No. 4 — February 2, 1996

Feature Article


Douglas Ostrom

Issue Index aaaa 1996 Archive Indexaaaa Search aaaa Subscriber Area aaaa Home



American analysts never stop looking for the secrets of Japan's remarkable postwar success, even during periods when, as at present, economic growth has slowed remarkably. Some analysts have identified public-sector investment as one of the keys of that success. The downside of such spending is that it often is associated in this country and elsewhere with pork-barrel politics.

The positive aspect of public-sector spending, providing it has certain characteristics, is that it is potentially productivity-enhancing — a point that may have gotten lost in recent American discussions about downsizing government. Whether, in fact, such expenditures improve overall productivity depends on the details and the objectives for which the funds are spent and the purposes from which they are diverted.

While the most direct measures of public-sector investments suggest that Japan does spend substantially more than this country and other major industrial nations relative to gross domestic product, other data indicate that Japan may get relatively little for its money. A rigid political structure is only one of several reasons for this outcome. Japan actually has less infrastructure of various kinds than does the United States despite higher outlays in this field. The data taken together suggest that public-sector spending almost certainly has not been responsible in any major way for Japanese increases in productivity vis-a-vis this country.



Experts in the United States have identified in more than two decades of ever-closer examination of Japan numerous economic and social characteristics that differ between the world's two largest economies. In many cases American analysts have indicated that the differences are a factor in the Asian nation's superior postwar economic growth rate. In not a few instances Japanese analysts have looked at nearly the same attributes and drawn the opposite conclusion: that the characteristic in question involves an area in which Japan to its own detriment has lagged behind the United States.

Infrastructure is one such example. Some American economists have looked at infrastructure spending in the two countries as well as that in other industrialized nations and found that those countries with higher levels of infrastructure spending relative to gross domestic product, particularly Japan, have experienced higher rates of overall growth than those competitors with lower rates, the most prominent example of which is the United States. Further, declines in public-sector investment in any given country have tended to coincide with lower economic growth rates.1 This argument, which largely originated with David Aschauer, now of Bates College in Maine, provoked a vigorous debate among American economists when it was made in 1989. Critics of the approach question the causal linkage in which infrastructure spending seems to lead economic growth and the implausibly high rates of return on public-sector investment that were found by Mr. Aschauer, among others.2 Data, moreover, provide weak support, at most, for proponents' conclusions (see Table 1).

For some Japanese analysts not only is the linkage between infrastructure spending and economic growth not obvious but they imply investment levels may be lower in Japan than those in other industrialized countries. They routinely cite their country's relatively meager urban park land and congested highways as among evidence that the nation's infrastructure clearly is underdeveloped compared to its counterparts in North America and Europe. Inasmuch as both American and Japanese researchers are looking at the same phenomenon both cannot be correct.

Both the American and the Japanese arguments have been marshaled in support of increases in infrastructure spending. If public investment is as useful and productive as suggested, it should increase in this country, taxpayer revolutions and suspicious attitudes toward government notwithstanding..

Similarly, if Japanese writers represent a consensus of public opinion, then policymakers in Tokyo presumably would not resist American arguments that Japan's public-sector investment spending should rise.

Official Washington's interest in Tokyo's level of infrastructure spending is more than an apparently blatant, if typically unnoticed, case of chutzpah. How can the United States legitimately declare that Japanese spending is too low when most recent data suggest that government at all levels in this country invests less than one-third the proportion of Japan's central and local governments? The answer has to do with the state of the Japanese economy and the relationship among savings, public spending and Japan's current account surplus.

Public-sector investment spending, illustrated as recently as Japan's September economic stimulus package (see JEI Report No. 36B, September 29, 1995), traditionally has been given an important role in countercyclical economic policy. Given the stagnation of the Japanese economy since 1991, one effect of which has been to reduce imports below the levels they otherwise would reach, public-sector investment is an obvious subject of attention in both Tokyo and Washington.

Public-sector spending for investment directly affects Japan's external balance by influencing the overall balance between supply and demand in the economy. Higher levels of government spending, other factors equal, soak up some of the excess of domestic production over domestic demand — also known as the current account surplus. Such spending offsets Japan's high savings rates, which by themselves tend to widen the gap between production and consumption. Moreover, with higher government investment, Japan's global external surplus decreases, as, presumably, does the surplus with the United States. For this reason Washington's advocacy of higher public investment spending in Japan is not as hypocritical as it appears.


What Is Infrastructure?

