WPS7614 Policy Research Working Paper 7614 Deal or No Deal Strictly Business for China in Kenya? Apurva Sanghi Dylan Johnson Macroeconomics and Fiscal Management Global Practice Group March 2016 Policy Research Working Paper 7614 Abstract Existing work on China’s economic influence in Africa consumption-driven economy which will increase demand refers to Africa in broad terms, thereby generalizing the for services, a growing strength of Kenya’s economy (World results to an extent that is unhelpful for policy-makers Bank Country Economic Memorandum 2016). The in a specific country. Moreover, the emphasis is on oil paper emphasizes that Kenyan policy makers should be exporters. This paper remedies this by focusing on a single, less concerned about bilateral trade imbalances and worry oil-importing country: Kenya. The paper examines China’s about Kenya’s overall trade balance. However, the Stan- economic presence in Kenya and some of the popular myths dard Gauge Railway and Thika superhighway experiences surrounding Chinese economic activity. The first myth suggest that Chinese firms offer relatively few technology is that Chinese companies do not employ local workers. transfer or supplier opportunities for local firms and aca- In fact, 78 percent of full-time and 95 percent of part- demia. Third, the popular focus of Chinese competition time employees in Chinese companies are locals. Second, is on the impact on well-organized Kenyan producers and although China represents a large potential market for local not on consumers, thereby underestimating the benefits exporters, the study finds that China has a better chance of Kenyan consumer derive from the availability of more expanding its exports to Kenya than Kenya does to China affordable Chinese goods. The paper concludes with based on existing specializations. This may change with policy directions for improving export competitiveness and recent oil discoveries in Kenya, increasing the space for transparency in infrastructure projects, and local content. Kenyan exports to China, as well as from China’s shift to a This paper is a product of the Macroeconomics and Fiscal Management Global Practice Group. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://econ.worldbank.org. The authors may be contacted at asanghi@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Deal or No Deal: Strictly Business for China in Kenya? Apurva Sanghi∗ Dylan Johnson†‡ JEL-Classification: F14, F21, F35 Keywords: Trade Imbalance, Foreign Direct Investment, Foreign Aid ∗ Lead Economist for Kenya, Uganda, Rwanda and Eritrea, World Bank Group. † Consultant, World Bank Group ‡ We are grateful to Michel Botzung, Deborah Br¨ autigam, Paul Brenton, Kevin Carey, Guang Zhe Chen, Shanta Devarajan, Nora Carina Dihel, Marcelo Giugale, Justin Lin, Manuel Moses, Thomas O’Brien, Anand Rajaram and David Tarr for excellent comments and feedback. All remaining errors are our own. A. Sanghi and D. Johnson Contents 1 Introduction 1 2 Kenya and China’s Trade Relationship 3 2.1 The common belief: Exports from poorer countries are commodity-dependent . . . . 3 2.1.1 Kenya should pay more attention to the overall trade balance . . . . . . . . . 3 2.2 China is a large source of imports for Kenya . . . . . . . . . . . . . . . . . . . . . . . . 4 2.2.1 Kenya imports rubber manufactures from China . . . . . . . . . . . . . . . . . 6 2.3 China is still a small export market for Kenya . . . . . . . . . . . . . . . . . . . . . . . 7 2.3.1 Kenya exports raw goods and metals to China . . . . . . . . . . . . . . . . . . 8 2.4 Winners and Losers from Kenya’s current trade patterns . . . . . . . . . . . . . . . . 10 2.4.1 The net benefit to the economy is positive . . . . . . . . . . . . . . . . . . . . . 10 2.4.2 Consumers and retailers gain . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 2.4.3 Producers are worse off; some benefit from Chinese intermediate goods . . . 10 2.5 What drives Kenya’s exports? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 2.5.1 Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 2.6 Brighter prospects for services exports as China rebalances . . . . . . . . . . . . . . . 16 2.6.1 China’s rebalancing will help reduce poverty by 2030 . . . . . . . . . . . . . . 17 3 Foreign Direct Investment in Kenya 18 3.1 Chinese FDI in Sub-Saharan Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 3.1.1 Chinese FDI in Sub-Sharan Africa is relatively small . . . . . . . . . . . . . . . 20 3.2 Chinese companies in Kenya: From large state companies to small private ones . . . 24 3.2.1 Chinese invest the most in metals, communications, and automotive original equipment manufacturing (OEM) . . . . . . . . . . . . . . . . . . . . . . . . . 24 3.2.2 Corruption is the biggest obstacle for Chinese firms . . . . . . . . . . . . . . . 26 3.2.3 Chinese firms source most materials from China . . . . . . . . . . . . . . . . . 26 3.2.4 Chinese firms employ a large share of local workers . . . . . . . . . . . . . . . 27 3.2.5 Chinese firms face some competition from the informal sector . . . . . . . . . 30 3.2.6 Chinese companies less likely to take credit line . . . . . . . . . . . . . . . . . 30 4 Official Development Assistance from China 30 4.1 Chinese aid is small compared to commercial activities . . . . . . . . . . . . . . . . . 30 4.1.1 Kenya may rely more on Chinese aid because of volatile aid flows . . . . . . 31 4.2 Most Chinese financing does not qualify as ODA . . . . . . . . . . . . . . . . . . . . . 32 4.2.1 China loans the most to ministries of energy and petroleum and transport and infrastructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 4.2.2 Chinese stands out in education aid . . . . . . . . . . . . . . . . . . . . . . . . 33 4.3 Lack of quality Chinese aid data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 4.3.1 Drawbacks of existing Chinese aid data: Media based data collection is prob- lematic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 2 5 Review and future directions 34 5.1 Consider the long term growth of local industry . . . . . . . . . . . . . . . . . . . . . 35 5.1.1 SGR spotlight: Non-existent capacity building . . . . . . . . . . . . . . . . . . 35 5.2 Diversify FDI sources to avoid overreliance on China . . . . . . . . . . . . . . . . . . 36 5.3 Monitor debt levels from China . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 5.3.1 Debt to China is growing quickly . . . . . . . . . . . . . . . . . . . . . . . . . . 37 5.4 Supply-side Shortages: Reducing labor costs . . . . . . . . . . . . . . . . . . . . . . . 38 5.4.1 Encourage technology transfer and capacity building with infrastructure projects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 5.4.2 Bring more transparency to loans and infrastructure projects . . . . . . . . . . 39 5.4.3 Special Economic Zones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 5.4.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 Appendices 47 Appendix A Gravity model of trade 47 A.1 Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 Appendix B Gravity Data 49 Appendix C Results 50 List of Figures 1 Kenya’s overall trade balance is a bigger concern Central Bank of Kenya 2015 2 China takes a sizeable share of total Kenyan imports (2012-2014) IMF World Economic Outlook Database 2014 3 China and India are major sources of imports (2004-2013) . . . . . . . . . . . . . . . . 5 4 Kenya imports rubber manufactures from China (2010-2014) UN Comtrade database 2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 5 Kenya imports rubber footwear, tires, and fabrics from China (2012-2014) UN Comtrade database 2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 6 Kenyan exports to China small relative to total exports (2012-2014) IMF World Economic Outlook Database 2014 7 Kenya exports metals and hides and skins to China (2010-2014) UN Comtrade database 2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 8 Kenya exports titanium ores and copper to China (2012-2014) UN Comtrade database 2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 9 Metals prices have been falling since January 2010 (1990-2015) World Bank Commodity Price Data (The Pink Sheet) . . . . . . . . . . . . . . . . . . . . . 9 10 Kenya’s imports of intermediate goods is rising quickly UN Comtrade 2015 11 China’s share of Kenya’s intermediate goods imports is growing UN Comtrade 2015 A. Sanghi and D. Johnson 12 Kenya’s apparel exports to the US (2000-2014) UN Comtrade 2015 13 Kenya’s services exports overall are strong UN Service Trade 2015 14 Kenya’s FDI is low. Trend in red (1980-2014) World Development Indictors World Bank 2015 . . . . . . . . . . . . . . . . . . . . . . . . 19 15 Kenya underperforms in attracting FDI relative to potential. Trend in red (1980- 2014) S World Development Indictors World Bank 2015 . . . . . . . . . . . . . . . . . . . . . . . . 19 16 Kenya’s Gross domestic savings has fallen sharply since 1993. Trend in red (1980- 2014) S World Development Indictors World Bank 2015 . . . . . . . . . . . . . . . . . . . . . . . . 20 17 Chinese FDI in Kenya is growing quickly since 2009 (2003-2012) UNCTAD FDI/TNC database 2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 18 China’s FDI represents large share of total FDI (2003-2012) Authors’ own calculation based on World Development Indictors World Bank 2015 . . . . . 21 19 Investment from China rising; investment from UK and US falling KenInvest 2015 20 China and France top sources of FDI inflows for Kenya (2012) UNCTAD FDI/TNC database 2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 21 China and UK hold the most FDI stock in Kenya (2012) UNCTAD FDI/TNC database 2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 22 Kenya’s mobile telephone subscriptions grew quickly over last decade International Telecommunications Union 2016 23 China invests most in metals and communications in Kenya (2003-2015) fDi Intelligence from The Financial Times Ltd 2015 . . . . . . . . . . . . . . . . . . . . . . 25 24 Chinese companies invest most in manufacturing (2003-2015) fDi Intelligence from The Financial Times Ltd 2015 . . . . . . . . . . . . . . . . . . . . . . 25 25 Chinese investment per project highest in headquarters creation and manufacturing (2003-2015) fDi Intelligence from The Financial Times Ltd 2015 . . . . . . . . . . . . . . . . . . . . . . 26 26 Chinese companies create most jobs in the communications sector (2003-2015) fDi Intelligence from The Financial Times Ltd 2015 . . . . . . . . . . . . . . . . . . . . . . 28 27 Chinese companies in manufacturing sector create the most jobs (2003-2015) fDi Intelligence from The Financial Times Ltd 2015 . . . . . . . . . . . . . . . . . . . . . . 