percent of U.S. petroleum consumption was derived from foreign sources. On the other hand, in an environment of limited and relatively high cost domestic supplies and relatively inexpensive foreign supplies the utilization of imports serves to conserve and prolong those limited domestic supplies, should a real emergency devel op. U.S. petroleum independence in the near-term—the relatively more rapid depletion of domestic supplies—risks the possibility of becoming more heavily dependent on ECONOMIC PERSPECTIVES foreign supplies in the future, barring major technological innovation, such as, for example, economically viable solar or fusion power. From an energy security perspective then, it is not a clear cut decision that reduced import dependence now is preferable to necessarily greater dependence later. This is an argument that policy makers must not continue to ignore. 6It is interesting to note how the region of source has changed over time. From 1973 to 1988 the proportion of total oil consumption by the six countries that was derived from the Persian Gulf states generally declined: For Japan, from 74 percent to 55 percent; for France, from 84 percent to 28 percent; for the U.K. from 66 percent to 15 percent; for Germany, from 49 percent to 9 percent; and for Canada, from 19 percent to 4 percent. The U.S. share increased from a comparatively low level of 6 percent to 9 percent. ’American Gas Association (1990), p. 124. ®Gross domestic product is used as the measure of a coun try’s output in order to obtain a more consistent data series. GDP is the more commonly used output measure abroad. GDP differs from GNP in that GDP excludes net factor income from abroad. In U.S. national income statistics, for example, most of the net factor income from abroad is in the form of U.S. firms’ corporate profits from abroad— profits earned abroad are not part of U.S. GDP, but are included in GNP. ’Other economists, for example, Summers and Heston (1984), have done extensive work that has focused on developing meaningful measures of real national product across countries. ^Organization for Economic Cooperation and Develop ment (1991). n OECD data are available for energy and oil consumption by transportation (air, road, and rail). An important restric tion in the transportation case is the lack of contribution to GDP by sector data, as was the case for industrial/nonindustrial sectors—that is, contribution to GDP by the trans portation sector and contribution to GDP by the non transportation sector. The technical efficiency measures, therefore, do not refer specifically to this sector. Rather, they are a hybrid that relates total GDP output to energy inputs for transportation/nontransportation. REFERENCES American Gas Association, 1990 Gas Facts, 1990 Central Intelligence Agency, International Energy Statistical Review, Washington, select ed issues. Energy Information Administration, Interna tional Petroleum Statistics Report, U.S. Depart ment of Energy, Washington, selected issues. International Energy Agency, Oil and Gas Information, Organization for Economic Coop eration and Development, Paris, selected issues. Organization for Economic Cooperation and Development, Energy Balances of OECD Countries, Paris, selected issues. Organization for Economic Cooperation and Development, National Accounts 1960-1989, Main Aggregates, Volume 1, Paris, 1991. Organization for Economic Cooperation and Development, OECD Economic Outlook, No. 48, Paris, December 1990, pp. 35-46. Summers, Robert and Alan Heston, “Im proved international comparisons of real prod uct and its composition: 1950-1980,” Journal of Income and Wealth, June 1984, pp. 207-262. Worldmark Press, Worldmark Encyclopedia of the Nations, John Wily & Sons, Seventh Edition, Volumes 3, 4 and 5, New York, 1988. Organization for Economic Cooperation and Development, National Accounts, Detailed Tables, Vol. 2, Paris, selected issues. FEDERAL RESERVE BANK OF CHICAGO 21 Balancing act: Tax structure in the Seventh D istrict Richard H. M atto o n £ gan and Wisconsin—focusing on their utiliza State and local governments tion of the property, income and general sales are facing their toughest fiscal tax. Finally, it will review the tax options now situation since the recession under consideration to relieve state and local years of the early 1980s. As fiscal pressure. tax revenues throughout the country fall below projections, state and local C rite ria fo r ju d g in g a good governments are grappling to find the right s tru c tu re ta x course between spending cuts and tax increas Requirements for a good tax structure can es. Owing to budget problems of its own, there be distilled from a large body of analysis and is little hope that the Federal government can thought.1 Criteria often include the following: help the states in these recessionary times. • The distribution of the tax burden should be As state and local governments begin to equitable, with everyone paying a fair share. examine their expenditure and revenue options, it would be wise for policy makers to bear in • Interference with the operation of efficient mind what constitutes a good tax system, as private markets by taxes should be minimized. well as the major advantages and disadvantages In particular, taxes should not distort econom of the tax sources available to state and local ic choice by placing excess burdens on indi governments. viduals. Many analysts have described what they • Taxes can, however, be used to correct ineffi believe the objectives of a good tax structure ciencies in the private sector in situations should be. While the criteria developed share where markets do not behave efficiently. many common concerns, no two lists are exact ly alike. Depending on which aspect of the tax • The tax structure should help stabilize the structure receives emphasis, goals can conflict national economy while providing adequate and blueprints for tax policy can become mud revenue growth. dled. However, keeping well-defined criteria at • The tax structure should permit efficient and the forefront can help focus the tax policy de nonarbitrary administration and should be bate and can force policy makers to be explicit comprehensible to the taxpayer. in recognizing the trade-offs that occur in se lecting one tax over another. • Administration and compliance costs should This article will begin by reviewing the be reasonable. criteria often cited for establishing a good tax structure. Against these criteria, it will The author is a regional econom ist at the Federal Reserve Bank of Chicago. The author thanks examine the tax structure of the five Seventh W illiam Testa for guidance and com m ents on District states—Illinois, Indiana, Iowa, Michi previous drafts. 