The term "infrastructure" clearly refers to fundamental building blocks of an economic system. Of course, many economic activities could be described as fundamental. For that reason two concepts — collective consumption and natural monopoly — can serve to define infrastructure and to narrow its scope. A third — public ownership — is a different, albeit overlapping, concept that many analysts use as a convenient proxy for infrastructure. Yet another concept — public need — is unrelated to infrastructure, appearances notwithstanding.

Collective Consumption - This refers to the fact that, for a small number of goods and services, one individual's consumption does not come at the expense of another's. Television broadcasts, for example, can be received by all households with television sets within the range of the signal. Unlike dry cleaning services, for example, an additional household can be added to a television viewing audience without having to either increase the supply of the service or inconvenience an existing customer.

Collective consumption is a relative concept. If a plane operating on a scheduled route is not full, another customer deprives no one of a seat. In this sense the service has collective attributes. Yet, once every seat is taken, a single additional customer can be seated only if someone is bumped or another plane is added. Hence, airlines combine the characteristics of television broadcasting and dry cleaning, generally falling outside the realm of infrastructure.

In some instances another customer actually benefits the existing customers and the provider by making the good or the service more attractive than it would be in his or her absence. The Internet is an appealing service precisely because of the large number of other users.

Collectively consumed goods present a problem of pricing, however. Because the product can be supplied to customers so easily, the supplier often has difficulty collecting for the good or the service provided. Yet, the service may have great appeal based on what customers would be willing to pay if forced. One way of covering the cost is through tax revenues, in which case the government typically provides the good or the service.

In Japan and many other countries television broadcasting is in part a government enterprise because the service is viewed as having some of the characteristics described above. More viewers do not deprive existing viewers of service, an informed electorate supposedly benefits everyone and getting individuals to pay for this presumed benefit through the market is difficult. The result in Japan is Nihon Hoso Kyokai (Japan Broadcasting Corp.), a nationwide network that is publicly owned and financed through semivoluntary "contributions" collected door-to-door and by other means rather than commercials or on-air solicitations. (Many Japanese admit that for years they have "just happened" not to be home at the time of the NHK solicitation visit.) Not surprising to American students of public enterprise is that NHK has leveraged its premier status in broadcasting to an important position in printed materials and cable television, two areas where the collective consumption argument appears to fall flat. NHK's private-sector rivals in broadcasting and other areas indeed complain of unfair competition.

NHK models its structure on British Broadcasting Corp. and other European public broadcasters; America has nothing that comes close. Public broadcasting in this country receives money from federal, state and local governments, but genuinely voluntary contributions cover a large share of the costs of providing its relatively modest output. Both public television and radio are far too busy trying to maintain their small roles in the United States to even think about becoming a miniconglomerate on the scale of NHK.

Natural Monopoly - This concept refers to the number of providers required to supply a good or a service at minimal cost. If servicing one customer is relatively expensive but adding extra customers imposes little additional burden, then a single firm or other organization will be the low-cost means of providing the good or the service. Market forces would tend to drive such industries to be dominated completely by one firm, even absent government intervention. Hence, such monopolies are "natural."

Traditionally, electrical power generation and distribution, which involve huge power plants and an extensive electrical grid, have been considered a single natural monopoly; indeed, most Americans have trouble even imagining two or more companies building power plants and stringing competing electrical wires into homes in a neighborhood. At the same time, though, isolated examples of competition among electric power companies in this country and moves to separate generation and distribution have begun to illustrate that natural monopolies are not immutable.

Infrastructures dominated by natural monopolies also involve a pricing problem. If a company sets prices independently, the lack of competition is likely to permit the firm to charge fees that will enable it to earn an exceptional return. If so, customers are likely to balk at buying much, even though the burden of substantial purchases on the economy's resources is modest. Alternatively, if the firm is forced to price the good or the service at a price that does not cover its high overhead costs, then it eventually will go out of business.

As a practical matter, the prices charged by natural monopolies are set in one of three ways. They may be subject to rate regulation on the part of government. This very widespread practice is typical of providers of local telephone service and electricity in the United States and Japan. Of course, this procedure creates numerous perverse incentives. For example, in order to get permission to raise rates an electric utility could be tempted to inflate costs. In theory, behavior of this sort may make the regulatory cure worse than the disease of monopoly. In such a case a second option appears more sensible: leaving natural monopolies alone to set prices as they wish. This alternative seldom is followed in either the United States or Japan. A third option is government ownership. Ideally, the government can set a low price reflecting the modest cost to society of attracting additional customers for the good or the service provided and cover the difference, if necessary, with tax revenues. In some instances, as with electricity in the United States, the natural monopoly dilemma is addressed through regulation in some areas and local public ownership in other parts of the country. In Japan, by contrast, electric power is provided exclusively through regulated private monopolies.