28 28 Chinese companies in manufacturing sector create the most jobs per project (2003- 2015) S fDi Intelligence from The Financial Times Ltd 2015 . . . . . . . . . . . . . . . . . . . . . . 28 29 China is fifth largest creator of jobs. India creates the most jobs (2003-2015) fDi Intelligence from The Financial Times Ltd 2015 . . . . . . . . . . . . . . . . . . . . . . 29 30 US is a top donor for Kenya. Chinese finance does not meet OECD/DAC aid criteria (2013) OECD DAC aid database 2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 4 31 Kenya receives the most aid in health and populations services (China excluded) (2013) OECD DAC aid database 2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 32 Ministry of Energy and Petroleum receives most loans from China (2014) Estimates of Development Expenditure Government of Kenya (2014) . . . . . . . . . . . . 33 33 Kenya’s debt to China is growing quickly (2010-2014) Kenya National Bureau of Statistics Economic Survey 2015 . . . . . . . . . . . . . . . . . . 37 34 Kenya owes most debt to China (2015) The National Treasury of Kenya (2015) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 35 Kenya’s debt to China outpaces the rest (2015) The National Treasury of Kenya (2015) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 C36 Trade and Distance Kenya and Rest of the World (1948-2014) CEPII Gravity Database 2010 and UN Comtrade 2015 C37 Trade and Importer GDP per capita Kenya and Rest of the World (1948-2014) CEPII Gravity Database 2010 and UN Comtrade 2015 C38 Trade and Importer GDP per capita Kenya and Rest of the World (2014) CEPII Gravity Database 2010 and UN Comtrade 2015 List of Tables 1 Estimates of Kenya’s Trade Flows (1948-2014). Dependent variable: Kenya’s exports 14 2 Estimates of China’s Trade Flows (1948-2014). Dependent variable: China’s exports . 16 3 FDI Projects and related employment (2007-2011) . . . . . . . . . . . . . . . . . . . . 29 4 Export Processing Zones sector contribution in 2012 (%) . . . . . . . . . . . . . . . . . 41 C5 Estimates of Kenya’s Trade Flows (1948-2014). Dependent variable: Kenya’s exports 50 C6 Estimates of China’s Trade Flows (1948-2014). Dependent variable: China’s exports . 53 C7 Estimates of China’s Trade Flows (1948-2014). Dependent variable: China’s exports . 56 C8 Estimates of Kenya’s Trade Flows (1948-2014). Dependent variable: Kenya’s exports 59 1 Introduction In recent years, China’s economic presence in Sub-Saharan Africa has risen rapidly. China’s growth in the region is driven in part by its strong demand for raw materials, and resource rich countries that manage the boom well may also translate the gains to the broader economy, work- ing to pay down high public debt or alleviate poverty. But the countries that benefit from the boom are also more vulnerable to China’s economic slowdown. Oil-importing countries such as Kenya will be shielded from China’s slowdown and should even see an increase in their exports. Kenyan exporters of services such as tourism will fare well as China transitions to a consumption- based economy by 2030. Greater Chinese consumption may also benefit Kenyan producers in the horticultural sector that are taking advantage of the trend of selling directly to large supermarkets in Asia. Supermarkets in China can also recieve Kenyan flowers if Kenya succeeds in negotiating duty-free access for cut flowers as part of the 404 duty free products from African countries. Exporters of flowers are performing well, but producers of manufactured goods face more competition from China in both domestic and foreign markets. Many fear that local producers will be hurt by Chinese imports; cheap plastic shoes and clothes from China, and second-hand clothes in general, are much more popular than local products. In addition, Kenyan exports of clothing to the United States, for example, lost market share to China between 2004 and 2006, and have only recently begun to recover. The manufacturing sector grew slowly at only 3.4 percent in 2014, down from 5.6 percent in 2013, and some worry that slower growth could be a sign of a pre- mature decline of industry (Chen, Geiger, Fui 2015). Without a turnaround in manufacturing, the growth potential of the economy is limited. But Kenya can enhance its growth in manufacturing if it continues to attract foreign direct investment from China. A large share of foreign direct investment (FDI) already comes from China, allowing Kenya to diversify its sources of FDI and increase investment in manufacturing. Lagging behind countries such as Ghana, Nigeria and South Africa, Kenya performs poorly in attracting manufacturing FDI. To increase the low investment, Kenya wants to market opportunities to China because Chinese firms are attracted to the low cost of labor in Kenya. The lower wages, however, come with lower productivity, raising the unit cost of labor; at the moment, the unit cost of local labor is higher than in China, making Kenyan workers more expensive than Chinese ones. If Kenya reduces the unit cost of local labor, it will attract more Chinese investment in labor-intensive industries, providing jobs and helping reduce poverty. There is strong potential for poverty reduction in the textile and garments industry because it mainly employs women, who tend to increase the household sav- ings rate. China also offers critical financing in sectors that traditional investors overlook: infrastructure and construction. China’s loans compete with loans from traditional donors that attach conditions of good governance and transparency. Uninterested in the politics of the country, China funds major infrastructure projects in Kenya. One such project is the Standard Gauge Railway linking Nairobi and Mombasa by the China Road and Bridge Corporation, and other Chinese construc- tion companies are taking advantage of the real estate boom in Nairobi. Following the slowdown in China, marketing for construction services should increase globally, and even more Chinese companies may come to Kenya to undertake major infrastructure and construction projects. The 1 A. Sanghi and D. Johnson improvement in insfrastructure will help lower the cost of doing business, attract more invest- ment, and enhance productivity. We contribute to the literature by investigating China’s impact on single, oil-importing coun- try, Kenya. Oil-importers receive little attention in the existing literature, and researchers and journalists usually highlight Chinese demand for land and natural resources in Africa, ignoring the useful role China plays as an infrastructure provider and source of cheap goods for consumers and retailers. In Kenya, Chinese firms invest in more than just natural resources. They also invest large amounts in the communications and automotive original equipment manufacturing sectors. Previous work on China in Africa refers to Sub-Saharan Africa in broad terms and fails to pro- vide specific examples and guidance for individual countries navigating relations with China.1 We avoid overgeneralizing by examining the trade, aid, and foreign direct investment between China and Kenya. Chinese investment is more than just state-owned companies negotiating directly with the government. Many are private companies looking for access to the domestic market or to produce goods for export to Europe or North America (Br¨ autigam 2013; SACE 2014). The manu- facturing and service sectors attract a number of small and medium enterprises, and construction draws larger companies. Some bid for tenders from the Ministry of Commerce and receive sup- port based on performance; others raise capital from family and friends in China; state-owned firms can access subsidized credit from the China export-import (EXIM) bank. But the size, oper- ations, and financing of Chinese firms is quite diverse, a diversity that is often overlooked when discussing Chinese investment in Kenya. We also find that businesses employ a majority of local workers in full-time and part-time roles; the majority of surveyed firms also report having a policy to localize its workforce, chal- lenging the stereotype that Chinese firms only use Chinese labor. Workers also receive basic skills, safety and hygiene training (SACE 2014). Chinese firms can offer even more training if Kenya promotes local capacity building and technology transfer —The WTO’s trade related investment measures (TRIM) forbids local content requirements. Mega-infrastructure projects undertaken by Chinese companies are valuable learning opportunities for local industry and training institutes; they allow well-organized firms to upgrade equipment and supply materials for both current and future projects. Experience from the Standard Gauge Railway (SGR) linking Mombasa to Nairobi has shown that without a strategy for knowledge sharing, local firms will miss out on the spillover effects from investment, a crucial part of increasing competitiveness of the domestic economy. 1 Similiar works include Onjala (2008), Chege (2008), Subramanian (2008) and Fiott (2006) that address Kenya specif- ically and provide a detailed account of China’s involvement in Kenya from 1960 to 2006. Kaplinsky (2006, 2007) explains China’s impact on clothing and textile exports from Lesotho, South Africa, and Kenya after the expiry of the MultiFibre Arrangement (MFA). Zafar (2006) sheds light on the effects of China’s global macroeconomic presence on Sub-Saharan Africa, where he identifies winners (mainly oil exporting economies) and losers (mainly oil importers such as Kenya). 2 2 KENYA AND CHINA’S TRADE RELATIONSHIP 2 Kenya and China’s Trade Relationship 2.1 The common belief: Exports from poorer countries are commodity-dependent Many African economies have benefitted from China’s strong demand for energy and metals. One expects Kenya to mostly export commodities to China, and China to export a greater variety of manufactured goods to Kenya. China is a richer middle income country and has greater oppor- tunities to expand its exports in different sectors; as a lower middle income country, Kenya relies more on agricultural and commodity exports. But Kenya’s exports are relatively more diverse: it exports metals and plastics, but also vegetable textile fibers and leather rawhide skins, presenting an opportunity to meet the demand for finished and crust leather in the EU and China (Hansen, Moon, Mogollon 2015). China mostly exports rubbers and plastic products to Kenya, suggesting an overspecialization in manufacturing. Its focus on manufacturing has come at the expense of domestic consumption and services. As China’s economy changes to emphasize consumption, Kenya may take advantage of the opportunity to export financial, tourism, and business services to China. For instance, Kenya has the opportunity to export the MPESA system to China and other countries, especially those with poorly developed financial services. Exporting more services to China and to other countries will help upgrade the services industry and strengthen the overall balance of trade. 