The Bear’s Lair: Sound fiscal management beats infrastructure

The Indian budget on February 28 slackened the much needed effort to move that country’s fisc towards balance in order to spend more on infrastructure. It was accordingly much praised by the Keynesian media worldwide. While the Lord knows infrastructure is much needed in impoverished India, and its cost there is presumably not as bloated as in the West, this choice still betrays a misguided sense of priorities. Budget deficits suck resources from the private sector to the public sector and leave the private sector impoverished and feeble. A drastic cutback of the public sector, especially in its regulatory aspects, would do much more for the Indian economy than even the most attractive “stimulus” infrastructure spending. A fortiori, that is true for rich economies such as Japan and the United States where infrastructure is already adequate.

Public choice theory quickly shows us why large infrastructure projects may be highly economically damaging. Politicians don’t have to pay for them, yet get the glory of opening them and in some cases getting major projects named after them. Generations untold, for example, have a better opinion of the economically disastrous President Herbert Hoover by contemplating the sublime Hoover Dam (which by losing re-election in 1932 Hoover himself was unable to open). Built at a time when infrastructure projects were still moderately efficiently constructed, it has provided massive amounts of electricity to the western United States for 80 years. It thus epitomizes the case for infrastructure spending, although it should strictly be named the Coolidge Dam, construction having been authorized by Hoover’s modest and effective predecessor in 1928.

The British High Speed Rail 2 project is a classic example of the bloated infrastructure genre. Its economic benefits are modest at best, cutting around 30 minutes off the journey time from London to Manchester, yet its cost is estimated at an astronomical 43 billion pounds ($65 billion), even at the doubtless over-optimistic figures of its sponsoring government department. It will doubtless project great glory on the next prime minister but three at its opening in 2027 (it’s not clear why David Cameron favors it, since he’s most unlikely to attend the opening, except as a second-tier geriatric guest in the background.)

To see why this is a bad idea, consider the original railway line from London to Manchester, which took the form of three separate projects, the London to Birmingham Railway, the Grand Junction Railway and the original 1830 Liverpool to Manchester Railway. Both the later projects were approved by Parliament in 1833, and construction was begun on both later that year. The London to Birmingham was opened in September, 1838, while the Grand Junction had opened a year earlier in July 1837. The three lines merged in 1846, and the combine, including numerous branch lines and the Crewe construction facilities, a total of 535 miles of track, was capitalized a few years later, according to an 1853 “Shareholders’ Key” at 31 million pounds, being 21 million of equity and 10 million of debt – a sum equivalent to about $5 billion today (and at that time the largest capitalization of any company in Britain.)

Note the difference in planning and construction efficiency from today. In spite of the far less efficient construction methods used, the London/Manchester railway complex was built in less than 5 years for a cost of less than $5 billion equivalent – the “Shareholders’ Key” gives a cost for the three original lines of 12.4 million pounds, equivalent to about $1.9 billion today. A functionally equivalent railway today is expected to take 12 years to build – beyond the 6 years’ planning already taken – and to cost about 30 times as much, in real terms.

This is not solely because today’s railway is built through built-up areas – the areas concerned were already heavily populated in 1830. It’s partly due to the lower intrinsic efficiency of public sector projects against private sector enterprises, but above all it’s the result of the insane costs of bureaucracy and regulation that are added to large infrastructure projects in today’s world.

You can do the same calculation in respect of the proposed tunnel under the Hudson River recently killed by New Jersey governor Chris Christie versus the functionally identical Holland Tunnel opened in 1927, and come to the inexorable conclusion: In today’s world, at least in wealthy “developed” countries, large public sector projects cost at least ten times what they should cost, or (in real terms) would have cost a century ago.

The economic implication is thus clear: if infrastructure costs ten times as much as it should, because of regulatory, environmental and bureaucratic cost inflation, then the world should stop building infrastructure altogether unless its need is so painfully obvious that it can be built and funded by the private sector.