Many goods and services have both collective good and natural monopoly characteristics. Additional zoo visitors can be accommodated almost without limit; in that sense a zoo has aspects of a collective good. The extra visitors require some more work by the park staff in terms of providing information and the like, but, even if no visitors show up, the animals have to be fed. Hence, a zoo has elements of a natural monopoly.

In general, collective goods and services on the one hand and natural monopoly on the other are overlapping but not identical concepts. Collective consumption does not necessarily imply natural monopoly. As indicated by the multiple yet profitable television networks in the United States and Japan, for example, a single dominant broadcaster in most markets has not emerged, even when, as in Japan, it is able to tap what amounts to tax revenues. Similarly, a natural monopoly can exist without collective consumption. Electric power has some attributes of a collective service. During nonpeak hours one customer may be able to increase usage without denying usage to another, but in principle the amount of electric power is finite and divisible. Hence, electric power is not a collectively consumed service, even though it is a natural monopoly.

Public Ownership - As noted above, either collective consumption or natural monopoly provides a rationale for public ownership, although each characteristic can be handled through other means as well. As a practical matter, infrastructure sometimes gets equated with public ownership. That linkage is problematic, not only because some infrastructure is privately owned but because certain government-owned firms operate in industries not characterized by either collective consumption or natural monopoly.

Tobacco provides a particularly striking example of state enterprise not linked to collective consumption or to natural monopoly. In the United States several companies compete vigorously in the cigarette business. In Japan, by contrast, tobacco manufacturing long was a state monopoly, as it remains in some other nations. (Imported tobacco products in Japan have faced fewer restrictions over time, and in 1985 the former Japan Tobacco and Salt Public Corp. became Japan Tobacco, Inc. See JEI Report No. 32A, August 17, 1990, for details.) Even the most addicted smokers would deny to the industry an infrastructure status; in addition, cigarette manufacturing has none of the cost characteristics of a natural monopoly.

Generally, the obvious temptation for people with political influence and a reason to promote the interests of a particular industry is to try to expand the scope of industries considered to be characterized by collective consumption, natural monopoly or otherwise having characteristics typical of infrastructure. If this political exercise is successful, the range of goods and services provided at low or subsidized prices expands and the risk of failure diminishes for firms operating in the particular industry.

As the tobacco industry example and the importance of political factors illustrate, infrastructure and public ownership overlap, but the two descriptions cover far from the same thing. Industries that can be characterized as being part of a nation's infrastructure are often, but not always, government-owned. Publicly owned industries may or may not be infrastructural in character, even when the near-universal pattern is state ownership. Industries properly identifiable as having a strong economic rationale for public ownership, in fact, may be publicly owned in some nations but not in others or may at some points of time but not at others.

The collective consumption and natural monopoly concepts establish important benchmarks for evaluating public ownership. They provide a rationale for public ownership but, as noted above, fall short of creating an imperative for that action. Conversely, absent either collective consumption or natural monopoly considerations, the case for public ownership is weak, given the well-known tendency for government enterprises to be inefficient. For this reason alone public ownership should not be expected to equate necessarily to higher productivity.

Unfortunately, the available data sort industries not by degree of natural monopoly or some other economic criteria but by public ownership. Researchers, therefore, are obligated to accept government-owned industries or firms as proxies for infrastructure despite the deficiencies of this methodology. As one consequence, measured infrastructure is broader in those countries in which government ownership extends across a broad range of activity.

Social and Economic Infrastructure - Researchers sometimes try to separate the production and the quality of life dimensions of infrastructure investment into economic and social infrastructure. In fact, the two concepts overlap to the point of typically obscuring any useful distinction. For example, a new highway permits faster commutes to work, leading to a combination of greater leisure or longer working hours. At the same time truck drivers can do their work faster, permitting more work to be done in one day, thereby raising the productivity of truckers and trucks. If the new highway results in less wear and tear on the trucks, productivity goes up even more, inasmuch as the vehicles need less maintenance and less frequent replacement.