22 ECONOMIC PERSPECTIVES In order to design tax systems, state policy makers often need more detailed guidelines which incorporate strategies specific to state and local government.2 In this regard, Robert Kleine and John Shannon have developed the following guidelines of taxation. • Revenue sources for taxation should be diver sified. In particular, they should balance the three principal tax bases available to state and local government: income, property and sales taxes. • Revenue sources should be stable and taxes should be moderate. This includes using broad tax bases with minimum volatility and insuring that expenditures and state revenues grow at consistent and reasonably parallel rates. • Taxation should be fair and equitable. That is, the tax structure should be progressive enough to protect the lowest income members of society from bearing a disproportionate share of the tax burden. • There should be state fiscal equalization pre venting property tax disparities from creating local fiscal disparities in funding for education and other programs. Specifically, this calls for state government to take a “senior” role in the state and local fiscal system and assume at least 50 percent of the cost for education, health and hospitals, and all of the cost of nonfederal public welfare. In lieu of picking up these costs directly, the state should offer revenue sharing so that communities can fund these programs without sole reliance on the property tax. • Changes to the tax structure should be politi cally accountable. Tax increases should be the product of deliberate legislative action and not inherent structural features of the tax system which permit automatic tax hikes. To cite two examples, “bracket creep” for income taxes and changes in assessment practices for property taxes often permit governments to raise revenues without approval from either taxpayers or their representatives. In contrast, indexing income taxes and adopting truth in property tax laws can help safeguard residents from automatic tax increases. • There should be property tax equity, defined as uniform assessments both within and be tween towns. FEDERAL RESERVE BANK OF CHICAGO • State and local governments should strive for tax competitiveness. Tax rates and policies should avoid creating an image of a poor business climate. Each state should be cau tious that its tax policies do not provide incen tives for desired companies to invest in lower tax jurisdictions. Neither of these lists exhaust the possible criteria for designing a good tax system and it is impossible to meet all of these criteria simulta neously. In trying to meet the goal of one as pect of the tax system, trade offs with other goals are bound to occur; compromises must be made. For example, offering tax incentives to business may undermine revenue stability. Many states have created special tax incentives in order to attract and retain business and to create a “competitive tax climate.” In doing so, states narrow the base on their business taxes, often making the revenue performance of the tax more volatile and thereby undermining revenue stability. Similarly, those sales tax policies that exempt “necessities,” such as food and clothing, are usually intended to promote tax fairness. In doing so, the revenue stability of the tax system may again be reduced by narrowing the tax base. Furthermore, some of the goals of a good tax system are ill-defined or not universally accepted. Having everyone pay their fair share is a goal of tax equity; however, not everyone agrees on what constitutes a fair share. Should a person’s fair share be based on their ability to pay or on the benefit they re ceive from the public services provided by taxation? Given that it might not be possible to satis fy all of the objectives listed above simulta neously, the purpose of reviewing lists such as these is to provide the framework for examin ing what specific needs and pitfalls should be considered when trying to reform or modify a state’s tax structure. Revenue systems are reformed because one feature of the system has fallen out of line. Too often, in the process of re-alignment, other important features are ne glected. Currently, revenue adequacy is the one goal of taxation which is of most immediate interest to policy makers. Many states are finding that revenues are not adequate. None of the three major state and local tax bases (prop erty, sales, income) appear to be easy targets for revenue expansion to relieve the currently 23 strained state and local fiscal condition. Yet states are largely limited to the property, sales and income tax bases which have traditionally been the focus of major tax changes. FIGURE 1 Distribution of state/local tax revenue by source percent of total tax revenue S even th D is tric t focu s The District’s revenue structure Seventh District states have established a fairly conventional tax structure, raising the bulk of their tax revenues from property, sales and income taxes. District states differ some what from the rest of