The Economist June 4th 2022 Finance & economics 69Red elephants?
R
arely canso much have been used by so few. During Shang-
hai’s long lockdown, which mercifully eased this week, the
city’s impressive infrastructure stood in splendid isolation from
most of the citizens it is meant to serve. The metro (all 831km of it)
was eerily quiet. The two airports, which handled 120m passen-
gers in 2019, operated at 99% below their normal level. The famous
mag-lev train neither magnetised nor levitated. Six-lane highways
provided an ocean of road space for handfuls of scooters. China is
renowned for creating “ghost cities”: new, sparsely populated dis-
tricts that gradually come to life as people move into them. Shang-
hai’s lockdown reversed this process, turning a lively metropolis
into something undead.
This surreal underuse of existing infrastructure notwithstand-
ing, the government’s best hope for reviving the economy is to add
more of it. Much more. Spending on transport, water conservation
and renovating old neighbourhoods will be a “strong driving
force” for the economy, helping to employ China’s 290m migrant
workers, said Li Keqiang, the prime minister, in an emergency
teleconference with thousands of local officials on May 25th. The
government will also “vigorously” promote 102 “major projects”,
listed in the country’s five-year plan, such as flood prevention, ul-
tra-high-voltage power lines and four-lane expressways—includ-
ing one to a city in Yunnan renamed Shangri-La.
If Omicron resurges, recurring lockdowns may prevent China
spending its way out of trouble this year. But even if everything
goes to plan, a successful stimulus will raise a deeper question.
Does China need all that additional infrastructure? Or will the ex-
tra spending leave behind superfluous “white elephants”, as un-
disturbed by human traffic as the airports, roads and railways of
locked-down Shanghai?
The question is tricky to answer, because infrastructure in Chi-
na is hard to measure or even define. The definition used by the
National Bureau of Statistics (nbs) often leaves out areas such as
gas and electricity, as well as social sectors like education and
health care. Worse, the official investment figures, designed with
central planners in mind, are not consistent with modern nation-
al accounting. Nor, owing to shifts in classification and reporting
thresholds, are they consistent with themselves over time. As Car-sten Holz of the Hong Kong University of Science and Technology
once noted, if one intended to make this data “as unusable as pos-
sible, one could probably not do a better job than the nbsdoes”.
In a paper published by the World Bank in 2020, Richard Herd
nonetheless estimates that China’s stock of infrastructure and
government capital rose from 64% of gdp on the eve of the global
financial crisis in 2007 to 107% in 2016 (the most recent figure in
his paper). This new prominence of infrastructure (and housing)
in Chinese investment may help explain the country’s productiv-
ity slowdown over the past decade. Another measure by the imf
adds up all the investment undertaken by China’s central and local
governments. According to this method, the stock of public capi-
tal was even larger: 151% of gdpin 2019, among the highest shares
in the world.
Both of these measures compare the scale of China’s infra-
structure with the size of its gdp. This convention makes some
sense: a bigger economy needs a larger backbone to support it.
Conversely, a small economy, where people are few in number or
limited in their means, can fit into a smaller infrastructural frame.
If few people can afford cars, flights or smartphones, a country
will have less need of roads, airports and 5gtowers. According to
this logic, infrastructure is a kind of “input” that should be sized
according to the scale of production.
But gdp is not the only relevant comparison. Indeed, saying
that a country’s infrastructure should be kept in proportion to its
gdp is tantamount to saying that poor countries should have poor
infrastructure. Some common components of infrastructure are
more like amenities than inputs to production. A cleaner environ-
ment, a faster bus trip or a more comfortable train journey are
things people of all income levels can appreciate. On this view,
what matters is the amount of infrastructure per person, regard-
less of their income.
Sadly, China’s infrastructure is less impressive when compared
with the size of its vast population. For example, it has 120km of
motorways per 1m people, compared with 179km in France and
326km in America. And it has 106km of railway per 1m people,
compared with 236km in Britain and over 400km in Germany.
China’s metro lines are more than 20 times as long as those in
France. But China’s population living in cities (of over 500,000
people) is also more than 20 times as large. China also has only 4.4
intensive-care beds for every 100,000 people, according to some
estimates, compared with 14 in America—a catastrophic shortage
of medical infrastructure that helps explain its lack of tolerance
for covid outbreaks. Indeed, there are 37 economies in the imf’s
database that have a higher stock of public capital per person than
China. Presumably those economies do not think that their extra
infrastructure is entirely superfluous.The longer road to Shangri-La
Critics of China’s proposed infrastructure stimulus worry that it
will crowd out other, more productive forms of spending. But in
China’s covid-wracked economy, other spending is unusually
weak. Without government help, the level of demand might not be
enough to fully employ the country’s labour and capital, including
its existing infrastructure. A recession can impose the same kind
of compulsory idleness on an economy as a lockdown. The time
and energy that China’s workers will devote to extending power
networks, waterways and roads to Shangri-La might otherwise be
lost to the economy for ever. Wasteful spending is a curse in Chi-
na. But underspending can be the most elephantine waste of all. Free exchange
China’s infrastructure stimulus will be less wasteful than a recession