Too much of a good thing? Infrastructure in China

by Jing-Li Yim, Research Analyst, EDHECinfra

A review of the paper “Does Infrastructure investment lead to economic growth or economic fragility? Evidence from China” by Atif Ansar, Bent Flyvbjerg, Alexander Budzier, and Daniel Lunn, in Oxford Review of Economic Policy, Volume 32, Number 3, 2016, pp. 360-390

In a world where infrastructure investments can be a scarce commodity, markets where a lot infrastructure is procured all at once can hide many risks. A recent paper on the quality and impact of Chinese infrastructure highlights this issue, amongst others. 

This paper by Ansar, Flyvbjerg, Budzier and Lunn challenges the prevalent view that infrastructure investment is always a precursor to economic growth. Given China’s three decades of rapid economic growth, development and urbanisation since its economic reform, China’s growth model is of particular interest for testing the macro-level effect. With its level of infrastructure investment surpassing that of many parts of the world, it is also worthwhile determining whether China’s delivery of infrastructure projects is superior and should be emulated.

This paper provides new data on infrastructure projects in China, based on the largest existing dataset of its kind, and an evaluation of the effect that China’s domestic infrastructure investment has had on its economy.

The authors used ex-post data from a sample of 74 road and 21 rail projects in China, built from 1984-2008, to determine whether these projects were economically viable (i.e. Benefit-to-cost Ratio (BCR) ≥ 1). Transport projects were chosen due to the conventional economic hypothesis that more and better infrastructure leads to lower transport costs. The authors measured performance in terms of outturn cost (real outturn costs to real estimated construction costs), schedule (actual to estimated project implementation period) and benefits (proxied by forecasted to actual traffic volume).

The results when compared to a dataset of projects in OECD countries (806 road and rail projects for cost data, 195 projects for schedule data), were not found to be significantly different. In both China and the rest of world, a majority of the projects suffer from cost and schedule overruns. Although, the average schedule overrun for Chinese projects was significantly lower, the authors attributed this to the prioritisation of speed of delivery over quality and safety in China (rather than more effective forecasting), citing the high road fatalities in China caused by design and quality issues. However, given the proliferation of Chinese investment in international infrastructure projects, there is increasing pressure on Chinese companies to maintain their reputation, shifting the focus from speed to higher quality and greater social and environmental considerations. Thus, there is a need to consider data on more recent projects in China.

The authors conclude that China’s performance in delivering infrastructure projects does not exceed that of developed countries. In fact, a generalisation from the sample analysed suggests that less than half of China’s infrastructure investments made in the last three decades have delivered a positive NPV. The study’s results support Flyvbjerg’s classic arguments regarding megaprojects and how they are typically a “survival of the un-fittest”, being “over budget, over time, over and over again”.

The ballooning debt which has accompanied China’s rise in capital investment has macroeconomic implications such as volatile interest rates and unpredictable asset price movements, as the paper points out, some of which can already be seen. Based on the average cost overrun of 30.6% obtained from their sample, the authors estimate that roughly a third of China’s US$28.2 trillion debt pile (2014 level) can be attributed to its cumulative absolute level of cost overrun, warning that other countries should not follow China’s path. The weight of cost overruns on Chinese debt has not been very well documented, and although the authors’ figure is merely an estimate, their findings suggest that is not something investors should ignore.

Surely, poorly planned infrastructure projects leading to wasting public funds, is also synonymous of higher counterparty risk for investors and greater systemic risk. As a result, investing in Chinese infrastructure means being exposed to China’s banking sector through a simple feedback loop if multiple projects experience financial difficulties at the same time.

The authors are not the first to allude to wasteful and ineffective investments in China. While China’s progress in connecting its cities and suburbs has been impressive compared with many developing countries, its ghost cities and ‘duplitecture’ projects – many of which had investors scrambling to pull out – have raised concerns about the true value of China’s infrastructure-led fiscal stimulus. However, concluding whether China has been making poor infrastructure investments is not so simple. Particularly for projects in rural locations, cost recovery, let alone profitability, may never be reached, but the resulting increase in disposable income from increased connectivity and trade has led to poverty reduction and increased factor productivity on a gigantic scale: benefits that are not captured by the authors’ measure of economic benefits proxied by traffic forecasting errors.

In effect, focusing simply on one metric to evaluate projects risks understating their gains and costs. Eijgenraam et. al (2000) suggested a method of evaluating infrastructure projects where rather than reducing outcomes to a single figure, a cost-benefit analysis should comprise a summary of profitability calculations and quantifiable information on externalities. Furthermore, the benefits and time frame needed for the expected outcomes to materialise vary between different types of infrastructure. Thus, the sample’s representativeness can be challenged.

Still, this paper is a useful reminder of the fact that building a lot of infrastructure, even if it leads to significant and lasting economic growth, does not necessarily create an investable asset class for long-term institutional investors to pour money in. Many infrastructure project are not economic and never will be, still they can be justified on public policy grounds. Moreover, governments can decide to build infrastructure in the wrong place, for the wrong reasons, and that makes the number of uneconomic infrastructure projects even greater.

For long-term investors in search of value and predictable income streams, only a handful of projects will qualify amongst the many other infrastructure assets that governments will decide to build, whether they should or not procure them.

This is why, as we have argued before, infrastructure investment is not about concrete, but all about contracts.