the nation in that they rely on property taxes for a greater proportion of revenue. Income and sales tax bases grew faster during the 1980s than the property tax base. As a result, District tax revenues grew more slowly during the 1980s than the rest of the nation. The property tax Table 1 compares the District with the U.S. on property tax reliance. The Advisory Com mission on Intergovernmental Relations (ACIR) tax effort index reported in the Table is designed to measure a state’s tax effort relative to the U.S. (see Box for details). Table 1 shows that, as measured in terms of effective tax rate, share of personal income and the ACIR tax effort index, District states tend to utilize the property tax more heavily than the national average. It is therefore not surprising that in terms of revenue raised, the property tax com prises the largest share of state and local tax revenues in the District states, as shown in Figure 1. District S O U RCE: Governm ent Finances: 1988-89, Bureau of the Census. Table 2 compares the growth of the proper ty, sales and income tax revenues for District states and the U.S. during the 1980s. From 1981 to 1989, real property tax revenues in the District grew by only 15 percent while sales and personal income tax revenues were up more than 25 percent, in spite of the fact that the District attempts to tax property more vigor ously than the other tax bases. At the national level, property tax revenues were the slowest growing revenue source with receipts up 31 percent while sales and income were up 40 and 49 percent, respectively. Before discussing the advan tages or disadvantages of the TABLE 1 relative reliance on property Property tax comparisons for the District and the U.S. taxes for District states, it is nec essary to keep several caveats in Effective Share of ACIR tax rate* personal income index mind. It is very difficult to make 1987 1988 1988 meaningful generalizations about how the property tax burden Illinois 1.59 3.8 129 uniformly affects any individual Indiana 1.28 3.2 103 state due to the wide variation in Iowa 1.96 4.5 152 local property tax administration. Michigan 2.26 4.7 157 Given the local nature of the tax, W isconsin 2.27 4.5 147 the property tax burden can be U.S. 1.16 3.5 100 significantly different from town ^Effective tax rate reflects the estimated taxes on a FHA insured to town and can reflect differenc single family home mortgage. es in the level, quality and effi SOURCES:State Policy Data Book 89, Table D-40, Brizius & Foster. States in Profile, 1990, Table D-9, Brizius & Foster. ciency of town services, variation 1988 State Fiscal Effort, 1990, Advisory Commission on Intergovern in the local price of providing mental Relations. services, assessment practices, 24 ECONOMIC PERSPECTIVES vestment is equal in both jurisdic tions. This will leave the high Real tax growth by type of tax: 1981-1989 tax town with a diminished share (Percent change) of capital stock, which can be a General Personal Total major concern for a town’s eco Property sales income tax nomic growth and development.3 The District’s reliance on the Illinois 15.6 19.5 19.1 13.2 property tax probably has little to Indiana 9.7 31.1 109.1 34.9 do with the debate over who Iowa 4.7 27.1 20.4 13.8 bears the burden of the tax. High Michigan 12.8 20.6 29.2 19.5 property tax reliance is usually W isconsin 30.2 46.2 7.6 25.6 related to a variety of other fac tors which relate taxes to the U.S. 31.4 39.9 49.9 32.49 funding of public services ren District 15.0 25.5 27.6 19.58 dered rather than to who bears SOURCE: Government Finances: 1980-81, 1988-89, Bureau the tax burden.. If there is a of the Census. political preference for local governments funding a large share of public education expense or other local and other factors. Simply noting that the aver programs, the property tax burden will be high age property tax burden for a given state is high er. For example, in Illinois, the existence of fails to recognize the extremely varied distribu thousands of special districts and other govern tion of this tax burden. ments explains why the state’s reliance on the Determining the ultimate burden of the property tax is higher than the U.S. average. property tax is also an issue of dispute. Origi Also, property tax reliance is often a product of nally, tax theory held that property tax was historical factors. Agricultural states tend to passed on to tenants and consumers of property. rely on property taxes because, historically, the As such, the effect of the tax showed up in property tax was a good way to relate the tax either the tenants’ rent or in the price of a prop burden to the ability to pay. In an agrarian erty. In practice, this made the property tax economy the more farm land held, the greater regressive with respect to income since lower the ability to produce income.4 income households tend to devote a larger share Critics often attack the property tax on of their income to housing. For this reason, several grounds. The first criticism is that it many tax analysts held that reliance on the serves as a poor proxy forjudging an individu property tax should be held in check due to its al’s ability to pay the tax if the tax is viewed regressive sting. solely as a tax on the consumption of housing. However, more recent theory has held that The property tax is often claimed to be “hori the property tax is actually a tax on capital. As zontally inequitable” in the sense that individu such, its incidence is reflected in a reduced als with equal incomes or wealth rarely have an return on capital investment which ultimately is equal property tax burden due to variations in also borne by labor. Labor is affected because their housing consumption. For example, two the reduction in capital stock caused by the houses may be assessed and taxed at identical property tax reduces labor output. The effect is rates