As the Belt and Road turns 10, it is a good time to assess the successes and failures of the initiative. Such a task is far from simple, however, complicated by the fact that the initiative has ballooned to 148 countries over the last decade and that it has never had specific goals or targets. Rather, the initiative has taken the form of a “campaign”, common in Chinese policy making and governance as a way of mobilizing actors via a call to action, but often leaving the target and methods open to interpretation, with ample space for actors to pursue their own objectives.

Nevertheless, some researchers are taking on the challenge of quantitatively assessing the BRI’s impacts in specific geographies and sectors. An article recently published in Oxford Development Studies sought to answer whether investment along the China Pakistan Economic Corridor (CPEC) has resulted in increased economic activity, one of the stated objectives of the cooperation program. The findings suggest that, while government spending in regions of Pakistan along the economic corridor saw a significant increase, economic activity not affiliated with government spending has not seen marked growth.

The findings should be cause for concern to Pakistani decision makers (and taxpayers), and raise alarm bells in other Belt and Road countries. They also point to the need for more such attempts to assess the impacts of BRI collaboration. Ten years on, all stakeholders in Belt and Road cooperation should be asking what has been achieved, what has been lackluster and what improved cooperation could look like. The answers will no doubt vary significantly across geographies.

More fundamentally, the findings of the paper are also a reminder of how difficult it is to assess the successes and failures of the BRI over the last decade when no one can seem to agree on what the original objectives were. This is more than a technical difficulty for research and economic modelling, it has become a critical weakness of the initiative.

The economic impacts of CPEC

In the Oxford Development Studies paper, David Landry, a researcher at Johns Hopkins University School of Advanced International Studies and World Bank consultant, seeks to answer two questions:

• Have Pakistani districts along the CPEC corridor seen an increase in government spending since 2013?
• Have Pakistani districts along the CPEC corridor seen a larger boost in economic activity than their counterparts not located on the corridor since 2013?

Landry uses two World Bank datasets; the Subnational Public Expenditure Database from 2012-2014 to calculate changes in regional government expenditure for the years immediately prior to and after the announcement of the BRI, and a dataset on subnational economic activity in Pakistan from 2012-2019 compiled using satellite measurements of “nightlight intensity”.

The study uses what is called a “difference in difference” methodology, a form of econometric modelling to assess changes between two groups of datasets. It is worth noting from the outset that the data is somewhat limited in the picture it provides and based on an assumption that government spending in the fiscal year 2013-14 should have an impact on non-government related economic activity up to 2019. The study also uses a quite restricted definition of the “CPEC corridor” as a single route between Kashgar in Xinjiang to Gwadar on the Pakistani coast, on which many of the “priority projects” are located. The official CPEC website in fact shows a more complicated corridor with multiple branches passing through almost all provinces of the country, while official documentation speaks of the project as consisting of “one belt, three axes and several passages” and divided into “core” and “radiation” zones.

CPEC as defined by Landry (left) v. as defined for highway networks by the official CPEC website (right).

Nonetheless, Landry’s analysis provides interesting insights.

On the first question, the analysis finds, firstly, that “districts located along the originally planned CPEC route saw a 21% greater increase in government spending than their counterparts between the 2012–2013 and 2013–2014 fiscal years.” Secondly, it found that over that period there was a shift in total government spending towards these regions.

This is perhaps not surprising. During these years construction either began or was in preparation for numerous projects including coal power plants, hydropower dams, roads and industrial zones. Given the state-led nature of these projects, that would necessarily involve the greater mobilization of government resources at a local level.

One concern Landry raises here is how CPEC appears to have pulled public expenditure toward Pakistani provinces that are already more developed, such as Punjab and Sindh, through which the main route of the CPEC runs. This risks, Landry argues, “fueling public resentment against CPEC or even the government more generally.”

Of greater interest are the findings on the second research question. Landry’s analysis of nightlight intensity finds that, while economic activity in CPEC regions did increase between 2012 and 2019 compared to non-CPEC regions (regions off the central CPEC corridor), the increase is “largely negligible after controlling for government spending.” That is to say, economic activity not associated with the increase in government spending saw little growth.

While one would expect a time lag between the initial input of public investment and the output of strengthened (non-government driven) economic activity, policy makers and the tax paying public would surely hope that over the eight years of nightlight data assessed by Landry, an economic performance would see positive impacts.

What does this mean for CPEC and the Belt and Road?

It needs to be stressed that this is a single study with limitations. The analysis model used also means that many potential variables have been excluded from the study and, as mentioned earlier, the analysis takes a somewhat narrow definition of the initiative. This points to the need for far more research on the economic impacts of the Belt and Road.

Nonetheless, the study does raise some thorny questions. Firstly, the implementation of the infrastructure-for-development model relies heavily on loans, with the understanding that subsequent economic growth will provide the returns to pay back the debt. It is a logic that has underpinned much of China’s domestic economic growth and has been raised on this blog in the context of Africa before. Landry’s study of the economic impacts of CPEC’s loan-based infrastructure development raises questions about this equation. With Pakistan already in a deep debt crisis, that is concerning.

Secondly, it raises questions about what exactly the BRI is intended to achieve in its diverse geographies. Is it economic growth, as presumed by Landry in the context of CPEC? Certainly there is a common understanding of the “Chinese model” as centered around building infrastructure in order to create markets and growth opportunities. (Though it is perhaps a simplistic understanding, given how important reforms of industry and land ownership have also been in China’s story of economic growth). But Belt and Road has been a vaguely defined and capacious initiative everywhere. At some times and for some actors it has been all about infrastructure, at other times and for others it is about trade with China, at other times it is about diplomacy and people to people exchanges. For CPEC, for example, there are a total of 10 broadly defined baskets of cooperation listed on the official CPEC website.

While for the last decade this has been a convenient “catch all” approach to mobilize domestic actors under the Belt and Road campaign and form “win-win” alliances based on long lists of promising opportunities, it is problematic when governments come to assess what has been achieved through the initiative.

This was on display recently in the debates around Italy’s potential departure from the initiative. While Italian politicians argued the initiative had done little to re-balance the country’s trade deficit with China, China’s foreign ministry spokespeople were pointing to the numbers on growth in overall trade. Both sides are correct. But neither side agreed on which of these targets the partnership was intended to achieve. Similar to CPEC’s stated aims, the original Belt and Road cooperation agreement signed by both sides listed six areas of cooperation, “trade and investment” forming just one. As a result, the Belt and Road can easily be pulled into domestic politics.

Are Chinese policy makers waking up to this? Comments relating to the initiative’s 10th anniversary from the Ministry of Foreign Affairs spokespeople in the last few months have included hard numbers on what has been achieved and what is expected to be achieved as a result of Belt and Road cooperation. They focus on poverty alleviation, jobs and income, and, most importantly, volumes of trade.

“Over the past ten years…the BRI has created 420,000 jobs and lifted almost 40 million people out of poverty outside China. According to data from the World Bank, by 2030, Belt and Road cooperation is expected to increase trade by between 2.8 and 9.7 percent for participating countries and between 1.7 and 6.2 percent for the world, and this is expected to increase global real income by 0.7 to 2.9 percent.” (Foreign Ministry Spokesperson Mao Ning. July 19, 2023) 

Is this the new, more focused definition of Belt and Road cooperation as the initiative enters its second decade? In its first decade, Belt and Road promised everything. Is it now narrowing down on trade? And just how well matched are these expected achievements with participating countries’ diverse objectives of addressing trade imbalances, moving up global value chains and tackling infrastructure deficits?