2.1.1 Kenya should pay more attention to the overall trade balance When talking about trade, many officials in both China and Kenya are primarily concerned with the bilateral trade deficit, but it is a meaningless statistic; a country can have simultaneous sur- pluses and deficits with many different trading partners and still have a positive balance overall. For policy makers, the overall trade deficit in Kenya is more relevant and a bigger reason for concern. A brief overview of Kenya’s weak exports The current account deficit, or imports minus ex- ports, reached 10.4 percent of GDP in 2014. The deficit was badly hit in 2011 when high oil and food prices and a weak shilling increased Kenya’s import bill so much that Kenya’s top four ex- ports were insufficient to cover the cost. To finance the gap, Kenya had to rely on both short and long-term debt. Even with lower oil prices, the deficit remains high at 9.8 percent of GDP because imports of capital and equipment increased more than 25 percent. But as imports soar, exports continue to dip. In 2015, Kenya’s manufactured exports fell 20.3 percent, its horticulture exports declined 5.5 percent, and its chemical exports fell 7.9 percent (Kenya Economic Update 2015). Even one of the major earners, tea, fell 1.1 percent. Tea and coffee still account for most of the growth, and Kenya must improve the competitiveness of manufacturing to diversify exports. It must also diversify export markets because the majority of growth is in traditional destinations, neglecting new opportunities for expansion. Why are Kenya’s exports performing so terribly? One can trace Kenya’s weak exports to an underperforming manufacturing sector. For over a decade, manufacturing has remained at only 10 percent of GDP. Manufacturing receives little investment because investors want to avoid the underdeveloped infrastructure and high cost of doing business, and have diverted funds to non- tradable sectors such as real estate and construction. The budding tradable sector has watched its 3 A. Sanghi and D. Johnson competitiveness erode through poor government policies and inefficiency. Price controls and mis- managed marketing boards, for example, have discouraged coffee farmers from exporting, and the sector is barely recovering from the damage (Kenya Economic Update 2010). The goverment must shift resources to production of tradable goods or risk getting into more debt, debt that will lead to slower future growth. Figure 1 shows Kenya’s overall trade balance between 2000 and 2014. Kenya’s net exports have fallen 14.74 percent per year over the period, reaching a low of negative US $12.2 billion in 2014. The trade balance reflects a larger need of preparing exports for competitive markets. Rather than focusing on exports to the Chinese market, Kenya should seek global markets and improve export competitiveness: curbing inflation and real exchange rate appreciation, reducing high tar- iffs on manufacturing inputs, and attracting more FDI into manufacturing. A key component to acheiving export competitiveness is the port of Mombasa. Greater efficiency at the port will cut the time for goods to reach Nairobi and help Kenya’s regional exports. To improve the export cli- mate, Farole and Mukim (2013) recommend enforcing competition law especially in the transport sector, creating an automated risk management system to speed up risk-free cargo through cus- toms, and creating a trade information portal on general tariff rates, preferential rates, and quality standards. 0 Trade Balance (USD Millions) −2500 −5000 −7500 −10000 −12500 Dec.00 Dec.01 Dec.02 Dec.03 Dec.04 Dec.05 Dec.06 Dec.07 Dec.08 Dec.09 Dec.10 Dec.11 Dec.12 Dec.13 Nov.14 Year Figure 1: Kenya’s overall trade balance is a bigger concern Source: Central Bank of Kenya 2015 2.2 China is a large source of imports for Kenya Figure 2 shows Kenya’s imports from China between 2012 and 2014. China’s share of Kenya’s total imports has increased significantly. In 2012, Kenya’s imports from China were 12 percent of total imports, but by 2014, they rose to 23 percent. Kenyan consumers benefit thanks to a larger quantity of cheap Chinese products in the market. From 2012 to 2014, consumers enjoyed a ten percent lower unit price on manufactured goods and a seven percent lower unit price on chemicals. Consumers are gaining, but policy makers fear 4 2 KENYA AND CHINA’S TRADE RELATIONSHIP that local producers are suffering from cheap Chinese goods. Some even argue that imports are hurting Kenya’s prospects of industrialization. Imports (USD millions) 20000 China, P.R.: Mainland World 10000 0 2012 2013 2014 Year Figure 2: China takes a sizeable share of total Kenyan imports (2012-2014) Source: IMF World Economic Outlook Database 2014 3000 3000 Imports (USD Millions) Imports (USD Millions) China, P.R.: Mainland China, P.R.: Mainland 2000 2000 India India Japan Japan South Africa South Africa 1000 1000 United Arab Emirates United Arab Emirates 0 2004 2006 2008 2010 2012 Year Country (a) China and India are a major source of imports (b) China and India were top two sources of imports (2013) (2004-2013) Figure 3: China and India are major sources of imports (2004-2013) Figure 3a shows trends in Kenya’s top import partners between 2004 and 2013. With strong growth since 2004, China and India have become major sources of imports to Kenya. China ranks second to India in number of exports to Kenya, and its low production costs and better positioning in global value chains may help it become the top source of imports for Kenya (figure 3b). Chinese imports grew at an annual rate of 33 percent, and Indian imports also grew quickly at 30 percent per year. Kenya likely imports even more from China and India because many imports from the UAE are re-exported manufactured products such as phones, computer monitors, or jewelry originally from China or India. It is is difficult to find the exact amount of re-exports from China, but the UAE re-exported US $384.5 million worth of goods to Kenya in 2014, a large fraction of those 5 A. Sanghi and D. Johnson goods originating from China (UN Comtrade 2015).2 2.2.1 Kenya imports rubber manufactures from China As in other Sub-Saharan African countries, Kenya mainly imports manufactured products from China. Figure 4 shows the top four import categories from China: manufactured goods classi- fied chiefly by material made up 35 percent, machinery and transport equipment were 31 percent, miscellaneous manufactured articles were 24 percent, and chemical and related products were 8 percent of total imports from China. In 2012, the top goods from China were rubber products, footwear with outer soles of rubber or plastic and woven fabrics of synthetic filament (5).3 5000 4000 Imports (USD millions) 3000 Chemicals and related products Machinery and transport equipment Manufactured goods. Ex: leather, rubber, fabrics 2000 Miscellaneous manufactured articles. Ex: apparel, footwear, furniture 1000 0 Commodity Figure 4: Kenya imports rubber manufactures from China (2010-2014) Source: UN Comtrade database 2015 Kenya imports a large amount of rubber products because China has a comparative advantage in cheap manufactured goods; its manufacturing sector and infrastructure is also much more inte- grated in global value chains. The special economic zones for manufacturing in Kenya have poor infrastructure and a weak rule of law, inhibiting its sales and growth. The barriers to production will prevent local producers from competing with Chinese goods. 2 UAE exports almost 60 percent of products originally from China according to the United Arab Emirates Informa- tion Guide http://alluae.ae/uae-imports-exports-re-exports/ 3 We use the 6 digit Harmonized System (HS) classification from the United Nations 6 2 KENYA AND CHINA’S TRADE RELATIONSHIP 200 Import Value (USD Millions) Mineral, petroleum oils 150 Plastics and polyethers Rubber products, pneumatic tyres for buses and lorries Leather, saddlery, handbags and suitcases Cotton and woven fabrics of cotton 100 Woven fabrics of synthetic filament Rubber of plastic footwear Ceramic products Television, image and sound recorders, telephone sets 50 Motorcycles and mopeds 0 Commodity Figure 5: Kenya imports rubber footwear, tires, and fabrics from China (2012-2014) Source: UN Comtrade database 2015 2.3 China is still a small export market for Kenya As shown in Figure 6, Kenya only sends one percent of its exports to China, exporting US $63 million in 2012, US $48 in 2013, and US $70 million in 2014. Kenya exports little to China be- cause it is an oil importer and relatively resource-scarce. With fewer natural resources, Kenya has been unable to take advantage of the commodity boom from China’s growth (Zafar 2007). What’s more, the growth does nothing for Kenya’s agricultural sector because it lacks a comparative ad- vantage in China’s main food imports (wheat, corn, beef, soybeans), making it difficult for Kenya to increase its exports of agricultural products. 6000 Exports (USD millions) 4000 China, P.R.: Mainland World 2000 0 2012 2013 2014 Year Figure 6: Kenyan exports to China small relative to total exports (2012-2014) Source: IMF World Economic Outlook Database 2014 7 A. Sanghi and D. Johnson 2.3.1 Kenya exports raw goods and metals to China Crude materials such as raw hides and skins are the major exports to China (figure 7). Between 2010 and 2014, exports of crude materials were 55 percent, manufactured goods were 21 percent, food and live animals made up 15 percent, and chemicals and related products were 9 percent of exports to China. In the crude materials category, major exports to China include hides and skins, scrap metals, and sisal; coffee and tea were major exports in the food and live animals category.4 In general, tea, coffee, sugar and flowers are sources of major foreign exchange earnings for Kenya, but it has managed some diversity in earnings (UNCTAD 2013). Value-added products such as chemicals and plastics have also reached China, a shift from the more common story of oil and resource exports. 