and yet disparities in the incomes of the even more pronounced when there are differen owners can make the burden of the tax minimal tials in property taxes between jurisdictions. on one owner and heavy on the other. For Here a distinction must be made between the example, the tax burden is often heavy for effect the tax has on land vs. its effect on capi senior citizens on fixed incomes who find that tal. While land is immobile, capital is not. property taxes consume a larger and larger Consequently, differentials in property tax rates fraction of income over time. will induce capital to move to locations with Another criticism is that the multiplicity of lower tax rates, thereby increasing the return on taxing districts and variations in tax structure capital. This will cause investment funds to often provides incentives for avoiding the tax. flow out of high tax jurisdictions and into low In states where mobile personal property such tax jurisdictions until the return on capital in TABLE 2 FEDERAL RESERVE BANK OF CHICAGO 25 The Advisory Commission on Intergovernmental Relations Representative Tax System The Representative Tax System (RTS) was created by the Advisory Commission on Intergov ernmental Relations (ACIR) in an effort to allow comparisons of state and local tax capacity and effort between states. RTS measures the amount of revenue that would be raised if a state applied national average tax rates to 27 different tax bases within the state’s boundaries. The figure arrived at through this process defines the hypothetical tax capacity of a given state. It is important to note that the RTS is a hypothetical measure that as sumes all states utilize all 27 tax bases and that all states tax at the same rate. Tax effort in the RTS attempts to determine a state’s relative utilization of its tax base. Tax effort is the ratio of actual state tax collections to the state’s hypothetical tax capacity. A tax effort above 100 indicates that the state’s utilization of the tax is above the U.S. average. For example, Wisconsin’s 1988 tax effort score for the personal income tax was 157. This indicates that the state imposes a tax burden which is 57 percent above the national average. Conversely, a score below 100 indicates that the state’s utilization of the tax is below the U.S. average. For example, Michi gan’s sales tax effort of 76 means that the state’s sales tax effort was 24 percent below the U.S. The RTS provides a basis for standard compari sons among states. as automobiles is taxed, there is an incentive to register cars in towns with low tax rates even if the owner resides in a higher tax town. A fur ther criticism is that gaps in the property assess ment cycle can distort the relative value of both personal property and new construction, forcing both to bear an unfair share of the tax load. This occurs when personal property and new construction assessment reflect current value while real property assessments are allowed to lapse for several years in between revaluations. In some states the gap between real property revaluations can be as much as 10 years. Evidence suggests that high property taxes can also serve as an obstacle to business and economic development, since the property tax can often be the largest tax faced by a business. A recent Wisconsin study found that the prop erty tax constituted 47 percent of the state and local total tax liability for a group of six manu facturing industries.5 This out-distanced the corporation business tax which comprised 35 percent of total state and local tax payments. The view that high property taxes impede eco nomic development is supported by economic theory, which holds that differentially high property taxes will tend to increase the cost of capital, thereby encouraging new investment to seek lower tax jurisdictions. Finally, high property tax reliance can create fiscal disparities between communities. Goals of equity and fairness can be undermined when property poor towns are forced to provide the bulk of their local services through the property tax. The primary defense of property taxation tends to be founded more on political than economic grounds. Because property tax growth is relatively inelastic, or unresponsive to swings in economic activity, it supplies a steady but sluggish revenue source for local governments. This steady growth can be a stabilizing factor for local governments in bad economic times. The relative stability of the property tax has proven to be an advantage during the recent recession. National Income and Product Ac count data placed the property tax as the fastest growing revenue source when measured from the first quarter of 1990 to the first quarter of 1991. Specifically, property tax revenues have grown by 6.5 percent while sales and income tax revenues have increased by only 2.7 and 3.8 percent, respectively. Recessionary effects which are impacting sales and income tax re turns are less likely to cut into property tax gains. Since property tax assessments often lag because of administrative features, this tax source continues to grow as updated assess ments reflect past gains in property values even though current values may have stabilized or declined. This administrative lag can often stabilize government spending in the early periods of a recession and generally reduce 26 ECONOMIC PERSPECTIVES reliance on faster growing but more volatile tax bases such as income and sales. In summary, the District’s reliance on property tax suggests a choice of slower but more stable revenue growth, more costly in vestment in property from a tax perspective and an overall tax structure that is probably more regressive with respect to income than is the case for the U.S. as a whole. Further, while the property tax may have the advantage of increasing local government accountability by requiring that more services be performed at the local level, it