Everybody likes infrastructure investment, but are its returns disappointing and its costs rising?
New research in America suggests that the costs of infrastructure have been increased by the increasingly complex planning and consultation processes involved.
Economics and economic policy are often highly contested arenas. The benefits of infrastructure investment for economic growth and economic welfare of communities is something that for many years has been a broad area of agreement. Both the Trump and Biden administrations have been committed to raising US infrastructure investment. EU structural funds have emphasised infrastructure in terms of EU growth and widening European prosperity. Much of the interest in new economic theories, the so-called endogenous growth theory, turned on high returns to infrastructure investment. The post-Brexit debate in the UK has focused on a ‘levelling up’ agenda significantly involving the use of infrastructure investment to improve the economic performance of less favoured regions that continue to suffer from the consequences of deindustrialisation in terms of GDP per capita.
The costs of public infrastructure are rising.
The latest edition of the American Economic Journal: Applied EconomicsVol. 15, Issue 2has caught my attention because it contains an interesting article on the costs of infrastructure investment and how it appears to be rising in America. The article shows, see chart below, that the cost of highway construction in the United States has increased significantly over the last 60 years.
I have taken an interest in the debate over the benefits of infrastructure spending over many years. This started with the anguished debate in the US in the late 1980s over whether America was under investing in public assets, the novel and arresting claims made for aspects of investment spending during the high point of interest in new endogenous growth theory in the early 1990s and the settled conviction among economic policy makers in the 21st century that public and private infrastructure investment is the a key to economic growth and to regional economic regeneration in particular, in the face of uncomfortable disappointing evidence accumulating around the world over many decades.
Perhaps American Economic Progress in the 19th century was not down to the railways.
Yet this entrenched and happy consensus surrounding the benefits of infrastructure investment merits critical and rigorous interrogation. The great example of a technology and a huge investment programme that transformed an economy is the completion of the trans-continental railroads in 19th century America. For the first seven decades of the 20th century railways were considered to be key to American prosperity and growth in the second half of the 19th century. This judgement was based on a strongly asserted set of a priori observations that were collaborated by much factual evidence of the growth the American economy occurring in tandem with railway investment. This accepted analysis was not informed by a rigorous interrogation of the available data.
In 1964 Robert Fogel upended this received judgement when he published Railroads and American Economic Growth: Essays in Econometric History in 1964. Robert Fogel, whose PhD had been supervised by Simon Kuznets the American pioneer of national income accounting in the 1930s, applied quantitative methods to historical economic data. Robert Fogel looked at the American economy in 1890 and compared it to a counterfactual economy in 1890 where transport was limited to wagons using turnpike roads, canals and natural waterways. Transport costs from farms to their primary markets would have been increased, particularly in the Midwest and the geological location of agriculture would have changed, but despite the increase in transport costs the social savings that arose from the railways was only the equivalent of 2.7 percent of GDP. Moreover, substitution effects available from the development of a more extensive canal system and improved roads would have further reduced the costs of not having railways. Robert Fogel’s conclusion was that railway development was much less important than had been thought and they were not indispensable for the development of the trans-continental American economy.
New Endogenous Growth Theory and the Clinton Administration and the American debate on infrastructure in the 1990s
In the 1990s an economist at the Federal Reserve Bank of St Louis John Tatom challenged the assertion that there was a shortage of publicly owned capital stock that was lowering private sector productivity and overall economic growth. The Clinton administration was heavily influenced by the new economic research into growth theory that emerged at that time. The standard neoclassical growth model could only explain growth by changes in the capital stock, population growth and technical progress. Each of these apart from changes in the capital stock were determined by factors that were exogenous to the model and therefore not really explained by the model at all. Among the interesting propositions of the family of research work that made up the new growth theory policy agenda was that some investments would generate high beneficial spill overs that yield not only constant returns to scale by increasing returns to scale rather than the normally observed diminishing returns at the margin. These unusually high returns in this research were associated with certain forms of manufacturing investment and public sector infrastructure investment. This work was done by Brad De Long, a Harvard economist on manufacturing, and Alicia Munnell and economist at the Federal Reserve Bank of Boston. They both served in the US Treasury Department between 1992 and 1994, when Senator Lloyd Bentsen was Treasury Secretary. By the Autumn of 1994, before the Democrats lost control of both houses of Congress, the Clinton administration had given up on an ambitious programme of large scale public infrastructure investment. Administration officials at the Treasury Department were candid: Alicia Munnell’s estimated returns to public sector infrastructure investment were recognised as being unrealistically optimistic and the administration was greatly influenced by another and less widely appreciated dimension of the new growth theory research work that suggested that private sector investment was sensitive to long-term interest rates and crowding out. Given the key role of compounding in economic growth it was imperative to do nothing to impede growth in the private stock of capital. This explains the Clinton administration’s determination to lower Federal borrowing and the creation of a surplus in 2000 under Treasury Secretaries Robert Rubin and Larry Summers in order to control the cost of long-term debt.