By contrast, some activities that are put under the rubric of social infrastructure are not infrastructural in character at all. Hospitals are one example. Hospital beds and the working hours of medical professionals are scarce resources that cannot be provided to one individual without potentially depriving another. Nor is natural monopoly the prevailing economic structure. While neighborhoods sensibly may regard a local hospital as essential, that fact does not make the facility part of the infrastructure any more than an equally — or, arguably, even more — essential local supermarket. In other words, the necessity for a good or a service has little if anything to do with whether or not it is part of the infrastructure.

In short, the concept of infrastructure is a mess. Yet, at some level nearly every expert agrees that infrastructure is important. Infrastructure properly defined makes work and play better. Infrastructure investment done right provides a combination of two things: it enables production to be accomplished more easily and it raises the quality of life. Either way, standards of living rise — the ultimate objective of economic activity.

Improvements in infrastructure may or may not be captured in official economic data. For example, a new national park provides a flow of services to its visitors. In terms of a nation's economic output those benefits are worth as much as extra cars rolling off an assembly line but may be far harder to measure.


The Costs Of Infrastructure

More infrastructure spending is not always good. Except during times of recession, increased public-sector investment spending, whatever its purpose, diverts resources from other uses. A government that builds a highway has to raise the funds somehow; whether the source is taxes or borrowing, other expenditures are sacrificed for the public purpose. If the government floats bonds and ultimately raises taxes to build a subway system that few people ride, policymakers have damaged the economy by reducing the ability of taxpayers to consume what they want.

Some analysts believe, however, that government almost never invests enough in infrastructure. Their argument is as follows. Purely private decisions, such as whether to buy a loaf of bread, can be made by an individual based upon his or her comparison of how much one wants the bread given its cost. A consumer does not as easily express a desire for a new highway. In fact, an infrastructure project typically makes sense if the sum of what potential users would be willing to pay is at least as great as the cost of the project. In a market economy no one individual has an incentive to take on this complicated calculation, leaving it up to the political process to react. The government may respond sluggishly, if at all.

Precisely because the popular will is expressed imperfectly, other analysts argue that public-sector investment consists of pork-barrel projects lacking an economic rationale. According to this way of thinking, elected officials and the bureaucracy do not necessarily have the public's overall interests at heart. They may push through projects that favor politically critical geographic areas or contributors to their political campaigns.


Distribution Of Infrastructure Projects

Pro and con cost arguments together suggest public investment may end up being either too large or too small, depending on the responsiveness of the political process to economic needs. This responsiveness depends, in turn, on the honesty of elected officials, the way in which political power is distributed and other factors. If, as in Japan, power tilts toward rural areas and the bureaucracy, then infrastructure investment decisions will reflect a similar bias.

Public works spending in Japan has shown a remarkably stable distribution by ministry for three decades, as shown in Figure 1. In some years the Ministry of Agriculture, Forestry and Fisheries has received the same portion of the public works budget as the year before, down to a tenth of a percentage point. From FY 1990 to FY 1993 the Ministry of Transport was allocated exactly 6.24 percent of the public works budget each year.3

Of course, infrastructure needs have changed dramatically over the period covered by Figure 1. Agriculture, forestry and fisheries constituted 9.9 percent of national output in 1965 against 20.1 percent of all public works funding allocated that fiscal year to the industry's namesake ministry. By 1993 the industry's share of GDP4 had dropped to 2.1 percent, but its cut of public works spending had risen slightly to 21.5 percent. Even within ministries shares have been remarkably constant. Hiromitsu Ishi, a prominent professor at Hitotsubashi University, notes that afforestation and conservation measures varied as a proportion of the MAFF public works budget only from a low of 2.6 percent to a high of 2.7 percent between FY 1980 and FY 1993.5 If for no other reason than that MAFF logically would move down its list of priorities over time, one would expect efficient budget allocation to imply far more variation in the share going to certain types of projects.

Other data regarding the apparent bias in public investment spending comes from its geographic distribution. In FY 1992, the most recent year for which data are available, Shimane prefecture won more public works spending per capita than any other prefecture in Japan. This rural prefecture, which borders the Sea of Japan, is neither the richest nor the poorest area of the country or even generally at the top or the bottom of socioeconomic rankings. It is, though, the home of Noburu Takeshita, prime minister from 1987 to 1989 and considered one of the most devoted Japanese exemplars of the famous dictum by the late Speaker of the House Tip O'Neill: "All politics are local." An early political associate of Mr. Takeshita's was the late Kakuei Tanaka, prime minister from 1972 to 1974, whose Niigata prefecture was long at the top in terms of per capita public works spending. (Niigata in FY 1992 had slipped to eleventh place among Japan's 47 major political subdivisions.)