reduces the redistributive role of state government by reducing the potential revenue raising capacity of state government. FIGURE 2 Sales tax effort: 1988 U.S. average-100 The general sales tax District wide reliance on the sales tax var ies significantly. While both Indiana and Illi nois draw 25 and 29 percent of total state and local revenues from the sales tax, the remaining three District states all draw less than 20 per cent. This variation is also evident in the rela tive burden of the tax as a percentage of person al income. In Indiana, the sales tax is 3.1 per cent of personal income, in Illinois it is 2.6 percent and in Wisconsin it is 2.5 percent. In Iowa and Michigan the percentage falls to 2.1 percent. When viewed relative to the U.S. average, only Indiana exceeds the national figure of 2.8 percent. This variation occurs despite the fact that state sales tax rates among the five District states are quite similar, ranging from 4 percent in Michigan to a high of 6.25 percent in Illinois.6 The significant difference in reliance has come about because states differ in authorizing localities to levy their own sales tax. Iowa, Illinois and Wisconsin all permit other govern mental units to levy a sales tax in addition to the basic state sales tax rate. In Chicago, for example, the total sales tax rate levied by all overlapping governments is 8 percent. Mean while, just over the state line in Indiana, the rate is 5 percent. Such disparities can create an incentive for people to avoid additional sales taxes by crossing state borders to make pur chases. Even in the ACIR tax effort figures, Dis trict states rank very low nationally on the sales tax effort index (see Figure 2). Only Indiana and Illinois at 105 and 104 respectively are above the national average. When compared to the very high property tax effort scores for FEDERAL RESERVE BANK OF CHICAGO August 1990, ACIR. District states, it is clear that the District pro vides something of a break with respect to consumption taxes. This reduced reliance does not arise because of any smaller tax base in the District. Rather, it most likely reflects differing exemption policies for certain goods (all five District states exempt manufacturing machinery and materials as well as food and prescription drugs from the sales tax) rather than lower spending tendencies by District residents. Low reliance on the sales tax is perhaps the most unusual aspect of the District’s tax struc ture. Nationally, the sales tax is consistently seen as the most popular tax in surveys of tax payers. Much of this popularity is due to ad ministrative features of the tax. The relatively small amount of the tax attributable to the pur chase price of a good makes the tax less visible on all but the largest purchases. In addition to its relative popularity, the sales tax is fairly easy to administer even though sales tax audits are necessary and often expose significant fraud. The sales tax also has the added advantage of being an exportable tax in that it is paid by those nonresidents such as tourists and conventioners who make purchases within the state. One problem often noted with the sales tax is that it is not progressive. That is, if a wealthy person and a poor person buy the same good, the tax bite as a share of income is much greater for the poor person. Attempts to correct this feature and reduce the regressivity of the 27 tax by exempting food and other necessities often fail because the tax break tends to provide relief to both income groups. As a further drawback, exemptions of necessities, such as food, increase the volatility of the tax base by making the base more reliant on “big ticket” consumer goods such as cars. Further, the relative regressivity can be capricious, because it is greatly influenced by an individual’s particular consumption and savings habits. This feature is moderated to some extent because it is possible to escape some of the sales tax burden by lowering the level of consumption and increasing savings. Indeed, some analysts favor this tax because it encourages national savings. But low income individuals have much less latitude in their consumption vs. savings decisions. Much of the decision to consume is based on the relative need for a product. The more price inelastic the demand for a product, the more regressive the sales tax. For goods that are price elastic, such as luxuries, the seller of the product may have to absorb some of the tax burden.7 Neces sities are relatively price inelastic, consequently the buyer must absorb all of the tax burden. By underutilizing the sales tax, District states generally forego revenues from the least politically objectionable tax source. Their gain may be greater revenue stability because they are less reliant on this relatively more volatile base. Furthermore, by permitting local sales taxes, the states provide local governments with an option for revenue diversification which can lessen the dependence on the property tax. However, taxpayers find it easy to avoid the local sales tax by making purchases in lower tax jurisdictions, thereby affecting shopping location decisions. nois at 1.7 percent is significantly below the average. With the exception of Illinois, with its low reliance on the personal income tax (this will change given the recent tax surcharge which raised the income tax rate from 2.5 to 3 per cent), the District tends to conform fairly close ly to the ACIR criteria of raising a quarter of revenues from this source. However, it does so by imposing a slightly higher tax burden on residents relative to the nation. Both the tax as a share of personal income numbers and the ACIR effort index reflect a relatively high burden for this tax (see Figure 3). Differences in reliance among District states arise primarily from differences in the tax structure. Three states—Illinois, Indiana and Michigan—have flat