150 Export Value (USD millions) 100 Chemicals and related products Crude materials. Ex: raw hides and skins, metal ores Food and live animals Manufactured goods. Ex: plastics, vegetable textile fibres 50 0 Commodity Figure 7: Kenya exports metals and hides and skins to China (2010-2014) Source: UN Comtrade database 2015 In figure 8, we show the top products in the crude materials and manufactured goods cate- gories. For crude materials, titanium ores and concentrates exports were the highest, followed by copper, vegetable textile fibers such as cotton, hemp, or sisal, and plastics.5 Kenya’s exports feature minimal value addition, and the prices of Kenya’s major exports to China are low on in- ternational markets. Between January 2010 and January 2015, copper prices dropped 4.62 percent annually, and iron ore prices declined 11.57 percent per year (figure 9). A further 11 percent de- cline in the price of metals was forecasted by the World Bank Commodities Outlook (2015) because of weak import demand from China and new supplies of metal globally. Titanium and iron ores and copper are important exports to China, and with falling metals prices, Kenya will likely gain little from its current export pattern to China. 4 Here we use the Standard International Trade Classification (SITC) Rev.4 from the UN Statistics Division 56 digit HS classification 8 2 KENYA AND CHINA’S TRADE RELATIONSHIP Export Value (USD Millions) Fish and crustaceans 20 Coffee, tea, and spices Iron ore and concentrates Titanium ores and concentrates Plastics Tanned or crust hides and skins of bovine Tanned or crust skins of sheep or lambs 10 Raw hides of other animals Vegetable textile fibres Copper 0 Commodity Figure 8: Kenya exports titanium ores and copper to China (2012-2014) Source: UN Comtrade database 2015 10000 7500 copper 5000 USD per metric ton 2500 200 150 iron.ore 100 50 1990 1995 2000 2005 2010 2015 Year Figure 9: Metals prices have been falling since January 2010 (1990-2015) Source: World Bank Commodity Price Data (The Pink Sheet) Finished leather has a window of opportunity in the Chinese market Because of economies of scale and low input costs, China is a much more competitive producer of leather products. Kenya, however, has room to export finished leather products to China. China has a limited presence in the high end leather market, giving Kenya a chance to supply more value-added leather products. (Hansen, Moon, Mogollon 2015). At the moment, Kenya exports raw hides and semi-processed leather, products that are in high demand in China. Just as Ethiopia attracted Chinese investment in the leather sector to meet demand, so may Kenya work to bring more investment to footwear manufacturing, lifting exports. Domestic leather demand is also higher than supply; local producers may use the new technology from Chinese firms to meet the growing domestic demand for finished leather (Hansen, Moon, Mogollon 2015). 9 A. Sanghi and D. Johnson 2.4 Winners and Losers from Kenya’s current trade patterns 2.4.1 The net benefit to the economy is positive Cheaper Chinese finished and intermediate goods provide an overall benefit to the economy. Chi- nese competition forces uncompetitive firms out of the market and eliminates the deadweight loss in the domestic economy. Firms that remain are able to improve efficiency and upgrade standards to supply inputs to Chinese companies. During the Standard Gauge Railway construction, local cement producers upgraded their production to meet international standards and supply part of the railway. Firms that use Chinese goods as intermediate inputs improve their efficiency, and informal sector firms that use intermediates increase their retail margins and create more employ- ment. Some well-organized producers may lose out, but economic theory tells us that the overall gain is positive. 2.4.2 Consumers and retailers gain Consumers benefit from a greater variety of cheap consumer electronics and plastic and rubber products. Medicine, footwear, clothing, textiles and office supplies are now available to consumers at much lower prices, prompting side businesses reselling consumer products to enter the market. Between 2013 and 2014, the wholesale and retail trade sector grew at 6.9 percent, and shopkeepers in western Kenya had an average annualized return of 33 percent, although the median firm in the study’s sample had an annualized return close to zero (Kremer et al 2011). Chinese goods only seem to help small retailers. Feinberg (2010) finds that small retailers in the United States are generally unaffected by currency appreciation, an indicator of higher imports; small retailers can cope with greater imports and other economic shocks. Although the context is different, the study agrees with anecdotal evidence of more Kenyan retailers selling goods from China. Chinese goods help small kiosks and shops earn greater profits, and since small shops make up 70 percent of shopping, Chinese goods appear to have benefitted retailers on a large scale. China dominates the second-hand clothing and shoe market in Kenya Most consumers buy leather shoes and clothes from the mitumba, or second-hand markets. Mitumbas offer consumers quality brands from North America, Europe, and Asia at lower prices than local clothes. But China is still a major player in the mitumbas: 60 percent of second hand shoes come from China and Hong Kong, and 60 percent of global leather footwear production and 40 percent of world exports of leather footwear are also from China and Hong Kong (Hansen, Moon, Mogollon 2015). The prices of Chinese products are often significantly cheaper. A bale from China is half the price of a bale from Germany, but the shoes are of different qualities. 80 percent of shoes in a bale from Europe are leather, but only ten percent of shoes in a Chinese bale are leather, the rest being rubber or plastic. The rubber and plastic shoes sell better in urban mitumbas; leather shoes sell better in rural areas because people often walk long distances and need something durable (Hansen, Moon, Mogollon 2015). 2.4.3 Producers are worse off; some benefit from Chinese intermediate goods The literature generally concludes that existing and potential local producers in Sub-Saharan Africa are displaced by Chinese imports. Chinese imports have hurt textile and clothing pro- 10 2 KENYA AND CHINA’S TRADE RELATIONSHIP duction, a sector that is 20 percent of all formal manufacturing employment in Kenya. Clothing enterprises employ mainly women, so a weaker industry worsens gender equality. Some producers benefit from using Chinese intermediate goods in production When local producers use intermediate goods, they can access goods unavailable locally to increase their pro- ductivity. Between 1990 and 2014, imports of intermediate and capital goods have grown 12.6 percent annually (figure 10). Over time, many domestic businesses have switched to cheaper Chi- nese goods. Figure 11 shows imports from top trading partners: since 1990, Kenya’s imports from China have increased 26.7 percent per year, and are now 21.9 percent of imports, replacing goods from other major sources. In 2014, only 6.9 percent of imports came from India and 4.5 percent came from South Africa, two former major sources of intermediate inputs.6 Intermediate Goods imports (USD Millions) 10000 8000 6000 4000 2000 0 1990 1995 2000 2005 2010 2014 Year Figure 10: Kenya’s imports of intermediate goods is rising quickly Source: UN Comtrade 2015 The top categories from China were processed industrial supplies, parts and accessories for transport equipment, capital goods, primary industrial supplies, and food and beverages for in- dustry. Industry supplies were 77 percent, and transport equipment and capital goods were each ten percent of imports of intermediate goods. Imports of industrial supplies and transport equip- ment are high because of strong demand from the Standard Gauge Railway construction. In gen- eral, transport equipment contributes a large share to import growth, and the switch to Chinese imports cuts the cost of production. 6 We use the Broad Economic Categories (BEC) classification for intermediate goods 11 A. Sanghi and D. Johnson China India Japan Percentage of total intermediate goods imports 80 60 40 20 0 South Africa All countries All −−− All United States 80 60 40 20 0 1990 1995 2000 2005 2010 2014 1990 1995 2000 2005 2010 2014 1990 1995 2000 2005 2010 2014 Year Figure 11: China’s share of Kenya’s intermediate goods imports is growing Source: UN Comtrade 2015 Kenyan apparel exports to the US fell initially but are now improving Under the African Growth and Opportunity Act (AGOA) and the MultiFibre Agreement (MFA), the US received over over 90 percent of Kenya’s clothing exports; exports increased 292 percent from US $78 million to US $306 million (Kaplinsky 2008). After removal of the MFA, Chinese competition decreased Kenyan exports, and threatened to erode gains Kenya made in the US market. Between 2004 and 2006, the value of Kenyan clothing exports to the US dropped 5.1 percent after the first two years of quota removal and dropped 4.6 percent between 2005 and 2011 (Kaplinsky 2008; Onjala 2008). Five factories closed in 2004-2005 with 4,603 job losses —though much fewer than the predicted 25,000 job losses. After the renewal of AGOA in 2008, Kenyan apparel exports rebounded and are steadily grow- ing (figure 12). Between 2010 and 2014, exports grew at a rate of 13.4 percent per year, faster than the Sub-Saharan Africa growth rate of 5.3 percent. By 2014, Kenya was 37.1 percent of Sub-Saharan Africa’s apparel exports to the US, up from 29.4 percent in 2012. But even with the help of AGOA, Kenya’s exports are still much smaller than China’s. In 2014, Chinese exports were 81 times as much as Kenya’s. Chinese exports grew at a slower rate of 1.2 percent from US $28.8 billion in 2010 to US $30.5 billion in 2014, but they are still 36.7 percent of total US apparel imports. China is becoming a supplier of choice for most US importers because it can meet the large volumes required by large US retailers and apparel companies. China also has lower labor costs, better technology, and better connections in global value chains. If China can manage its market share even with AGOA, if AGOA is removed, Kenyan exports may experience a major loss of their US market share. 