John Tatom’s critique of the public sector infrastructure agenda was published in a Federal Reserve Bank of St Louis research note Is an infrastructure crisis lowering the nation's productivity? in 1993. Its analysis turned on the reactive role of the Federal Government in infrastructure spending and the causality of the asserted connection between public investment and private sector productivity. He argued that Federal spending only accounted for a small proportion of total public sector capital accumulation. A central part John Tatom’s critique was that ‘the purported link between public capital and private sector productivity has been widely criticised for distorting the role of public capital, yielding implausible estimates of the private sector productivity gains that could arise from public sector capital formation, and reversing the connection between the two. The fundamental problem with earlier estimates is that they result from spurious or unrelated movements in the quantity of public capital and business sector output and productivity’. Moreover, the movements and growth in public and private sector productivity at that time did not offer any guidance to the direction of causation between the two. The two appeared in the 1980s to move inversely to one and another. For John Tatom a matter of ‘special note is the rebound in private sector productivity growth until 1988, which was accompanied by an accelerated decline in the stock of public capital per hour. The bottom line here is that no one has produced evidence that an increase in the nation’s public capital stock will boost private sector output or productivity, within the year or even some future period’.
The merits of public sector investment in infrastructure has become an almost theological belief.
Economists that publish research that reports reassuring high returns to public sector investment are widely reported and thanked by national politicians, local authority leaders and the wider public sector and its agencies. Economists that rigorously analyse returns to public sector investment and regeneration programmes are rarely thanked and can find themselves shunned if they are closely connected to the local community where they have studied results from public investment. Public bodies rarely offer cogent analysis of the expected returns to be yielded from public sector investment. The UK Treasury’s Green Book that summarises the cost benefit criteria to be used in assessing proposed investment is subjected to an energetic political lobby of local authorities, local government representative bodies, activist think tanks and construction companies and other commercial interests that expect contracts to lower the discounts and expected rates of return on investment. One of the main centre pieces of the 2022 UK Budget was a commitment to infrastructure spending as part for the Government’s levelling up agenda. Yet HM Treasury officials explained that the Treasury did not have estimates for the rate of return expected on the investment but were assured by the evidence of the wider economic literature that the investment would yield benefits.
There are few public sector or political thank yous for questioning or rigorously examining returns to infrastructure
A Danish economic geographer Bent Flyvbjerg who now works at Oxford University has undertaken extensive research into the expected returns to public infrastructure investment and the claims made for it. He learnt early on in his career that questioning the benefits of public investment had the potential to earn you few friends and to invite the heavy handed opprobrium of government officials.
When he received a large research grant from the Danish Transport Council to develop his first research group on megaprojects, at Aalborg University, Denmark, he was pleased to see that it attracted the attention of a senior Danish official. That was until the civil servant took him to a restaurant on the canal close to his ministry, where:
‘we had lunch and pleasant conversation, until towards the end of the meal, when he congratulated me on my new research grant and told me in no uncertain terms that if I came up with results that reflected badly on his government and ministry he would personally make sure my research funds dried up’.