rate income taxes which range from 3 percent in Illinois to 3.4 percent in Indiana and 4.6 percent in Michigan. Iowa and Wisconsin have graduated rate income taxes, with Iowa’s tax consisting of nine tax brackets ranging from a tax rate of .4 percent to 9.98 percent, and Wisconsin’s consisting of three brackets ranging from 4.9 percent to 6.93 per cent. All five states use some variant of federal adjusted gross income as a starting point for the tax base. As a group, none of the District states can be characterized as having generous deduc tions or tax credits in arriving at state taxable income. Both Illinois and Indiana permit only a $ 1,000 standard deduction against individual income before levying their flat rate tax. Iowa and Wisconsin provide limited tax credits. FIGURE 3 ACIR income tax effort: 1988 U.S. average-100 The personal income tax The District tends to be slightly above the national average in utilizing the personal in come tax. When measured as a share of state and local revenues, the District tends to draw on the personal income tax more heavily than the U.S. as a whole. This is also reflected in the burden of the personal income tax as a percentage of income. The U.S. average for state and local personal income taxes as a per centage of income is 2.1 percent. District states above the average include Indiana at 2.3 per cent, Michigan at 2.5 percent, Iowa at 2.6 per cent and Wisconsin at 3.3 percent. Only Illi 28 August 1990, ACIR. ECONOMIC PERSPECTIVES Iowa’s credit is $20 for the filer and $15 for each dependent while Wisconsin limits its tax credit to $50 per dependent. Michigan provides the most generous standard deduction at $2,000. The impact of these limited deductions and credits is to make the income tax more regressive in District states, particularly in those with low standard exemptions and flat rate tax structures, than in those states with progressive rate tax structures or large personal exemptions.8 There are several advantages to raising revenues through income taxes. It is considered the best tax in terms of progressively relating the tax burden to one’s ability to pay. And assuming that the tax uses a fairly broad mea sure of income, it is also horizontally equitable, that is, it treats equally-situated individuals the same. Tax compliance costs are also lessened through conformance with the federal income tax base guidelines. The tax is highly produc tive as a revenue source. As personal incomes rise the tax base grows. As such it can provide a substantial revenue source for the states. Also, because most states choose not to index the income tax base, state income tax revenues can grow through bracket creep in which in creases in personal income push tax payers into higher tax brackets even if their inflation-ad justed incomes have not increased significantly. Since three of the District states utilize flat income tax rate structures, bracket creep is less of an issue within the District. One disadvantage of the personal income tax is that as it reduces disposable income, it can serve as a drag on spending and investment. A relatively higher income tax burden can mean that people will adjust their spending and savings habits in order to absorb the tax. This behavior is likely because the income tax in most cases cannot be transferred or shifted to another party. As such, the District’s slightly heavier burden on income taxes may tend to reduce both consumption and savings. Further more, the preference toward flat tax rates with limited deductions erodes much of the progres sive structure usually associated with income taxes. E x p o rta b ility Tax exportability—the ability to levy taxes in such a way that the burden is borne by outof-state residents—is quite understandably a desirable goal from the point of view of the FEDERAL RESERVE RANK OF CHICAGO state levying taxes.9 The advantage for a state in having a significant share of its tax base exportable is that it lessens the tax burden on state residents. Three states noted for having highly exportable tax bases are Alaska, Nevada and Hawaii. Alaska is able to export a great deal of its tax burden through severance taxes on the sale of oil. Because oil deposits are mostly owned by large multi-national compa nies, which are in turn owned by non-Alaska residents, severance tax incidence falls on outof-state residents. A 1980 study1 found that 0 Alaska was able to export 36 percent of its tax burden largely because of the severance tax. The same study found that Nevada was able to export over 20 percent of its tax burden through the sales taxes on gambling by out of state residents. Hawaii exports taxes by taxing spending by tourists. By comparison, Seventh District states are not so blessed. At present, there is no large industry generating profits for out-of-state resi dents which can be tapped by the tax system. Very little of existing revenue vehicles are able to export the burden. This is due in part to the District’s relatively heavy reliance on property taxes, which are usually not exportable, and the light reliance on sales taxes, which are the most exportable. The 1980 study found that while the average rate of tax exportation for the U.S. was 9.6 percent, the range in the District states was 7.7 percent in Illinois, 6.5 percent in Mich igan, 5.8 percent in Indiana, and 5.4 percent in both Iowa and Wisconsin. In a sense, taxes can also be exported to the federal government when states take advantage of the so-called federal offset permitting deduc tions from the federal income tax for selected taxes paid to states and municipalities. In par ticular, income and property taxes are currently eligible as a deduction for those who itemize deductions on federal returns. District states fare better with respect to federal tax exporta tion because they rely heavily on property and income taxes. In 1980, the national average