12 2 KENYA AND CHINA’S TRADE RELATIONSHIP Log Exports to US (USD Millions) 10 Exports to US (USD Millions) 300 World 8 China Kenya 200 Kenya 6 SSA 100 4 2000 2005 2010 2014 2000 2005 2010 2014 Year Year (a) China’s apparel exports strong even with AGOA (b) Kenya’s apparel exports hit hard in 2004, but rebounded from 2010 Figure 12: Kenya’s apparel exports to the US (2000-2014) Source: UN Comtrade 2015 Kenya’s exports to Uganda and Tanzania are falling Chinese goods may have also hurt Kenya’s exports to its neighbors. Exports to Tanzania and Uganda are quite similar to China’s, compared to both countries’ exports to the United States or the UK. The greater overlap in East Africa suggests that Chinese goods will likely displace Kenyan exports. Between 2008 and 2014, manufacturing exports to Tanzania fell 36.1 percent; exports to Uganda increased slightly by 4.5 percent, but com- pared to previous years, the growth was slow. 7 Some fear that Chinese imports could lead to deindustrialization Because Kenya produces and trades few intermediate goods, researchers have concluded that Chinese imports could lead to a de-industrialization. Many suspect a premature decline of industry because manufacturing growth was only 3.4 percent in 2014, down from 5.6 percent in 2013 (Chen et al 2015). The man- ufacturing sector is ten percent of GDP, but the government wants it to be 20 percent of GDP as part of its Vision 2030 program (KNBS 2015). Kenya will need to promote more FDI into man- ufacturing, improve labor productivity and infrastructure, lower transport costs, and lighten the regulatory burden of trade it if hopes to boost exports and the share of manufacturing in GDP (Farole 2011). But Ethiopia is building a strong manufacturing industry Ethiopia does a much better job of attracting FDI in manufacturing. Between 2010 and 2013, the FDI to GDP ratio in Ethiopia was 1.39 compared to only 0.67 in Kenya, and the FDI to Export ratio was 0.1 in Ethiopia and 0.03 in Kenya; it also has a lower cost of doing business, offering lower taxes, electricity, and labor costs 7 We calculate the export similarity of Kenya and China in Tanzania to be .0247 and in Uganda to be .164; in contrast, the export similarity in the United States is only 0.06. Let Xij be the share of product i in country j’s exports and Xik be the share of product i in country k’s exports. The export similarity is n XS jk = ∑ min{Xij , Xik } i =1 13 A. Sanghi and D. Johnson than Kenya (Chen, Geiger, Fui 2015). Ethiopia’s manufacturing received 76 percent of total FDI for projects in operation, and China is the second largest investor in manufacturing, investing US $545 million between 2008 and 2013; it also has 196 projects in operation, the greatest number of projects among all investors (Chen, Geiger, Fui 2015). Within manufacturing, Ethiopia’s textiles, clothing, and leather sub-sector attracted US $2.5 billion in investment. FDI in the textiles and clothing sub-sector is crucial for employment growth. In Ethiopia, 40,000 permanent jobs came from manufacturing FDI between 2008 and 2014, with 22,000 manufacturing jobs originating from the textiles, clothing, leather and shoemaking sub- sectors (Chen, Geiger, Fui 2015). But many jobs in textiles and clothing come from two major investors, China and India. Between 2008 and 2014, China created 24 percent and India created six percent of total jobs. If Kenya wishes to attract more manufacturing FDI from China and India, it will have to work towards lowering the cost of doing business to compete with Ethiopia’s more favorable investment climate (Chen, Geiger, Fui 2015). Kenya has preferential access to the US market under AGOA, and may attract more manufacturing FDI in textiles to create permanent jobs and boost exports to the US market. 2.5 What drives Kenya’s exports? 2.5.1 Estimation We estimate a gravity equation to explain what influences Kenya’s exports, and if Kenya under or over exports to China relative to comparable countries. The model is analogous to the physical equation of gravity; just as the gravity is proportional to the mass of two planets and inversely proportional to the distance between them, so are exports positively related to the sizes of the two economies and negatively related to the distance between the two economies. The detailed model is in appendix A.1. Using a full set of bilateral trade flows between all pairs of countries as described in appendix B, we estimate a Poisson Quasi Maximum Likelihood regression model described in appendix A. Kenya exports less to China than to economies of similar size Under the Poisson model in table 1, the coefficients on importer GDP per capita and distance are significant and have the expected signs: distance reduces trade and a higher GDP per capita of the importer increases trade. Distance deflects Kenya’s exports by a factor of 0.18. Kenya’s exports increase by a factor of 10.02 to countries that share a common border with Kenya. The China dummy coefficient is -1.74, so Kenya exports less to China by a factor of 0.18 compared to economies of similar size. If Kenya exports one million USD worth of goods to a country similar to China, it only exports 820,000 USD worth of goods to China. Kenya may export less to China after controlling for other independent variables, but the effect is constant over time —none of the China-year dummy coefficients are significant (table C8). Table 1: Estimates of Kenya’s Trade Flows (1948-2014). Dependent variable: Kenya’s exports Poisson Quasi Maximum Likelihood Constant −0.93 (0.41)∗ ∗∗∗ Kenya GDP per capita 0.00 (0.00) 14 2 KENYA AND CHINA’S TRADE RELATIONSHIP Poisson Quasi Maximum Likelihood Importer GDP per capita 0.00 (0.00)∗∗∗ Population Kenya −0.01 (0.01) ∗∗∗ Population Importer 0.00 (0.00) Distance −0.00 (0.00)∗∗∗ Regional Trade Agreement 1.16 (0.10)∗∗∗ ∗∗∗ Contiguous 2.30 (0.09) ∗∗ Common Language 0.19 (0.06) GATT Kenya 0.78 (0.23)∗∗∗ ∗∗∗ GATT importer 1.50 (0.09) Colonial History 2.53 (0.08)∗∗∗ China dummy −1.74 (0.42)∗∗∗ AIC BIC Log Likelihood Deviance 98819.06 Num. obs. 6118 ∗∗∗ p < 0.001, ∗∗ p < 0.01, ∗ p < 0.05. Standard errors in parentheses. Kenya’s exports falls with distance. China is far Figure C36 presents the relationship between distance and trade. The farther Kenya is from its destination, the less it exports. Kenya may under export to China because they are 9,201.3 kilometers apart, nearly in the third quartile of distances in the sample. Kenya exports more to richer countries Figure C37 shows the relationship between the per capita GDP of the importing country and exports in 2014. We see that a higher importer GDP per capita is linked to higher trade. Given its already high per capita GDP, the UK stands out as a bigger trading partner, and it probably trades more with Kenya because of the former colonial link. Kenya’s exports to China are still low as figure 6 shows, but the large per capita GDP of China presents a potential market for exports. China’s exports to Kenya are “normal” We estimate a similar Poisson regression taking China as an exporter to the world. Table 2 presents estimates of China’s exports from the Poisson model: the importer GDP per capita and distance are significant predictors of trade; the coefficient of distance is close to zero, meaning that a one kilometer increase in distance reduces China’s exports by a factor close to 1. The effect of sharing a border with its trading parter is insignificant for China, but it benefits Kenya. China exports more to Kenya by a factor of 1.08, so if China exports goods worth one million USD to a country comparable to Kenya, it exports 1,080,000 to Kenya. The coefficient, however, is statistically insignificant. Since the results from Poisson are less biased than the ordinary least squares model, we observe that China’s exports to Kenya are as expected when we hold the other explanatory variables fixed. Over time, China’s exports to Kenya are 15 A. Sanghi and D. Johnson roughly the same as its exports to countries similar to Kenya; neither the year dummies nor the Kenya year dummies are significant (Table C7). Table 2: Estimates of China’s Trade Flows (1948-2014). Dependent variable: China’s exports Poisson Quasi Maximum Likelihood Constant 8.82 (15.22) China GDP per capita 0.00 (0.00) ∗∗∗ Importer GDP per capita 0.00 (0.00) Population China −0.01 (0.03) ∗∗∗ Population Importer 0.00 (0.00) ∗∗∗ Distance −0.00 (0.00) Contiguous 0.09 (0.06) ∗∗∗ Common Language 1.94 (0.06) GATT China 0.90 (18.60) GATT importer 0.79 (0.08)∗∗∗ Colonial History −1.45 (0.91) Kenya dummy 4.42 (32.00) AIC BIC Log Likelihood Deviance 13045625.79 Num. obs. 8215 ∗∗∗ p < 0.001, ∗∗ p < 0.01, ∗ p < 0.05. Standard errors in parentheses. Gravity model suggests room for Kenyan export growth The results suggest that Chinese ex- ports to Kenya are essentially normal, but Kenya under exports to China compared to other larger markets. Since China has specialized in manufacturing exports, it is simply a large exporter in global value chains; Kenya imports the same amount from China as countries at similar income levels do. Kenya, however, has some room for growth of exports. Leather products and cut- flowers have shown some promise, and Kenya could reap the benefits of reorganizing its leather sector and negotiating for duty free access of its cut-flowers in China. 2.6 Brighter prospects for services exports as China rebalances Kenya is unlikely to expand exports to China at the moment, but it may look forward to real ben- efits towards 2030. A recent World Bank Africa’s Pulse report (2015) shows that China will import more from Kenya once it rebalances its economy to a consumption-driven path. Focusing on do- mestic consumption, China’s GDP growth will slow from 7 percent to 6 percent per year between 2016 and 2030, eventually reaching 4.6 percent per year in 2030. Initially, Kenya’s GDP will fall by 1.3 percent compared to if China continued with seven percent yearly growth. Kenya’s GDP will suffer from weak demand for commodity exports, lower world commodity and food prices, and 16 2 KENYA AND CHINA’S TRADE RELATIONSHIP higher spread on its sovereign bonds, the spread being already 600 basis points (6 percent) over US Treasuries (World Bank Africa’s Pulse 2015). Kenya’s GDP will rebound, however, when China decreases investment and allows consump- tion to take a larger share of its GDP. Greater consumption usually leads to more imports, and demand for tourism, travel and business services will increase relative to commodities, an in- crease that should help Kenya’s service sector, an already strong sector (figure 13). From the boost from China’s greater consumption, Kenya’s GDP will 7.5 percent higher than it would have been if China stayed with seven percent growth. By 2030, the higher GDP should also bring higher wages and real exchange rate appreciation (World Bank 2015). 