This lunch and pre-emptive rebuke in Copenhagen from a high-ranking government official in infrastructure planning in Denmark, whom he had met on several occasions, provoked Bent Flyvbjerg into curiosity about the benefits of large scale public investment where there is of as much political capital at stake as public money. This has stimulated him to examine rigorously public investment projects and spending on investment, their costs, their returns and the wider economic and social benefits that were claimed for them.
Cost overruns, expected benefits do not materialise and wider economic and social benefits appear hard to identify
Bent Flyvbjerg and various academic collaborators that include Cass Sunstein have looked at thousands of large-scale projects over 80 years. As he puts it:
‘the world’s largest database of big projects now exists and its bottom-line numbers are grim. Nearly 48 percent of big projects finish on budget, eight-and-a-half percent finish on budget and on time, and only half of a percent of big projects finish on budget and on time, delivering the expected benefit’s’.
Moreover, he and his colleagues report that most types of the investment big projects involved in this analysis have what statisticians call “fat tails.” That means disasters are far more likely than conventionally believed. Disasters like Boston’s “Big Dig” transportation project, which went 228 percent over budget, the Montreal Olympic Games, which went 720 percent over budget, or Scotland’s Parliament which, when it opened in 2004, was an almost unimaginable 978 percent over budget’.
How the costs of infrastructure investment have increased over the last fifty years
In Infrastructure Costs by Leah Brooks and Zachary Liscow which appears in this edition of the American Economic Review of Applied Economics the authors explore how real spending per new mile over the history of the Interstate Highway System has increased. They found that spending per mile increased more than threefold from the 1960s to the 1980s. This increase persists even conditional on pre-existing observable geographic cost determinants’ They identify some evidence that may explain this increase in costs . Apparently ‘input prices explain little of the increase’. While statistically, changes in income and housing prices explain about half of the increase where a wealthy community may choose to spend more on highway services.
Apparently, the principal cause of increased costs has been that of the process of getting permission to undertake and carry out the projects. This relates to the cost of consultation, protracted and expensive litigation and the costs that arise from several pieces of US Federal legislation that impede the process of reaching agreement on new highway programmes as well as various state level pieces of legislation.
This federal and state legislation will not only impede and increase the costs of highway investment but the construction of infrastructure more widely such as the building of dams for water supply and the construction projects involved in the ambitious programme of green infrastructure laid out in the recent Inflation Reduction Act and the other infrastructure plans that have formed part of the Trump and Biden administration's economic agendas. Ironically many of the citizens who are as engaged and active in promoting the agenda of decarbonisation often find themselves being as energetically involved in opposition to the projects that would contribute to accomplishing it.
The citizen voice hypothesis suggests the costs of infrastructure expending have been increased
Leah Brooks and Zachary Liscow ‘propose the citizen voice hypothesis, arguing that key social changes in the late 1960s and early 1970s empowered people to demand changes in government behaviour that yielded increased costs’. Their research identifies ‘suggestive evidence consistent with this hypothesis’. They are scrupulous in offering a balanced interpretation of the potential consequences of this enhanced citizens’ voice hypothesis. Leah Brooks and Zachary Liscow write that ‘one benign interpretation of these findings is that more expensive highways are an efficient response to the overall rise in US incomes. Another benign interpretation is that the rise of citizen voice causes governments to internalise what were previously external costs of highway construction, raising the cost of highways but increasing social welfare’. However the authors also note that ‘a malign interpretation is that the tools of citizen voice accentuate the voices of favoured parties, yielding increased spending with little social value’. What the articles does show is that the costs of infrastructure spending as a result of the process has increased significantly in over 50 years and an increase in costs that will reduce economic and social returns from it. These sorts of process and public policy transactional cost increases are not confined to the USA but can be identified across advanced OECD economies in the EU, Canada and the UK. There is little doubt that the construction of dams, roads and even simple local community regeneration schemes often generates intense and sometimes toxic controversy.
Warwick Lightfoot
Warwick Lightfoot is an economist and was Special Adviser to the Chancellor of the Exchequer between 1989 and 1992, he is the author of America’s Exceptional Economic Problem.