What does Shimane prefecture get for its ¥553,548 ($5,535 at ¥100=$1.00) in per capita public works spending that last-ranking Saitama prefecture does not get with its measly ¥268,754 ($2,688)? Lots of highways, for one thing. In fact, some local critics have complained that the Shimane area, with its abundant historic sites and some of the country's most unusual ecological features, is all but being destroyed by the heavy levels of spending on roads. In this regard the complaints echo those regarding Hokkaido, which ranks number five in per capita public works spending. (For a brief description of public works and resulting controversies in Hokkaido, see JEI Report No. 2A, January 20, 1995.)

These and related figures raise the distinct possibility that, at least in Japan, the political process does not work well in allocating public investment either across different projects or by region. Agriculture most likely either was getting too small a share of investment in FY 1965 or too large a cut in FY 1993, for example. If the allocation goes haywire, the overall level of public investment is not likely to give a good indication of the gain in productivity from infrastructure types of investment.


Measuring Infrastructure Spending

As noted above, some American analysts have operated under the assumption that public-sector investment is allocated efficiently and that it has a high, arguably implausibly elevated, return. If so, those nations that do more public investment are more productive generally, other factors equal. Adherents of this view might choose as an illustration Japan's famous and largely publicly financed bullet train, the shinkansen, which permits business travelers to get from place to place faster, raising the amount of work that can be done in a day. In their eyes, then, America's lack of an equivalent high-speed rail network, in addition to being a disappointment to a handful of rail aficionados, might be one of the reasons why this nation's productivity growth has been arguably suboptimal.

Those making this argument sometimes bolster their case with data regarding levels of public-sector investment in the United States, Japan and elsewhere. As noted in Table 1 presented earlier, Japan ranks at the top and the United States at the bottom when such spending is calculated as a fraction of gross domestic product. For reasons to be spelled out below, however, those numbers must be regarded with caution.

In mid-January this year the Department of Commerce released for the first time as part of its GDP figures data on public-sector investment since 1982. Those new numbers, along with data for Japan, are shown in Figure 2. Use of the new Commerce yardstick for the United States rather than the measure employed by the Organization for Economic Cooperation and Development narrows the margin of difference with Japan, particularly when military spending is counted in the U.S. total.

High Japanese prices suggest one possible distortion in the transpacific public investment comparison. While comparable figures for the public sector are not available, the discrepancy is probably greater than, say, for housing, which is less subject to political influence on pricing. At current exchange rates residential construction prices in Japan are estimated to be double those in the United States.

One intriguing but rough comparison is possible from a pair of recent infrastructure projects in the United States and Japan. Within the past 17 months large new airports have become operational in both nations. Kansai International Airport, which opened in September 1994, and Denver International Airport, which opened at the end of February 1995, have received similar and decidedly mixed reviews from critics. While both airports are praised for their beauty (one is said to resemble a soaring eagle, the other the Rocky Mountains) and their use of advanced technology, they are criticized for their high cost, distance from city centers and costly, unanticipated difficulties with the new technology. While overall assessments eventually may change, at the moment "beautiful boondoggle" would appear to be a consensus description of both.

The differences between the airports are revealing, however. While Denver is a much smaller metropolis than the combined Osaka, Kobe and Kyoto region served by KIA, the Denver airport has at least as much — and by some estimates, twice as much — passenger capacity as KIA.6 The Denver airport has five runways against one at KIA, although both are designed so that more runways can be added. In actual size DIA is several times larger. On balance, however, the Denver airport clearly has the greater capacity to handle aircraft and passengers.

The Japanese airport, however, was much more expensive to build — nearly ¥1.5 trillion ($14.5 billion) against $4.2 billion. Technical difficulties and cost overruns boosted KIA's costs 40 percent above the projections made prior to the commencement of construction in 1987. Even if the airport had been completed at budgeted cost, it would have been more than twice as expensive as DIA, which is roughly two times larger.

These arguments suggest that KIA may have been as much as four times as expensive per passenger to build as DIA, which itself has been criticized as too expensive. While a single example is hardly conclusive, the KIA-DIA comparison is consistent with the widespread view that the cost of infrastructure is far higher in Japan than in the United States.