5000 4000 Export Value (USD millions) 3000 2000 1000 0 2009 2010 2011 2012 2013 Year Figure 13: Kenya’s services exports overall are strong Source: UN Service Trade 2015 Threats to growth scenario: Strong Chinese downturn and terrorism Kenya may not realize the benefits it China’s economy slows too much. A sharper downturn than expected will push down commodity prices even further, raising the cost of financing for Kenya; Kenya has a large current account deficit of ten percent of GDP (World Bank Africa’s Pulse 2015). Slower Chinese growth may also mean less FDI in Kenya; Kenya performs poorly in attracting FDI and inflows are volatile (see sections 3 and 4.1.1). Less FDI in Kenya may stall growth plans, plans that would upgrade much need infrastructure. Because China’s domestic consumption will shift toward final consumer goods, demand for natural resources and commodities will fall, and Kenya’s recently discovered oil will face lower global prices. 2.6.1 China’s rebalancing will help reduce poverty by 2030 Lakatos et al (2015) use a Global Income Distribution Dynamics (GIDD) model that incorporates micro and macro data for simulation of the effects of economy-wide changes in Sub-Saharan Africa. Changes in demographics, sectoral employment, per capita consumption growth, relative wages across sectors, and relative food to non-food prices are recorded by the model. It also uses 17 A. Sanghi and D. Johnson data from 130 countries and explicitly assesses the long-term behavior of income distributions, tracking demographic and educational changes over time. Wealthier households in Sub-Saharan Africa benefit from demographic changes and wage changes across sectors The wealthiest 40 percent of households will have higher per capita in- come growth from demographic changes, but the poorest 40 percent of households will see no benefit. The poorest 20 percent of households will experience slower per capita income growth when demographic changes cause changes in relative wages across sectors. Upper middle income households —those between the 60th and 80th percentile of the income distribution —will gain the most from the wage changes, earning the fastest per capita income growth8 Poorer households are hurt by changes in food to non-food relative prices, but gain overall from Chinese slowdown and rebalancing Changes in food to non-food prices leave the poorest 40 percent of households worse off: per capita income growth is 2.9 percent compared to the 3.07 percent if China had continued to grow at seven percent. But thanks to greater Chinse consump- tion, the bottom 40 percent will increase their incomes; the number of people living in extreme poverty will fall by an additional 4.04 million people. The Chinese slowdown scenario increases poverty initially, but the rebalancing reduces poverty enough for an overall drop. More important, the bottom 40 percent in Kenya will see a per capita income increase of 2.7 percent, the highest in Sub-Saharan Africa (Lakatos et al 2015). 3 Foreign Direct Investment in Kenya Kenya performs poorly in attracting foreign direct investment (FDI) given the size of its economy. Despite a larger economy, Kenya attracts even less FDI than Uganda and Tanzania. Figure 14 shows Kenya’s net inflows of foreign direct investment (FDI) from 1980 to 2014. Kenya’s invest- ment levels dropped to less than 10 percent of GDP near 2000, but has since returned to levels experienced in the mid 1990s. Corruption, poor infrastructure, and poor investment climate have reduced foreign direct investment flows compared to pre 1980 levels. In 2007, Kenya received US $729 million in FDI, but post-election violence in 2008 cut down flows to only US $96 mil- lion. Kenya has since recovered, but it only recently surpassed its former peak. Figure 15 shows Kenya’s FDI as a percentage of GDP. The average FDI to GDP ratio between 1980 and 2014 was 0.54. When compared with Kenya’s domestic savings rates, Kenya’s FDI signals a vulnerable cur- rent account because of weak domestic savings and investment. Figure 16 shows Kenya’s gross domestic savings as a percent of GDP. From a peak of 22 in 1994, Kenya’s gross domestic savings rate dropped to only 4 percent in 2014. Kenya’s savings rate is much lower than the average in Sub-Saharan Africa of 20.4 percent. One reason for the low savings is Kenya’s large scale infras- tructure projects: the Standard Gauge Railway, Lamu berths, and Northern Corridor Integration Projects. Kenya’s high fiscal debt puts it at a sovereign risk level of B1 (Moody’s 2015; KPMG 2013).9 8 9 AAA is least risky and D is most risky. 18 3 FOREIGN DIRECT INVESTMENT IN KENYA 750 FDI (USD Millions) 500 250 0 1980 1985 1990 1995 2000 2005 2010 2014 Year Figure 14: Kenya’s FDI is low. Trend in red (1980-2014) Source: World Development Indictors World Bank 2015 2.5 2.0 FDI (% of GDP) 1.5 1.0 0.5 0.0 1980 1985 1990 1995 2000 2005 2010 2014 Year Figure 15: Kenya underperforms in attracting FDI relative to potential. Trend in red (1980-2014) Source: World Development Indictors World Bank 2015 19 A. Sanghi and D. Johnson 20 Savings (% of GDP) 15 10 5 1980 1985 1990 1995 2000 2005 2010 2014 Year Figure 16: Kenya’s Gross domestic savings has fallen sharply since 1993. Trend in red (1980-2014) Source: World Development Indictors World Bank 2015 3.1 Chinese FDI in Sub-Saharan Africa China’s FDI consists of many medium to short-term loans with a focus on extractive industries such as oil, mining, gas, minerals and natural resources (Wang and Bio-Tchane 2007). Some loans are repaid with future exports from natural resources, especially in countries with poor credit ratings. Such resource-backed loans can be extended to Ghana, for example, for a hydropow- ered dam, and the loan is repaid with exports of cocoa beans (Br¨autigam, Gallagher 2013). Other loans go to support multinational or state-owned companies in accessing markets, or to support foreign firms in buying Chinese goods. In 2004, the Chinese Development Bank loaned the ma- jor telecommunications firm Huawei US $10 billion for overseas expansion, and Nigeria took US $200 million in loans to buy Huawei equipment (Executive Research Associates 2009). In 2012, Huawei was awarded a tender to build a national fibre-optic network in Kenya worth US $60.1 million, a deal financed the China EXIM Bank (ICT Authority Kenya 2015). The China EXIM Bank also gives cheap capital to state-owned firms to bid for large infrastructure projects. Their state- owned status allows them to report profits at longer intervals, instead of quarterly as most firms are required. Other foreign firms with shorter time horizons and a higher profit requirement face a unique challenge when competing for contracts in Sub-Saharan Africa. 3.1.1 Chinese FDI in Sub-Sharan Africa is relatively small In 2011, Chinese FDI stock in Sub-Saharan Africa was US $20.1 billion, or 3.2 percent of the to- tal FDI stock of US $629 billion in Africa. China’s relative focus on Sub-Saharan Africa is large —US $26 billion in Sub-Saharan Africa compared to US $22 billion in the United States in 2013 —but its share of investment is still small (Chen, Dollar, and Tang 2015). Unlike in oil-exporting countries, Chinese firms are interested in coffee and manufacturing in Kenya because the returns are higher than in oil exploration. In particular, the communications and automotive equipment manufacturing sectors (OEM) attract a large share of Chinese investment (figure 23a) (Financial Times Ltd 2015). Figure 17 shows the investment flows from China between 2003 and 2012. Its 20 3 FOREIGN DIRECT INVESTMENT IN KENYA FDI has grown rapidly, reaching US $23 million in 2008 from just US $1 million in 2003. Between 2009 and 2010, Chinese FDI increased 261 percent, the largest gain during the 2003-2012 period before returning to US $68 million in 2011. China’s FDI stock in Kenya has also grown 35.6 percent annually from US $26 million in 2003 to US $403 million in 2012. Figure 20 shows Kenya’s top FDI inflow sources in 2011 (UNCTAD 2015). China comes in second to the United Kingdom, and its share of Kenya’s FDI total inflows has also grown. Reaching a peak of US $101 million in 2010, or 57 percent of Kenya’s total FDI flows, it then fell to US $79 million in 2012, representing 31 percent of total flows (figure 18). China’s large FDI profile will lead to greater cooperation with Kenya, but Kenya must continue to lower the costs of business and investment and curb corruption to squeeze the most from foreign investment. 100 400 FDI Instock (USD Millions) FDI Inflow (USD Millions) 75 300 50 200 25 100 0 2003 2006 2009 2012 2003 2006 2009 2012 Year Year (a) Kenya’s FDI inflow from China (2003-2012) (b) Kenya’s FDI instock from China (2003-2012) Figure 17: Chinese FDI in Kenya is growing quickly since 2009 (2003-2012) Source: UNCTAD FDI/TNC database 2015 China FDI (% of total FDI) 40 20 0 2003 2006 2009 2012 Year Figure 18: China’s FDI represents large share of total FDI (2003-2012) Source: Authors’ own calculation based on World Development Indictors World Bank 2015 21 A. Sanghi and D. Johnson 800 Investment (USD Millions) 600 China India 400 Nertherlands UK 200 US 0 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Year Figure 19: Investment from China rising; investment from UK and US falling Source: KenInvest 2015 80 FDI inflows (USD Millions) 60 40 20 0 China France Italy Japan Luxembourg Switzerland Country Figure 20: China and France top sources of FDI inflows for Kenya (2012) Source: UNCTAD FDI/TNC database 2015 22 3 FOREIGN DIRECT INVESTMENT IN KENYA 800 FDI instock (USD Millions) 600 400 200 0 China France South Africa Switzerland United Kingdom United States Country Figure 21: China and UK hold the most FDI stock in Kenya (2012) Source: UNCTAD FDI/TNC database 2015 Chinese firms are well-represented in large-scale infrastructure projects China constructed 905.5 km of road in 2006 and invested AC200 million (US $227.6 million) to rehabilitate the Nairobi- Mombasa road (Fiott 2006). In 2012, three kilometeters of road at the KWS Gate/Bomas Junction have been built for KES 2.67 billion (US $27.1 million) by the China Wu Yi Co., and many Chinese firms have been pre-qualified by the government to develop 2,000 km of road in various counties (Kenya Urban Roads Authority 2015). For the mega-projects, the China Road and Bridge Cor- poration is building a 609 km section of the Standard Gauge Railway (SGR) linking Nairobi and Mombasa for US $3.6 billion, and the China Communication Construction company is building three berths at the port on Lamu island for US $467 million (Financial Times 2014). With the slow- down of the Chinese economy, China’s low cost infrastructure providers will seek to market their services outside of China, including in Africa, so Chinese infrastructure projects in Kenya will likely continue. But the bulk of Chinese work in infrastructure is not investment, but contracting The SGR and the berths at Lamu are not investments, but projects financed by China’s EXIM bank: the bank will finance 90 percent of the SGR, making SGR a Kenyan investment. Kenya has taken on debt for many major construction projects, and Chinese companies typically have no equity stake in the particular building, road, or port. The EXIM bank is responsible for promoting exports and investments of Chinese firms abroad, offering international guarantees, export buyers credit, and export sellers credit (Center for Global Development 2006). The bank meets demand for in- frastructure in Kenya and Sub-Saharan Africa, a sector that traditional donors and investors have neglected because of fears of corruption and high risk. The EXIM bank, for example, provided US $95.4 million for drilling of 26 wells in Olkaria, Nakuru for geothermal energy in 2014 (Geothermal development company 2014). Once one accounts for large, debt-financed projects, medium sized private Chinese firms are a sizeable share of investment, and they tend to operate in the manufac- turing sectors, creating many low-skilled manufacturing jobs. Futhermore, Chinese firms face the same problems that any firm faces in Kenya: poor electricity, roads, security and corruption. 