The Belt and Road Initiative (BRI) or One Belt One Road (OBOR) initiative was announced in the Fall of 2013 by President Xi Jinping of the People’s Republic of China (PRC). The President invoked the ancient Silk Road announcing the plans to bring Asia, Europe, Africa, and the Middle East closer together by constructing investment and trade networks and creating new institutional linkages. The BRI could promote trade, more efficient resource reallocation and strengthen economic growth across the region. It could also encourage countries to coordinate economic policy and improve regional collaboration. The goals of our analysis are to i) study the impacts of infrastructure improvements on BRI and non-BRI countries’ trade flows, growth and poverty; and ii) suggest policies that would help maximize gains from the BRI-induced trade cost declines. To our best knowledge, there are only a few CGE-based studies on the impacts of the BRI.
With moderate assumptions on BRI investment, the simulation results indicate a global welfare gain of 1.3% of global GDP by 2030 with a boost to global trade by 5%. The vast majority of gains applies to BRI countries. Some assume that BRI leads to a 25% reduction in road transport margins and 5% in sea transport margins, as well as a significant reduction in time to import by BRI countries due to the accompanying trade facilitation measures. The authors estimate that BRI GDP growth rates could increase by 0.1-0.7 percentage points, while total exports of BRI countries could increase by between $5 billion and $135 billion depending on the assumptions on trade costs reductions. Our analysis adds value to the existing research by incorporating the impacts of BRI-related infrastructure improvements based on bilateral trade time reductions from de Soyres et. al. (2018). In addition, while the CGE analysis provides impacts on several economic variables (e.g. trade, employment, economic growth), we also study the impacts on poverty and greenhouse gas emissions. We also assess the role of trade policy reforms in maximizing the gains from BRIinduced infrastructure improvements.
It should be noted that we abstract from the economic impacts of the infrastructure expenditures. These are unlikely to have major repercussions on the long-term real income gains but could affect the timing of the gains depending on the source of the financing, i.e. domestic versus foreign. Foreign inflows could impact the short-term gains through their effect on the real exchange rate, unless there is strong leakage of the inflows in terms of imported goods and services. If investment was funded by higher government taxes, it would likely lower household consumption with impacts on private investment and relative prices. In addition, our modeling approach misses some of the channels that would be likely to increase the gains from the BRI such as additional foreign direct investment flows or new products or markets where countries might become competitive following implementation of BRI as in our approach intensive and extensive margins are determined by the existing trade flows. Furthermore, there is significant uncertainty regarding the projects to be implemented under BRI and their potential for trade cost reductions. Our analysis uses the best available assumptions, but the impacts of the BRI will ultimately depend on the specific investments and their effectiveness in reducing trade costs.
This study employs the estimates of de Soyres et. al. (2018) to generate the BRI-induced trade costs reductions. The authors use network analysis to quantify the impact of the BRI on connectivity between countries before and after the infrastructure investments. The network model covers more than 1,000 cities and incorporates the current network of rail and maritime links to compute the shipment costs between all cities using a shortest path algorithm. The algorithm factors in the cost of a given transport path taking into consideration distance, travel time and vehicle operating costs. The analysis covers cities with population over 500,000 as well as the three most populous cities in the country. The network is generated by solving for each city pair the routing determined by the shortest time. The limitation of the methodology is that it fails to include road and air transport features in the current application. Other limitations are also discussed in the study by de Soyres et. al. (2018) such as lack of information on rail gauge, service frequency port facilities etc. Another key challenge was posed by the identification of BRI-related projects where timelines and specific locations were not always available. Nevertheless, the authors generate a consistent data set of BRI-related trade costs reductions in rail and maritime transport costs including switching across the two transport modes. These trade costs reductions are then applied to ad valorem equivalents of value of shipment day generated based on updated estimates of Hummels and Shaur (2013). The time in trade estimates are based on US imports data and variations in the premium paid for air shipping versus maritime shipping. The average delay of one day in shipping is estimated to be associated with ad valorem tariff of 4.9% in the baseline. However, the data is available at the HS2 digit level and aggregated to GTAP sectors and country pairs based on 2014 trade flows.