for the federal offset of state taxes was an estimat ed 7.1 percent. Among District states, the value of the offset was greater in all states ex cept Indiana; 7.5 percent in Illinois, 5.6 percent in Indiana, 7.3 percent in Iowa, 9.2 percent in Michigan and 10.2 in Wisconsin. Sales tax deductions were phased out following the 1986 Tax Reform Act and are no longer a source of 29 federal tax exportability. With the elimination of sales tax deductability, the size of the offset has undoubtedly declined but, given the Dis trict’s lesser reliance on sales taxes, the decline has probably been less precipitous than in high sales tax states. One final trend which also affects the rela tive value of the federal tax offset is the recent popularity in state and local governments of user fees and special charges. While these have the advantage of diversifying revenue sources and directly relating the cost of providing a service to the beneficiary, they have the disad vantage of not being deductible from federal taxes. User fees and charges often replace revenues from deductible sources, such as prop erty and income, which in turn reduces the value of the federal offset. O p tio n s fo r raisin g ta x revenues States are now facing the largest revenue gap since the 1981-82 recession. Then, most states initially tried to avoid major tax increas es. Nevertheless, states entered FY83 with record tax increases. At the beginning of FY83, 16 states raised their income tax, 11 raised their sales tax and 5 raised both. Furthermore, 10 states had a stated goal of trying to increase revenues by 15 percent that year.1 1 While states are today looking at similar options, fears of developing a less competitive tax climate than neighbors have many states attempting to raise revenues through means other than raising major tax rates. Some states, motivated by their current fiscal problems, are reducing tax rates that are perceived as being too burdensome, and replacing these taxes with new ones. For example, Tennessee and Con necticut considered instituting state income taxes and using the money raised to role back the sales tax rate. In Tennessee’s case, the state sales tax rate could be cut from 5.5 to 4 percent and in Connecticut from 8 percent to 4.25 per cent. The belief is that a more balanced tax system will ultimately be of greater benefit to the states’ residents while also enhancing eco nomic development.1 2 In a turnaround of trends from the 1980s, a number of states with progressive rate income taxes are considering raising the top marginal tax rate on individual income taxes. Proposals to do so have been offered in California, Dela ware, Montana, Maine, and New York. For example, a legislative package in California 30 proposed raising the top bracket marginal tax rate from 9.3 to 11 percent. While these pro posals have received some popular support, others have criticized them on the grounds that higher income individuals tend to be more mobile, so that high marginal rates may encour age the wealthy to leave, thereby diminishing the tax base.1 3 The sales tax is the remedy most frequently considered by states to relieve fiscal pressure. Two kinds of changes are being considered. Rather than increasing the tax rate, most states are looking at ways to eliminate tax exemptions for certain items such as snack food and maga zines, thereby broadening the tax base. Califor nia recently added a number of snack foods to its sales tax base, although problems defining what constitutes a snack food are being encoun tered. Alternatively, states are considering an extension of the sales tax base to include ser vices. Service taxation is still an area where states have been moving cautiously. A broadbased taxation of services has yet to emerge although certain professional and personal services are increasingly being taxed. Service taxation has been defended on several grounds. According to one argument, the purchase of services is a form of consump tion just like the purchase of tangible property. There is no reason to treat the choice of one type of consumption differently from another. Consequently, according to this argument, service taxation is a matter of tax equity as well as neutrality. It is unreasonable to penalize those who consume tangible goods rather than services, thus encouraging service purchases rather than goods. As a practical matter, services have not been taxed because of difficulties in administer ing the tax. Because some services, such as housekeeping and lawn cutting, are very infor mal, it is difficult to imagine that a tax on such services could be easily imposed and easily collected. However, other services, particularly those involving the repair and maintenance of tangible personal property, are particularly vulnerable targets for sales taxes. Subscription services such as cable television are also a frequent target for the same reason. From a practical point of view, service taxation is not without advantages. The rapid growth in ser vice activities and in service consumption makes this a particularly attractive potential tax base. Increased service taxes can broaden and ECONOMIC PERSPECTIVES reduce the volatility of sales taxes which have grown dependent on big ticket durable goods purchases. Objections to service taxation are based on the fear that this is such a huge pool of revenue that it will not only relieve current fiscal pres sures, but further encourage government spend ing. Another objection is that service taxation should not apply to many of the services pur chased by business because this would be a form of double taxation, i.e. the physical output (which embodies the service) is ultimately taxed once it is sold. According to this argu ment, just as most states try to limit taxes im posed on parts used in the manufacturing pro cess to avoid double taxation, services used by businesses should receive