By Manoj Atolia, Bin Grace Li, Ricardo Marto, and Giovanni Melina

November 9, 2017 

Versions in 中文(Chinese), Español (Spanish), Français (French), 日本語 (Japanese)

Low-income countries tend to spend less on schools than on roads as a share of GDP (photo: iStock by Getty Images).

Roads or schools? It’s a question akin to the “guns or butter” choice that governments around the world confronted in the 20th century: How to spend a nation’s finite resources to produce the maximum benefit for its people.

In our recent IMF Working Paper, we find that low-income countries tend to spend less on schools than on roads as a share of GDP—even though investment in education may be a more pressing need in their societies. 

We divide capital spending on infrastructure into two categories: economic infrastructure describes investments, such as roads, railways, ports, water, power, and telecommunications, that help the economy function better. Social infrastructure comprises capital spending, including schools, hospitals, and universities, that primarily delivers social services. We looked at data for a cross-section of low-income countries during 2000-2008 and labeled these categories roads and schools.

Shortchanging schools

Why do countries make these choices? Investment in roads can deliver a short-term increase in output, new economic opportunities, and lower poverty—even though, in the long run, capital spending on schools results in a much larger increase in output.

The tradeoff is rather stark and clear. For an average low-income country, in the long run, a permanent increase of public investment in schools of 1 percent of GDP raises output by about 24 percent, whereas an equal investment in roads boosts production by just 5 percent.

For political leaders, the critical factor may be which choice produces the quickest results, and that tips the scales in favor of roads. An investment in roads, instead of schools, produces faster economic growth for the first 13 years. By contrast, an investment in schools slows growth for nine years mainly because of the labor supply change.

Political myopia

Eventually, the payoff in growth from an investment in schools overtakes the gain from similar spending for roads. But that takes 24 years, and few leaders have such a long planning horizon. We call this condition “political myopia.”

In the meantime, the peak of the public debt increase associated with investment in schools is three times larger than that associated with investment in roads. The reason lies in upfront fiscal costs and the greater delay with which schools increase output and government revenues, which poses greater risks to debt sustainability.

This only adds to leaders’ reluctance to commit resources to schools.

Front-loading the investment, a strategy we call the “big push,” accelerates the payoff. With such a “crash” effort, the gain in growth from schools overtakes that derived from roads in about 20 years, or about four years earlier.

Short-term costs

Of course, the short-term costs of the “big push” are higher. A rapid scale-up of schools spending detracts from private consumption, labor supply, and output in the short and medium term. And the “big push” requires a higher tax and debt burden in the short run. But within 20 years, public debt as a share of GDP reverts to its original level, or lower, owing to the faster rise in output. The handicap of schools vis-à-vis roads from a fiscal perspective almost vanishes with a “big push.”

Even so, that may not be enough to overcome political myopia. Addressing the short-term concerns may require the help of multilateral agencies. Specifically, our paper recommends offering concessional financing and grants to give policymakers the incentive to emphasize investment in schools.

While tying aid to social infrastructure would address the issue of myopia, concessional terms would mitigate concerns of debt intolerance.