Why infrastructure costs more to build in Japan is controversial. To a large extent it may have to do with topography and other conditions about which little can be done. (The construction of an artificial island for KIA increased costs dramatically, but Denver's weather complicated the construction of DIA.) Land costs could be a factor in the reported figures for the two airports, although which way they would tip the balance is not clear given DIA's much larger size. Washington has argued that Japanese public works projects, including KIA, have not been as open as they should be to the participation of foreign contractors, which might be able to do the work for less. Finally, Japanese construction companies are believed widely to conspire, in effect, over the prices charged to government entities, apparently with the help of politicians. Japan's Fair Trade Commission, that country's antimonopoly watchdog agency, has become increasingly active in taking action against such practices (see JEI Report No. 6B, February 11, 1994), but the jury is still out as to the effectiveness of these moves. Despite the JFTC's newfound vigor, most analysts probably would not dispute the conclusion that contract charges in the public sector — combined with prices for the rest of GDP that are less elevated compared to abroad — raise the proportion of public works spending as a fraction of GDP in Japan relative to levels in other industrialized countries.

This argument would suggest that Figure 2 overstates the actual difference between the United States and Japan in public investment at any given time, but one still can conclude that such spending is on the rise in Japan. Since 1991 Japanese public-sector investment has risen sharply as a percentage of GDP. In August 1992 Tokyo adopted the first of what has become a series of economic stimulus packages, the centerpieces of which typically are increased public-sector spending. Elevated levels of public investment are likely to continue at least through the balance of 1996, inasmuch as the most recent package was announced only last September (see JEI Report No. 36B, September 29, 1995).

Even in the absence of recession, Japanese public-sector spending was poised to rise in the early 1990s as a consequence of the 1989-90 Structural Impediments Initiative discussions between Japan and the United States. In the final accord reached in June 1990 Tokyo agreed to spend ¥430 trillion (roughly $3 billion at the average 1990 exchange rate of ¥145=$1.00 or $4.3 billion in today's dollars) on public works between FY 1991 and FY 2000 inclusive. That objective will be reached if, combined with the spending to date, public works outlays expand a nominal 3 percent a year from now until the end of the period. Given that the economy cannot be guaranteed to grow that fast over the remainder of the interval, achievement of the goal may require an increase in the percentage of GDP accounted for by public works. In early October 1994 the cabinet approved a revised basic plan for public investment that included a projection of between ¥600 and ¥630 trillion ($6 to $6.3 billion) in cumulative spending between FY 1995 and FY 2000. Under this scenario public investment would continue to rise approximately 3 percent nominally each year.

Of course, the spending that Tokyo committed to in the SII agreement and in the subsequent cabinet action, assuming that it continues to materialize, does not assure that the economy will receive a proportionate boost. As already noted, merely spending money does not automatically mean that productivity goes up. Productivity and GDP growth actually may stagnate if a worthy capital spending plan gets less support than a pork-barrel favorite.


Alternative Measures Of Public-Sector Capital

Public works spending each year tells only part of the story, of course. Public investment by definition results in buildings, equipment and other capital goods. Another way to compare the United States and Japan is by the level of capital stock — rather than the annual flow of investment — relative to GDP. In 1992, the most recent year for which comparable data are available, the public-sector capital stock in Japan equaled 62 percent of GDP; the relevant figure for the United States was 45.8 percent. In short, not only does Japan invest a higher percentage of GDP in public investment each year, the cumulative effect is a public capital stock that also is larger relative to GDP.

Few Japan watchers would accept this conclusion without qualification, however. Some of the more obvious apparent examples of public capital stock turn out not to be relevant. While Japan does have a splendid rail system, for example, it now is privately owned and, therefore, not part of the public capital stock. Nor, for reasons explained below, should it be considered part of infrastructure. Also, as noted above, anticompetitive practices arguably lift prices in the public sector in Japan, obscuring simple comparisons with the United States. Indeed, for Japanese observers using very similar definitions of infrastructure, just the opposite argument from the one articulated by some American analysts is possible. In the Japanese view their nation suffers from a paucity of public capital. The argument is largely anecdotal but nonetheless convincing. While data and economic differences render U.S.-Japan comparisons difficult, in a number of important sectors U.S. infrastructure appears greater than that in Japan.

Parkland - One of the most widely cited statistics in Japanese analyses of infrastructure involves urban parkland per capita. Comparisons collected from several Japanese sources are shown in Table 2. With only a handful of exceptions, the cited Japanese cities have far less parkland than large cities in other industrialized countries.