23 A. Sanghi and D. Johnson 3.2 Chinese companies in Kenya: From large state companies to small private ones Kenya currently hosts around 400 Chinese firms spread across every sector. In February 2014, the Sino-African Centre of Excellence (SACE) foundation launched the Business Perception Index (BPI) survey to learn the views and experiences of Chinese companies in Kenya. The BPI surveyed 75 companies: 25 state owned and 50 privately owned enterprises. Many firms in the sample are in the manufacturing, construction, and resources sectors; Chinese firms are also involved in trade, tourism, hotels, and restaurants. Chinese electrical companies are working with Kenya Power, the government-owned power company, to establish 33 power lines in five Kenyan towns: Kitale, Awendo, Konza and Kutus (Business Daily 2015). Oil companies are sure to be attracted to Kenya’s new oil discoveries in Turkana and the northwest part of the Rift Valley, but the UK based Tullow Oil Plc has rights for drilling. 3.2.1 Chinese invest the most in metals, communications, and automotive original equipment manufacturing (OEM) Figure 23a below gives the amount of Chinese investment by sector between 2003 and 2015. Chi- nese companies invested US $178.9 million in metals, US $150.9 in communications, and US $68 million in automotive original equipment manufacturing (OEM). Investment in metals fits with Kenya’s high exports of titanium ores and copper to China. Companies also want to take advan- tage of the growing telecommunications market in Kenya: the number of mobile telephone has grown significantly in the last five years, reaching over 33 million subscribers in 2014; Chinese firms such as Huawei and ZTE wish to compete with other major companies such as Nokia or Ericsson (see figure 22). Figure 23b shows the overall distribution of Kenya’s FDI. The top sectors are communications at US $2.51 billion, alternative and renewable energy at US $2.36 billion, and coal, oil and natural gas at US $1.35 billion. In total, Chinese companies contributed US $464.75 million in capital investment with an average investment of US $35.75 million per project; other foreign companies in Kenya had a similar project size of US $35.5 million. 30 Subscribers (Millions) 20 10 0 2000 2002 2004 2006 2008 2010 2012 2014 Year Figure 22: Kenya’s mobile telephone subscriptions grew quickly over last decade Source: International Telecommunications Union 2016 24 3 FOREIGN DIRECT INVESTMENT IN KENYA 2500 Capital Investment (USD Millions) 150 2000 Alternative/Renewable energy Automotive OEM Capital Investment (USD Millions) Automotive Components Automotive OEM Chemicals Communications 1500 Coal, Oil and Natural Gas 100 Consumer Electronics Communications Financial Services Financial Services Food & Tobacco 1000 Metals Industrial Machinery, Equipment & Tools Real Estate 50 Metals Transportation 500 Warehousing & Storage 0 0 Industry sector Industry sector (a) Chinese FDI in Kenya by sector (b) FDI by sector Figure 23: China invests most in metals and communications in Kenya (2003-2015) Source: fDi Intelligence from The Financial Times Ltd 2015 Chinese companies invest the most in manufacturing activities Chinese firms focus on manu- facturing in Kenya, investing US $296.17 million between 2003 and 2015. Manufacturing was 64 percent of total investment for Chinese companies compared to only 24 percent of total investment among all companies. The average capital investment of China’s manufacturing projects was US $49.36 million with a total of six projects; the overall average project size of manufacturing projects was slightly larger at US $58.31 million with 59 projects. Compared to overall trends, Chinese FDI is absent in major sectors: electricity, construction, ICT and internet infrastructure, and logistics, distribution, and transportation. Even though companies have worked on high profile projects, Chinese investment in construction is tiny. The EXIM bank only provides loans: Chinese firms worked on the Standard Gauge Railway and the Thika Superhighway, but the government of Kenya must pay back the debt. Overall, construction had an average project size of US $215.4 million, with five projects in total. 300 3000 Capital Investment (USD Millions) Capital Investment (USD Millions) Business Services 200 Construction Business Services Design, Development & Testing Design, Development & Testing 2000 Electricity Education & Training Extraction Headquarters Headquarters Manufacturing ICT & Internet Infrastructure 100 Sales, Marketing & Support Logistics, Distribution & Transportation 1000 Manufacturing Sales, Marketing & Support 0 0 Industry Activity Industry Activity (a) Chinese FDI by industry activity (b) FDI by industry activity Figure 24: Chinese companies invest most in manufacturing (2003-2015) Source: fDi Intelligence from The Financial Times Ltd 2015 25 A. Sanghi and D. Johnson 80 Average Capital Investment (USD Millions) Average Capital Investment (USD Millions) 200 60 Business Services 150 Construction Business Services Design, Development & Testing Design, Development & Testing Electricity 40 Education & Training Extraction Headquarters 100 Headquarters Manufacturing ICT & Internet Infrastructure Sales, Marketing & Support Logistics, Distribution & Transportation Manufacturing 20 50 Sales, Marketing & Support 0 0 Industry Activity Industry Activity (a) Average Chinese FDI by industry activity (b) Average FDI by industry activity Figure 25: Chinese investment per project highest in headquarters creation and manufacturing (2003-2015) Source: fDi Intelligence from The Financial Times Ltd 2015 3.2.2 Corruption is the biggest obstacle for Chinese firms Chinese companies identified corruption as the most significant obstacle to doing business in Kenya: 53 percent of respondents said corruption was a “very significant obstacle” and 15 per- cent mentioned corruption as a “significant obstacle” (SACE 2014). Crime, theft, and personal safety was the second most significant obstacle: 37 percent said it was “very significant” and 25 percent responded “significant” (SACE 2014). 32.3 percent of foreign companies in the World Bank Enterprise Survey (2013) said corruption was a major constraint. According to a Kenya Na- tional Bureau of Statistics (KNBS) survey, 76.4 percent of respondents said corruption hurt their business. The negative perception of corruption was consistent across all sectors: manufacturing, transportation, financial services, and wholesale and retail trade (KNBS 2013). The BPI also finds that the Kenya Revenue Authority (KRA) is more likely to visit Chinese companies. The KRA visited Chinese companies 3.1 times on average during 2013 compared to only 2.2 times for all companies (SACE 2014). In addition, 60 percent of Chinese companies report that KRA officials asked for informal payments and gifts during their visit. Because of KRA corruption, Chinese companies see the tax system as a burden; 85 percent of Chinese companies surveyed have at least one Kenyan accountant to handle external audits and authorities, and 63 percent of companies have a Chinese accountant for internal audits (SACE 2014). 59.2 percent of foreign companies view the corporate tax, value-added tax (VAT), and the custom and excise duty as obstacles to business (KNBS 2013). 3.2.3 Chinese firms source most materials from China Chinese companies import 59 percent of goods from China and an additional seven percent indi- rectly from Kenyan suppliers. Chinese goods are much cheaper than the local equivalents, and firms simply find it more profitable to import inputs from China. Companies in the sample tar- get the domestic market with 96 percent of sales occurring in Kenya during 2013; the other four percent of sales were for Tanzania and Uganda (SACE 2014). Chinese companies also report a shorter time for customs clearance than other companies. The majority, however, report having a professional clearance agent, where corruption is rampant (SACE 2014). 26 3 FOREIGN DIRECT INVESTMENT IN KENYA 3.2.4 Chinese firms employ a large share of local workers Contrary to the popular belief that Chinese companies only hire Chinese workers, 93 percent of companies report hiring Kenyan employees; private enterprises are more likely to hire locals than state enterprises. In addition, larger firms are more likely to hire Kenyans than smaller firms: 40 percent of micro enterprises and all small, medium and large enterprises hire Kenyans (SACE 2014). Of the companies surveyed, Kenyans represent 78 percent of full-time and 95 percent of part-time employees. The companies had an average of 360 local employees: 252 were part-time (70 percent) and 108 were full-time (30 percent) (SACE 2014). All foreign manufacturing compa- nies in Kenya had 127.8 full time workers and 19.3 part-time workers on average. Manufacturing and construction companies are larger employers, hiring 762 employees on average compared to 45 in the services sector; all foreign manufacturing firms hire 158.3 full-time workers and 47.3 part-time workers on average (World Bank Enterprise Survey 2013). Chinese companies in the services sector hire 71 percent full-time employees, but the manfacturing and construction sectors hire only three percent full-time employees. 90 percent of manufacturing employees are local, and 82 percent of service sector employees are local. Chinese companies also hire more Kenyans over time: they had 102 full-time local employees upon establishment, and by the time of the BPI survey, they had hired 214 full-time local employees. 