In the baseline, the trade among BRI countries air transport accounts for nearly 12% of total transport cost, land is around 31% and sea is around 58%. These shares vary dramatically across exporting countries—with high sea shares for island economies such as Indonesia and the Philippines. Therefore, our analysis covers a significant share of total transport services. In the baseline, the highest time in trade estimates are estimated in products of animal origin, cereals, fruits and nuts, pulp and paper; while products the least sensitive to time in trade covered special woven fabrics, silk, man-made fibers and filaments.
The baseline scenario assumes the continuation of past trends with no decline in trade costs. In the counterfactual scenarios we study the impacts of the BRI-induced trade cost reductions. We apply the lower bound estimates of trade cost reductions due to BRI-related infrastructure improvements from de Soyres et. al. (2018) to generate the low case scenario. The upper bound estimates allow for switching across transport modes leading to deeper reduction of trade costs and form our high case scenario. Further gains could be realized if in addition to improving the transport infrastructure, the trade facilitation reforms would lead to halving of border delays for BRI countries. Finally, we augment the last scenario with two variants of complementary policies where BRI countries implement a 50% reduction of tariffs in trade with each other or reduce their tariffs in trade with all trading partners.
The shock occurs in two time periods. One-half of the shock is assumed by 2025, and the remainder by 2030. The fourth and last simulation involves tariff reductions applied on MFN basis. Most of BRI-induced bilateral trade costs reductions are below 10%, but several exporters experience higher trade costs reductions to most destinations. Those biggest cost reductions on trade costs of exporting are expected in Kyrgyz Republic, Kazakhstan, Ethiopia, the Lao People’s Democratic Republic and Cambodia. We would expect those countries to be the biggest beneficiaries of the BRI. When analyzing the bilateral trade costs reductions by using the example of imports by China, we note that the differences across sectors of economic activity are very small, but systematically vary by the source of exports. Again, South Asia experiences the biggest trade cost reductions, while countries that are not part of the BRI such as the United States or Latin American and Caribbean countries (LAC) register very small gains (less than 1%) from better maritime connectivity of China with other regions.
BRI related investments are expected to lead to an increase of returns to factors with workers being relatively better off than capital and land owners. Table 11 provides some insights to the distributional impacts of the BRI initiative. It shows the change to aggregate real factor returns (deflated by the regional CPI) in 2030 relative to the baseline. For the BRI area in aggregate, the return to labor has a higher increase than the return to capital (including land and natural resources), 1.37% versus 0.87%. Unskilled workers would see a gain (1.36%) and skilled workers a slightly higher gain (1.38%). However, there are wide variations across regions. For example, Lao PDR and Thailand witness a fairly sizeable increase in skilled wages relative to unskilled. Pakistan and Kyrgyz Republic see the opposite. Likewise, the relative gains between labor and capital vary widely across regions. Land returns increase in all BRI regions, especially in Pakistan and Bangladesh and some regions see a decline in returns to natural resources. Kyrgyz Republic is again a significant outlier with returns to natural resources increasing by 13.2%. The changes to real factor returns tend to be smaller for the non-BRI area. The impact of the BRI investments on labor displacement would be moderate.
For the BRI area, total displacement is some 12 million workers, or 0.48 percent of the baseline labor force. This is a relatively small amount, particularly as we assume that there is a transitional phase for the initiative. The EAP region is expected to lose agricultural employment of about 0.8 million, while South Asia would gain over 4 million workers in the agricultural sector. Overall as a result of the BRI infrastructure investments the largest share of labor force of 0.9 percent is expected to switch jobs in EAP, followed by the SSA and MENA region where about 0.6 and 0.5 percent of labor force respectively would switch jobs. The BRI area could see a net loss of almost 0.8 million agricultural workers (in 2030 relative to the baseline). The majority of this would be in China, though many other regions would see agricultural employment losses as well such as Malaysia and Thailand. Bangladesh, Pakistan and India would see some significant increases in agricultural employment, as would Kenya and Tanzania. The displacement effect is significantly lower in the non BRI area.