Other, less-often cited, figures show a similar story for parkland for the nation as a whole. For example, Japan has only 4,594 square feet of national parkland per capita. By contrast, the United States has in excess of 13,000 square feet per person. Of course, much of the U.S. total is in the West far from population centers, making it less than easily accessible for the casual traveler. The same is true for Japan, however; some national parks are accessible only by ship or air, with flights from regional population centers taking an hour or more. Nor do urban parks and national parks tell the whole story; America has more than 2,000 square feet of state parks per person, often within an easy one-day round trip from urban areas. Japan has nothing equivalent.

Japan as a densely populated island might not be expected to have large amounts of parkland, but it does not do particularly well even with areas of the United States that are just as urbanized. New Jersey, an area with roughly the same population density as Japan, does not have a strong reputation as an outdoors vacation spot, even with its own residents, despite its Garden State nickname. Yet New Jersey has two national recreation areas partly within its borders and roughly as much state parkland (1,757 square feet per capita) as the rest of the United States. Given abundant national parks in nearby states, the outdoor recreational opportunities available for residents of New Jersey are much greater than almost anywhere in Japan.

Highways - One might argue that parkland as social infrastructure does not count in the public capital stock tally, despite the argument made earlier that social and economic infrastructure distinctions are not meaningful. Yet, highways, which contribute directly to increased output, tell a similar story, albeit with somewhat more nuances. The United States has roughly three times the expressway mileage per capita of Japan and nearly the same advantage with major secondary roads. Japan actually has more road mileage in total per capita than the United States, but many of these are either winding affairs that motor vehicles share with bicycles and pedestrians or unpaved roads.

Unlike parkland, highway mileage comparisons change somewhat once corrected for population density. If Japan were to be compared once again to New Jersey, that country has 40 percent more miles of expressways per capita than New Jersey, although it has fewer major secondary roads. For roads of all types Japan has more miles per capita.

These comparisons would appear to suggest that Japan's highway infrastructure is fairly impressive, given the urban character of the country. In addition, people there have a close substitute for highway travel in their rail network. Together these arguments seemingly indicate that the highway network of Japan is as adequate as that of the United States, if not more so.

Few people who have spent much time on roads in both countries would buy this argument, however. One problem is the distribution of highways in Japan — more per capita in rural areas and less in cities and surrounding areas. Traffic reporters routinely report huge backups in urban areas, sometimes even late at night. By contrast, some high-speed motorways in Kyushu, the southernmost of the main islands, are so uncrowded that one may meet only a couple of cars an hour during midday, even though the region is not sparsely populated by American standards. While this apparent rural bias regarding highway construction is far from unknown in the United States, its extent appears to be much greater in Japan. Such a bias hardly is surprising, however, in light of the alignment of political forces in Japan and the tendency to favor rural interests in a wide range of policy areas.

Comparisons of expressways that include European countries tell a story more consistent with what one would expect from high levels of traffic congestion. As shown in Table 3, in comparison with Germany and Italy, for example, Japan has fewer miles of expressway per capita, per square mile and per 10,000 cars. While the United States, the United Kingdom and France rank below Japan in some dimensions of expressway capacity, by two criteria — per capita expressway length and expressway extent per car — Japan ranks last. In short, although some information suggests that the system of roads is adequate in Japan compared to other large industrial countries, the preponderance of statistical data as well as the experiences of everyday drivers suggest that Japan, in fact, does suffer from having fewer good roads, particularly in urban areas, than would be considered sufficient in comparable countries.

Railroads and Airlines - Common measurements of rail and air transportation focus on the amount of freight or number of passengers carried. Such data reflect demand for the services of the organizations, such as airlines, that use the infrastructure. Providers of such services typically do not meet the criteria described earlier for being counted as infrastructure, like providing a collectively consumed good or operating under conditions of natural monopoly. In Japan and the United States these carriers typically are not now government-owned, although important rail and airline companies were public enterprises in Japan until the mid-1980s.

The most revealing available data regarding public transportation infrastructure concerns airports, which do have many characteristics of both collective goods and natural monopolies. Typically in both the United States and Japan airports open to the public are operated by an arm of government. With a total of 17,846 public and private airports, this country has more airports than the rest of the world combined. Scheduled air service is provided at 800 of the 5,545 airports available to the public. By contrast, fewer than 100 airports in Japan have scheduled flights. These figures suggest that America has roughly four times the number of airports with scheduled air service per capita. The average Japanese, however, probably is fewer miles from an airport than his or her American counterpart.