Such a moratorium on new infrastructure projects is a perfectly economically viable strategy for a reformist government. If an Interstate (say) becomes so overcrowded that traffic jams abound, then a private sector toll road will be highly profitable, without the government intervening at all. That toll road will be subject to regulatory and environmental nonsense, but it will at least avoid the political process and absurdities such as the 1931 Davis-Bacon Act requiring “prevailing wages” to be paid on federally funded or assisted public works contracts, which in practice imposes union featherbedding on such contracts.

As the infrastructure deteriorated, public demand would build up for infrastructure spending, which a wise government would resist, while undertaking studies identifying all the cost bloating factors that have caused infrastructure costs to soar, and introducing legislation eliminating as many as possible of those costs. Eventually the cost of infrastructure would be reduced to no more than twice its inflation-adjusted 1900 level, at which point public works could recommence.

Overpriced infrastructure is especially pernicious because of its effect on budgets. Politicians can convince themselves, if nobody else, that infrastructure boondoggles are really “investments” and therefore should not count in budget calculations. The result is that public debt soars uncontrollably, while the overpriced infrastructure makes losses (however it is priced, and especially if its use is not priced at all) and hence does nothing whatever to offset the cost of the debt raised to fund it. Eventually, the result is bankruptcy, either of the polity concerned or even (in the case of Greece, whose infrastructure boondoggles were mostly funded by the EU) of the larger entity funding the waste.

Japan is the cause celebre for infrastructure waste. Ever since the boom ended in 1990, the Japanese government has tried to stimulate the economy by endless Keynesian programs of “stimulus” infrastructure spending. By 2000, Japanese infrastructure spending had reached 6.5% of GDP, twice the level of France, the next most profligate country. Junichiro Koizumi brought it down, and went some way to solving Japan’s fiscal problem, but his successors, including notably Shinzo Abe since 2012, have reverted to infrastructure spending stimulus. For example Japan’s transport infrastructure spending alone in 2008-10 averaged 30% higher than the OECD average as a percentage of GDP, 38% higher than the OECD average, 50% higher than the United States (even in those years of the Obama stimulus) 70% higher than Germany and 25% higher than even France. Since Japan is a geographically compact country that has in the last two decades hopelessly over-“invested” in infrastructure, its current overspending is especially egregious.

In India, with its abundance of low-cost labor, it is of course possible that public sector infrastructure suffers less from the sort of cost bloat that blights Western programs, but knowing what we do about Indian levels of corruption it has to be said that such a claim is barely credible. We do know that in 2012 the country spent 4.7% of GDP on infrastructure, less than China but more than almost all rich countries. We also know that the Indian public sector deficit, including the central government and the states, peaked at close to 10% of GDP in 2013, and is probably running at 6% of GDP in 2015-16 (the Federal deficit is forecast at 3.9% of GDP.)

India’s two greatest economic needs are to scythe back the “permit raj” that has made investment in that country such a problematic process and to reduce the public sector’s drain on the country’s resources, which will cause trouble in the next global credit crunch, due soon. Infrastructure spending, such as the 100,000 km of new roads proposed in last week’s budget, should not be expanded until these objectives have been met. It is simply not a priority.

The wealthy world’s budgets are all in some difficulty, disguised by nearly a decade of pathological interest rate conditions, with fatuous “stimulus” programs in many countries never having been properly reversed. Given the excessive cost of infrastructure in rich countries, and the need to control public spending, it would be sensible to allow a few bridges to fall down, while forcing the vested interests surrounding infrastructure provision to disgorge their ill-gotten gains and return its costs towards nineteenth century levels.

Public infrastructure provision has failed. It needs to be eliminated, and provision made for the private sector to step in, at much lower cost, where there is a need.


(The Bear’s Lair is a weekly column that is intended to appear each Monday, an appropriately gloomy day of the week. Its rationale is that the proportion of “sell” recommendations put out by Wall Street houses remains far below that of “buy” recommendations. Accordingly, investors have an excess of positive information and very little negative information. The column thus takes the ursine view of life and the market, in the hope that it may be usefully different from what investors see elsewhere.)