While impacts would vary by countries, income gains under the BRI scenario would lift several million people from poverty relative to the baseline scenario. There will be winners and losers in the short run, but if BRI investments materialize the poor in both BRI and non-BRI countries would likely experience welfare gains. To evaluate the effect on poverty in countries that are likely to be the most affected, i.e. low and middle-income countries, the use of extreme and moderate international poverty lines at PPP$1.90 and PPP$3.20 a day are preferred. Globally, BRI related investments could lift 7.6 million from extreme poverty and 32 million from moderate poverty. Developing countries in the BRI will benefit the most from the reductions in extreme poverty of 4.3 million and moderate poverty of 26.7 million.
In countries like Kenya and Tanzania, an additional 0.7 million poor people would be expected to be lifted from extreme poverty, at PPP$1.90 a day. This is approximately equivalent to an additional 1.0 and 0.9 percentage points reduction in the extreme poverty headcount rate. In South Asia, Pakistan would see some significant additional reductions in extreme poverty with 1.1 million people being lifted compared to the baseline. Bangladesh and India are expected to see a smaller number of people lifted out of poverty i.e. 0.2 million (0.11 percent of headcount) and 0.03 million (0.002 percent of headcount), respectively. In Nepal, the scenario with BRI infrastructure investment alone lifts an additional 60 thousand people out of extreme poverty compared to the baseline. In East Asia and Pacific, Philippines expects to see approximately 90 thousand people lifted out of extreme poverty compared to the baseline.
Several countries will be particularly affected by the infrastructure improvements related to the BRI. We focus on the upper bound infrastructure improvements scenario (BRIUBD) to understand the specific impacts on those countries. Pakistan presents the highest welfare gain of the countries involved in the initiative, with a gain of 10.5% by 2030 (relative to the baseline), and an 8.6% share of the total gain of the BRI area (Table 5) These gains are the result of trade cost reductions originated by ventures such as the improvement of the Port of Gwadar connections, through highway, rail and pipeline infrastructure, as part of the China-Pakistan Economic Corridor Plan. Other projects include a Peshawar-Karachi Motorway, and an expansion and reconstruction of a railway between Karachi to Peshawar. The sector that shows the sharpest reductions in import trade costs is petroleum and coal products, but sectors such as construction, trade services and the transport sectors also exhibit high reductions. As a result of trade costs reductions, imports of several products increase (Table 9b). The biggest increases are recorded in agricultural goods, textiles and petroleum and coal products. Pakistan is importing less oil, chemical, rubber and plastic goods and less transport equipment than in the baseline. Lower costs of imported inputs and higher demand from abroad lead to expansion of exports of Pakistan in several sectors (Table 8b). The highest increases are recorded in chemicals, rubber and plastics, processed foods and other manufacturing, which consequently contributes to the highest rates of output growth in these sectors. On the other hand, the sectors that show the highest reductions in exports include agriculture and leather goods.
The Kyrgyz Republic is the BRI country that has the second highest percentage welfare gains, with 10.4% by 2030 (relate to baseline), however it is one the lowest in absolute values, with only 1.3 billion dollars increase (in 2014 prices). Thanks to the gains obtained by the BRI projects that focus on transportation - mostly railway lines and roads, most sectors in the economy benefit from the high trade cost reductions. Kyrgyz Republic imports increase across all sectors, with machinery and equipment, textiles, leather goods and petroleum and coal products showing the highest increases. Agriculture records a sharp reduction in both exports and domestic products, creating a significant increase of agricultural imports. In terms of exports, Kyrgyz Republic sees an increase in the sectors of leather goods, coal, and machinery and equipment, as well as exports of trade services and other transport driving the expansion of output in those sectors. Gas and metal products show the highest reductions in output.
The Lao PDR is expected to record a welfare gain amounting to 3.1% by 2030 (relative to the baseline). This is directly linked to the high level of trade cost reductions due to infrastructure improvements (see Table 4). These important gains will be associated with a faster access to the sea therefore benefiting trade with partner countries that can be reached through maritime connections. Among the projects that impact Lao PDR, the BRI related investments cover the new rail link from Vientiane to the Bangkok port in Thailand, but also the improvement of the Sihanoukville port in Cambodia while all the rail improvements in Vietnam also play a role. In addition, Kra canal in Thailand is particularly beneficial for all sea shipments going west – this will prevent ships from taking a long detour. The biggest decline of trade cost on imports is expected in sectors such as construction, trade services, paper products and, coal. These sectors also experience the increase of imports, but the biggest volume increases are recorded in machinery and equipment, chemicals, rubber and plastics and processed foods. The largest increases in the volume of exports of Lao PDR are recorded in chemicals, rubber and plastics as well as in energy intensive manufacturing. On the other hand, the exports of agriculture, wearing apparel and processed foods show the highest decreases. Lower exports also drive output of these sectors to decline. Overall, Lao PDR output seems to be growing in most sectors, with trade services, other transport and hospitality services gaining the most value.
