Who Wins, Who Loses? Understanding the Spatially Differentiated Effects of Belt and Road Within Central Asia

Who wins, who loses? Understanding the Spatially Diﬀerentiated Eﬀects of Belt and Road within Central Asia

Somik Lall and Mathilde Lebrand

The World Bank

September 6, 2018 Motivation

I The Belt and Road Initiative (BRI) is a pipeline of new transport infrastructure and trade policy reforms, led by China, to enhance trade across countries.

I The BRI is likely to economically integrate countries but the eﬀects within countries will be geographically diﬀerentiated.

I To advise policy makers, it is important to isolate the mechanisms through which external and internal integration will aﬀect their economic development. Main questions:

I What are the spatially diﬀerentiated eﬀects of the BRI within Central Asian countries?

I What is the role of labor mobility?

I Are there complementary policies to mediate the economy’s response to external integration? Method and main results

We use a quantitative economic geography model based on Fajgelbaum & Redding (2014) that we enhanced with labor mobility frictions to produce counterfactuals to identify which locations will gain and which ones will relatively lose.

Four main results: 1. The relative winners in terms of population and real wages are districts that have a comparative advantage in exporting, would highly beneﬁt from reductions in transport costs, and are attractive places for workers.

2. Gains are concentrated in locations close to border entry points and in urban hubs.

3. Complementary investments in domestic transport networks and trade facilitation can help in spatially spreading the beneﬁts.

4. Barriers to domestic labor mobility exacerbate spatial inequality in wages whilst dampening overall welfare. What is the BRI?

The BRI: 6 main corridors with the potential of inﬂuencing 65 countries, 4.4 billion people, and leveraging 40 percent of global GDP.

SOURCE: Adapted from China-Britain Business Council, by A. Trubetskoy The quantitative general equilibrium model

Key elements of the model based on Fajgelbaum & Redding (2014): 1. Production side:

I Tradables (manufacturing, agriculture) and non-tradables (services and local manufacturing)

I Two factors of production: land (immobile) and workers (mobile) 2. Asymmetric locations (districts) in terms of productivity, amenity scores, transport costs to reach main international gateways 3. Addition of domestic labor mobility frictions:

I Preference shocks from a Frechet distribution with its main parameter as the elasticity of labor mobility to real wages.

I Iceberg costs from moving (Morten et co. 2018)

Key mechanisms:

I the role of internal geography in shaping the eﬀects of external integration

I ”the spatial Balassa-Samuelson eﬀect” Data and calibration

Data to measure productivity wedges, amenity scores and transport costs across districts for Central Asian countries (Kazakhstan, Kyrgyz Republic, Uzbekistan, Tajikistan).

I Population and employment data for districts (oﬃcial statistics, surveys, GIS maps)

I Land area for economic activities (ESA data)

I Transport times to compute transport iceberg costs (GIS analysis)

Model calibration

I Factor intensity and consumption parameters (Fajgelbaum & Redding 2014)

I Frechet parameter (Redding 2016, Morten et co 2018)

I Matrix of iceberg migration costs The scenarios for counterfactuals

Scenarios that aﬀect transport costs to reach main gateways (Moscow, Istanbul, Urumqi): 1. new or rehabilitated transport infrastructures from the BRI 2. new domestic transport hubs 3. lower trade barriers

BRI: New Silk Road Economic Belt

Moscow Perm Nizhny Novgorod Kazan Yekaterinburg RUSSIA Kiev 1767 km 485 km 60 Ufa Pre−BRI Voronezh Chelyabinsk Krasnoyarsk Kharkiv Samara Omsk Novosibirsk Odessa Ulan Ude post_BRI, without Volgograd Rostov-on-Don Harbin post_BRI, with Astana Vladivostok

Istanbul 50 Ulaanbaatar Athens Bursa 4375 km 4700 km Shenyang Izmir Ankara Anaklia 1273 km 4100 km MONGOLIA 211 km Jining Konya 77 km Tbilisi Aktau KAZ. Beijing Seoul Adana Gaziantep Baku Almaty 709 km Urumqi Tianjin 459 km Turkmenbashi Khorgos Bishkek 40 Tabriz Tashkent 300 km 789 km 300 km Cairo Tehran Ashgabat Dushanbe Kashgar Qom 365 km Shanghai 1325 km Mashhad 7322 km Isfahan Kabul CHINA 1537 km Ahvaz 870 km Havelian Peshawar Islamabad 30 IRAN Chengdu PACIFIC Faisalabad Gujranwala Lahore Legend Shiraz Quetta OCEAN Multan Existing rail Frequency Delhi Guangzhou Shenzhen 5633 km Guangzhou Shenzhen Kunming Rail improvement 3645 km Hong Kong

Gwadar 20 Hyderabad Nanning Karachi 866 km Dhaka 1824 km Planned dry port 532 km Hanoi Haiphong Kolkata Mandalay INDIA 50 km 230 km 293 km Manila Existing road 112 km Chittagong Naypyidaw Vientiane Da Nang Road improvement Mumbai 1637 km 1780 km Yangon 10 Proposed sea route 269 km Bangkok 493 km Phnom Penh 210 km Istanbul Cities over 10 million (2015) Ho Chi Minh City Chennai Bangalore Sihanoukville Can Tho Hanoi Cities over 1 million 0 Vientiane Other important cities

INDIAN OCEAN Kuala Lumpur 1441 km 100 150 200 250 Singapore Average time to reach the main gateways (hours) Results for Kazakhstan: who wins, who loses? Result 1 : The relative winners in terms of population and real wages are districts that have a comparative advantage in exporting, would highly beneﬁt from reductions in transport costs, and are attractive places for workers.

Figure: Higher real-wage growth for districts with larger decreases in transport costs (left) and higher amenity (right)

Change in real wages (in %) Change in real wages (in %)

2 3 4 5 2 3 4 5

-15 -10 -5 -3 -2 -1 0 1 2 3 Change in transport costs (in %) Log of Amenity Results for Kazakhstan: spatially diﬀerentiated eﬀects

Result 2 : Concentration of gains in locations close to border entry points and in urban hubs.

Figure: Map of the diﬀerentiated spatial eﬀects

(a) Diﬀerential of population (b) Growth of real wages (in %) Complementary policies: trade facilitation and domestic infrastructure

Result 3 : Complementary investments in domestic transport networks and trade facilitation can help in spatially spreading the beneﬁts

Figure: Welfare eﬀects without and with border-crossing time reduction (change in real wages- in %)

(a) No border improvement (b) With border improvements The role of labor mobility

Result 4 : Barriers to domestic labor mobility exacerbate spatial wage inequalities whilst dampening overall welfare.

(a) Spatial inequality in wages (b) Aggregate Utility

5 10 15 20

Change in Gini of Wages (in %)

Change in aggregate utility (in %)

0.0 0.1 0.2 0.3 0.4 0.5

1.00 1.05 1.10 1.15 1.20 1.00 1.05 1.10 1.15 1.20 Migration cost Migration cost Conclusion

I This paper uses a quantitative GE model to provide counterfactuals to understand the spatially diﬀerentiated eﬀects of BRI interventions in Central Asia.

I Districts that might lose are farther out districts that do not have a comparative advantage in exporting and are less attractive for workers to move in.

I Complementary policies, such as trade facilitation and domestic transport investments, as well as lowering domestic labor mobility barriers can mediate the economy’s response to external integration.