Sewage Diffusion Rates - Critics, particularly in Japan, have long cited less than universal sewer systems as a particularly glaring deficiency of social capital in Japan. A quarter-century or more after the subject first received widespread attention — and despite an official commitment to close the gap with other industrial countries — differences remain large. The most recent data suggest that 47 percent of homes in Japan were hooked up to sewage systems in 1992 against 68 percent in France in 1987, 96 percent in Great Britain in 1990, 86 percent in Germany in 1990 and 73 percent in the United States in 1986. Although alternate waste disposal systems, such as septic tanks, may serve the Japanese people well as evidenced by few pollution problems from the lack of sewage system hookups, this measure remains one often used to judge the adequacy of public infrastructure, especially in highly populated areas.



Japanese infrastructure, in short, is not as great as it looks from the GDP data. In fact, by many measures it continues to lag that in other large industrial nations. While examples of specific types of infrastructure cannot tell the whole story, the tentative conclusion is just the opposite of the one sometimes reached. To the extent that a good transportation system, parkland and the like increase productivity or the level of GDP properly measured, low degrees of infrastructure development probably have reduced Japan's GDP below what it otherwise would reach, other factors equal.

Other factors seldom are equal, though. At least until recently high returns from private-sector outlays logically may have implied that private, rather than public, spending has boosted productivity more usefully. The apparent high cost of public investment and politically inspired misallocation further have reduced the returns on such investment relative to that made by businesses.

In the mid-1990s, after three years of near-zero economic growth, the thesis of high returns in the private sector may have to be reexamined. In addition, the process of political reform bears watching. If, as some analysts believe, the recent realignment of Japan's legislative districts tilts the political balance toward urban areas — which seem to have received less than a fair share of infrastructure spending to date — then future public investment may be used far more efficiently. Finally, stepped-up antitrust enforcement along with new rules may make public-sector investment less costly, permitting more infrastructure investment without a bigger budget.

In short, the climate might be ripe for a serious effort by Japan to close the infrastructure gap with other industrial countries. To the extent that these efforts are financed through deficit spending one side effect could be to reduce Japan's global external imbalance. The conditions under which such spending would improve in any significant way the productivity and the standard of living of the Japanese people remain problematic and tangled in a web of competitive conditions and politics. Absent appropriate regulatory reform and improvements in the political structure, Washington's efforts to get Tokyo to expand infrastructure spending could be tantamount to urging the government to waste money. For that reason such an initiative would not necessarily enjoy broad public support in Japan. Within the next few years, as competitive practices change and elections are held in the realigned legislative districts, a clearer picture should emerge as to whether additional public-sector investment spending could be put to wise use in Japan. Both Japanese consumers and economic policymakers in Washington would have cause to applaud such developments.

Ai Ando provided research assistance.

The views expressed in this report are those of the author
and do not necessarily represent those of the Japan Economic Institute

Top aaaa Issue Index aaaa 1996 Archive Indexaaaa Search aaaa Subscriber Area aaaa Home


1aa For details of this argument, see David A. Aschauer, "Is Public Investment Productive?" Journal of Monetary Economics, March 1989, pp. 177-200. Return to Text

2aa One of those generally supportive of Mr. Aschauer, Alicia H. Munnell, formerly with the Federal Reserve Bank of Boston, has received a nomination from President Clinton to the three-member Council of Economic Advisers. For various views of the subject, see Alicia H. Munnell (ed.), Is There A Shortfall in Public Capital Investment? (Boston, Massachusetts: Federal Reserve Bank of Boston, 1990), and Edward M. Gramlich, "Infrastructure Investment: A Review Essay," Journal of Economic Literature," XXXII, September 1994, pp. 1176-1196. Return to Text

3aa See Hiromitsu Ishi, "Rigidity and Inefficiency in Public Works Appropriations: Controversy in Reforming the Budgeting Process in 1994," Journal of Japanese Studies, Summer 1995, p. 407. Return to Text

4aa Based on 1965 and 1993 calendar-year GDP. Government investment by ministry is available only on a fiscal-year basis; sectoral GDP is available only on a calendar-year basis. Return to Text

5aa Ishi, op.cit., p. 411. Return to Text

6aa However, unlike Kansai International Airport, an older airport was closed in Denver when the new facility opened. Nevertheless, the discrepancy in market size is relevant to the discussion below regarding numbers of airports per capita. Return to Text

The views expressed in this report are those of the author
and do not necessarily represent those of the Japan Economic Institute

Top aaaa Issue Index aaaa 1996 Archive Indexaaaa Search aaaa Subscriber Area aaaa Home