Ethiopia is not on our list of BRI countries, yet it is also expected to reap significant benefits from the reduction of trade costs and border cost delays among the BRI countries. The expected welfare gain amounts to 1.9% by 2030 (relative to the baseline). The biggest decline of trade costs on imports applies to Kyrgyz Republic, Kazakhstan, Cambodia and other countries in both East and South Asia and apply to coal, electricity and machinery and equipment. Faster increase of imports (relative to the baseline) applies mostly to metal products, and machinery and equipment and transport equipment. Ethiopia is expected to increase its exports of agricultural products, leather goods and energy intensive manufacturing. The business and hospitality services and metal products show a reduction in exports. The trade changes impact output generating significant expansion of output in agriculture, processed foods, leather goods and selected services sectors. Several other sectors exhibit a decrease of output, as resources are redirected to most profitable sectors. This applies to paper products, energy intensive manufacturing and metal products. The structure of the economy seems to move towards the primary sector – agriculture, with manufacturing growing only slightly and services experiencing a decline.
Finally, China also benefits significantly from the BRI-induced trade cost reductions. The welfare gain amounts to 0.7 % by 2030 (relative to the baseline). The sectors experiencing the largest trade cost reductions include petroleum and coal products, and water, air and other transport sectors. Imports of several products increase much more than in the baseline including electronics, chemicals, rubber and plastics, with agricultural goods experiencing the biggest increase in volume of imports. On the other hand, China shows a much higher increase of exports of electronics, machinery and equipment and chemicals. Since there is a small reduction of output of agriculture and to accommodate its growth of exports, China will meet the increased external demand by increasing imports of this sector. Other sectors with high reductions of output are coal, oil, air transport and electronics. Sectors that show the highest increases in output are wearing apparel, metal products and machinery and equipment. The BRI induced trade cost reductions contribute to restructuring of Chinese output away from agriculture towards services with some positive impacts on manufacturing.
Our findings indicate that the BRI would be largely beneficial, but some countries outside the initiative could suffer from trade diversion. As a result of the BRI, global real income increases by 0.7 percent (in 2030 relative to the baseline), which in comparative terms is relatively sizeable as upper estimates of the real income impact of global free trade are around 1 percent. This translates to almost half a trillion dollars in 2014 prices and market exchange rates. The BRI Area captures 82 percent of the gain, with China garnering 36 percent of the total global gain. In percentage terms, the largest return accrues to the East Asia regional aggregate seeing an increase of 2.2 percent in overall real income. The non-BRI area sees some gains with an increase of 0.2 percent, most of which is captured by the European Union and the Rest of high-income region (which is dominated by the high-income economies of East Asia). These latter two regions, though not formally part of the BRI area, are the most integrated economies with the BRI area. There are minor losses for the regions in the Western Hemisphere.
BRI related investments can contribute to lifting 8.7 million people from extreme poverty and 34 million from moderate poverty at the global level. Under baseline conditions, the percentage of people living in extreme poverty, with less than PPP$1.90 a day, is projected to decline from 10.1 percent in 2015 to 5.2 percent by 2030. With infrastructure investments the BRI can additionally lift from extreme poverty up to 8.7 million people. These benefits extend to both BRI and nonBRI countries: 5.1 million from BRI area and 3.7 million from non-BRI countries. Our estimates indicate that as a result of BRI related trade cost reductions changes in volume and structure of economic activity would have negligible impacts on CO2 emissions with an aggregate increase of 0.5% at a global level. The BRI region itself sees an increase of 0.6%, but with China showing no change. There is considerable heterogeneity across countries and regions.
Taken from: Maryla Maliszewska and Dominique van der Mensbrugghe: The Belt and Road Initiative Economic, Poverty and Environmental Impacts, published by the World Bank.