similar treatment. Finally, perhaps the strongest objection is based on tax competitiveness. States are very sensi tive about taxing those services that are per ceived as footloose. Such taxation might also put home state service companies at a competi tive disadvantage when competing against outof-state businesses.1 4 One final area that is receiving consider able attention (but limited action) is property tax reform. Several states (for example, Illi nois, Michigan and Kansas) have proposed reducing property tax burdens using either tax caps or roll backs in property assessments.1 5 While these proposals are popular with vot ers—Michigan voters will again try to put an initiative on the 1992 ballot to roll back assess ments by 20 percent—cash strapped state gov ernments have trouble identifying sources of funds to replace these lost local revenues. The consequences fo r ta x re fo rm m easures When examining the tax structure of a given state or local government it is important to realize that the effect of raising $100 million in revenues through an increase in the sales tax is not the same as raising 100 million through the income tax. For District states, concern about revenue sources is increased by the tight fiscal situation facing its governments and the resulting adjustments to state fiscal systems. In recent tax developments, many of the District states have attempted to reduce depen dence on property taxes. Wisconsin has passed property tax relief measures and Illinois has increased its personal income tax in order to pay for a larger share of education spending, thereby reducing the need for the local property tax to fund these expenditures. Furthermore, Illinois has considered capping property tax assessment growth at 5 percent per year. Mich igan’s new governor proposed raising state sales and/or income taxes in order to reduce property tax dependence. If the trend to reduce property taxes continues, the result will be a larger role for state government in the District. In the short run, if District policy makers are forced to raise revenues, they need to con sider the tradeoffs suggested by the varying theories of taxation presented here. If a state’s interest lies in sparking capital investment, then encouraging increases in local property taxes is probably a bad idea as such a measure could, if unaccompanied by services benefitting local business, increase the cost of capital. If, on the other hand, legislators want to increase tax stability, then broadening the tax base of the sales tax by reducing exemptions or taxing services could provide a valid avenue. In any case, the ability to export tax burden outside the state’s boundaries is another consideration. Given the tenuous position of state and local economies, revenue raising decisions must be made with care. Adjusting the tax rates of major revenue sources in an ad hoc manner fails to recognize the interaction between taxes and economic activity. District policy makers need to understand both how the current tax structure effects the economy and how pro posed changes may improve or hinder future economic activity. FOOTNOTES 'See Musgrave and Musgrave (1976), pp. 210-211. 4See Fisher (1969), Chapter 4. 2See Kleine and Shannon (1986), pp. 33-36. 5Wisconsin Department of Revenue (1990). T or more on the incidence and theory of the property tax see Phares (1980), P. Mieszkowski (1972), and Musgrave and Musgrave (1976), Chapter 19. T he Council of State Governments (1991), Table 6.17. Note also that the Illinois state government only collects sales tax revenues based on a fixed tax rate. Revenues FEDERAL RESERVE BANK OF CHICAGO 31 above this percentage are distributed to various other governments in Illinois, such as special districts, municipal ities, etc. Also, Iowa’s tax rate is 4 percent; in Wisconsin and Indiana the tax rate is 5 percent. nSnell (1991). 1 2State Legislatures (1991). l3State Budget and Tax News (1991). 7Musgrave and Musgrave (1976), chapter 20. l4John Mikesell (1986). 8Council of State Governments (1991), Tables 6.19 & 6.20. 1 5State Budget and Tax News (1991). 9Phares (1986), chapter 5. l(,Phares (1986), chapter 5. R E F E R E N C E S The Council of State Governments, The Book of the States 1990-91, Lexington, Kentucky, 1990. Fisher, Glenn. Taxes and politics, a study of Illinois public finance, University of Illinois, 1969. Kleine, Robert and John Shannon, “Character istics of a balanced and moderate state-local revenue system,” Reforming State Tax Systems, National Conference of State Legislatures, 1986. Mieszkowski, P. “The property tax: An excise tax or a property tax?” Journal of Public Eco nomics, 1972. Mikesell, John, “General sales tax,” Reforming State Tax Systems, National Conference of State Legislatures, 1986. Musgrave, Richard and Peggy, Public finance in theory and practice, McGraw Hill, 1976. Phares, Donald, Who pays state and local taxes, Oelgescher, Gunn and Hain, 1980. _________________________ , “The role of tax burden studies in state tax policy,” Reform ing State Tax Systems, National Conference of State Legislatures, 1986. Snell, Ronald. “In the deep weeds” State Legislatures, February, 1991. State Budget and Tax News, Volume 10, Issue 10, May 6, 1991. State Legislatures, “Tax reform gets sophisti cated,” May 1991. State Policy Research, Inc., State Budget and Tax News, Volume 10, Issue 9, April, 1991. _________________________ , State Budget and Tax News, Volume 10, Issue 10, May 1991. National Conference of State Legislatures, “Tax reform gets sophisticated,” State Legisla tures, May 1991. 32 ECONOMIC PERSPECTIVES ECONOMIC PERSPECTIVES BULK RATE Public Information Center Federal Reserve Bank of Chicago P.O. Box 834 Chicago, Illinois 60690-0834 U.S. POSTAGE PAID CHICAGO, ILLINOIS PERMIT NO. 1942 Do N o t Forw ard A d d ress C o rrec tio n R equested R eturn Postage G uaranteed FEDERAL RESERVE BANK OF CHICAGO