63 percent of Chinese companies said they had a policy of replacing Chinese employees with locals (SACE 2014). Larger firms were more willing to replace Chinese workers than smaller firms, and private enterprises were more will- ing to replace Chinese workers than state enterprises. A local technician is much cheaper than a Chinese technician because a work permit costs US $4,597. Local employees also receive basic insurance from Chinese companies: 44 of 68 companies noted that they offer basic insurance for all employees. Again, larger companies are more likely to provide basic insurance than smaller employers. 84 percent of manufacturing and construction companies offer insurance for Kenyans, and 72 percent of service sector companies offer insurance for locals. Overall, there are 663 for- eign and 20,790 local workers in managerial positions, 781 foreign and 87,589 local employees in skilled positions, and 633 foreign and 131,618 workers in unskilled positions. The average number of unskilled workers in foreign manufacturing companies is 49, or 37.3 percent of the workforce; detailed information about the workforce of Chinese companies is unavailable (KNBS 2013). Chinese companies create the most jobs in communications Figure 26a presents the number of jobs that Chinese companies have created in each sector. Chinese companies have created 2,170 jobs, 5.3 percent of the total 40,646 jobs created through FDI. The greatest number of jobs came from the communications sector at 931 jobs, followed by 500 jobs in the automotive original equipment manufacturing (OEM), and 342 jobs in the metals industry. Chinese companies have focused more on Automotive OEM than other countries. It makes up 23 percent of China’s overall jobs created compared to only 11.4 percent of total overall jobs. As figures 27 and 28 highlight, manufacturing takes the largest share of employment for both Chinese FDI and total FDI. Chinese manufacturing FDI created 1,600 jobs with 266.7 jobs on average, and manufacturing in total cre- ated 21,119 jobs with 357.9 jobs per project. Other major employers were customer contact centres, creating 2,218 jobs and 443.6 jobs per project. 27 A. Sanghi and D. Johnson 750 4000 Automotive OEM Automotive Components Building & Construction Materials Automotive OEM Business Machines & Equipment Jobs Created Jobs Created Communications Business Services 500 Consumer Electronics Communications Financial Services Consumer Electronics Food & Tobacco Consumer Products 2000 Industrial Machinery, Equipment & Tools Financial Services 250 Metals Food & Tobacco Metals 0 0 Industry Sector Industry Sector (a) Chinese FDI job creation by sector (b) Total FDI job creation by sector Figure 26: Chinese companies create most jobs in the communications sector (2003-2015) Source: fDi Intelligence from The Financial Times Ltd 2015 1500 20000 Business Services 15000 Construction Business Services Customer Contact Centre Jobs Created Jobs Created 1000 Design, Development & Testing Design, Development & Testing Education & Training Extraction Headquarters 10000 Headquarters Manufacturing ICT & Internet Infrastructure 500 Sales, Marketing & Support Logistics, Distribution & Transportation Manufacturing 5000 Sales, Marketing & Support 0 0 Industry Activity Industry Activity (a) Chinese FDI job creation by Industry Activity (b) Total FDI job creation by Industry Activity Figure 27: Chinese companies in manufacturing sector create the most jobs (2003-2015) Source: fDi Intelligence from The Financial Times Ltd 2015 1500 400 Business Services Average Jobs Created Average Jobs Created 300 Construction 1000 Business Services Customer Contact Centre Design, Development & Testing Design, Development & Testing Education & Training Extraction Headquarters 200 Headquarters Manufacturing ICT & Internet Infrastructure 500 Sales, Marketing & Support Logistics, Distribution & Transportation Manufacturing 100 Sales, Marketing & Support 0 0 Industry Activity Industry Activity (a) Average Chinese FDI job creation per project by (b) Average total FDI job creation per project by Industry Industry Activity Activity Figure 28: Chinese companies in manufacturing sector create the most jobs per project (2003-2015) Source: fDi Intelligence from The Financial Times Ltd 2015 China is also a major employer for Kenya. Figure 29 shows the number of jobs that each country creates through FDI. China ranks fifth overall with 2,170 jobs, and India is number one with 7,422 jobs. Hence, China and India are not only important sources of trade and investment, 28 3 FOREIGN DIRECT INVESTMENT IN KENYA but new jobs as well. Chinese companies have a higher number of jobs per project because they have fewer projects than other countries. Between 2003 and 2015, FDI from China created 166.92 jobs on average; total FDI generated 100 jobs per project.10 FDI has a strong ability to create jobs, and table 3 shows some of the progress. 6000 Jobs Created 4000 2000 0 China Egypt India Japan Nigeria South AfricaSouth Korea Spain UK United States Country Figure 29: China is fifth largest creator of jobs. India creates the most jobs (2003-2015) Source: fDi Intelligence from The Financial Times Ltd 2015 Table 3: FDI Projects and related employment (2007-2011) Year Number of Projects Employment Employees per project 2007 55 2,847 51.8 2008 73 4,341 59.5 2009 121 37,045 306.2 2010 129 15,753 122.1 2011 145 13,289 91.6 Source: Adapted from UNCTAD/DIAE/PCB/2012/6 Chinese firms already offer basic skills training 60 percent of Chinese companies offer formal training programs on skills, safety, and hygiene for local staff; 64 percent of all foreign firms in Kenya offer formal training (SACE 2014; World Bank Enterprise Survey 2013). Companies in the manufacturing sector and state enterprises are more likely to offer formal training than private companies and the service sector. A few Chinese projects aim specifically to provide skills train- ing for Kenyans, but businesses report that high turnover harms efforts for skills training and promoting Kenyans to higher level management because the high staff turnover raises the cost of training employees (SACE 2014). The BPI suggests that companies engage local employees in for- mal contracts, regardless of skill level, to build more trust and reduce the incentive to switch jobs and vandalize property. The language barrier also presents a challenge in training, but the grow- ing interest in Chinese langauge study among Kenyans should slowly improve communication 10 Kenya gained 40,646 jobs and 406 projects from FDI in total between 2003 and 2015. 29 A. Sanghi and D. Johnson (section 4.2.2). Chinese companies hire fewer female employees Chinese companies hire few female em- ployees: women only represent five percent of total employees for Chinese companies on average versus 29 percent for all companies in Kenya. Chinese companies in the services sector have 19 percent female employees compared to 31.8 percent female employees in the services sector over- all. In manufacturing, women are only three percent of employees in Chinese companies, but they are 22.7 percent of full-time employees overall (SACE 2014; World Bank Enterprise Survey 2013). 15 percent of the workforce in private companies is female, but only two percent of the workforce is female in state-owned enterprises. Hiring more female employees has a direct im- pact on poverty because women spend more on health, education, and household durables. When women work, they also tend to increase household savings (Lawson et al 2009). 3.2.5 Chinese firms face some competition from the informal sector 93 percent of Chinese companies in the sample have registered their business. 24 percent of com- panies reported competition from the informal sector, and informal competition was more likely in services (35 percent) than in manufacturing (13 percent) (SACE 2014). 3.2.6 Chinese companies less likely to take credit line Chinese firms are less likely to obtain a loan or credit line. Only 26 percent of Chinese companies obtained a loan; 36 percent of all companies in Kenya had a loan. (World Bank Enterprise Survey 2013). Potential borrowers are dissuaded by unfavorable interest rates and complex application procedures. 51 percent of Chinese companies reported not needing a loan because they had suf- ficient capital (SACE 2014). State owned enterprises had financing from headquarters in China, and private companies accumulated savings and loans from family and friends. Manufacturing and construction companies take as much credit as other companies in Kenya, but only 16 percent of companies in the services sector have taken a loan or credit line. Larger companies tend to have better access to loans and credit. 95 percent of Chinese companies, however, have a local bank account and can access basic banking services. 4 Official Development Assistance from China 4.1 Chinese aid is small compared to commercial activities Chinese aid to Sub-Saharan Africa was US $3.2 billion in 2013, and is quite small relative to its in- vestment stock of US $32.35 billion in Africa; the small amount of aid supports China’s no handout policy (China Africa Research Initiative 2013). China is more of a business partner than a donor, but it does provide grants and other forms of aid (Wang and Bio-Tchane 2007). Chinese aid only started to become significant after 2002, and it became 1.23 percent of total loans and grants to Kenya in 2003 (Mwega 2009; Onjala 2008). China also accounted for 10 percent of loans and 20 percent of physical infrastructure assistance to Kenya in 2005 (UNDP 2005; Onjala 2008). Unfor- tunately, as we discuss in section 4.3, we do not have access to quality aid data. 30 4 OFFICIAL DEVELOPMENT ASSISTANCE FROM CHINA 4.1.1 Kenya may rely more on Chinese aid because of volatile aid flows Chinese aid may continue to play a bigger role over time because Kenya’s foreign aid flows are volatile. The coefficient of variation of Kenya’s aid is 74.19 between 1960 to 2013; the coefficient of variation for Sub-Saharan Africa is 64.71 over the same period.11 Chinese official development assistance is a small fraction of Kenya’s total aid flows. Figure 30 shows Kenya’s top sources of official development assistance in 2013. 800 Gross ODA (USD Millions) AfDF (African Dev.Fund) 600 EU Institutions France Germany Global Fund 400 IDA IMF (Concessional Trust Funds) Japan United Kingdom 200 United States 0 Country Figure 30: US is a top donor for Kenya. Chinese finance does not meet OECD/DAC aid criteria (2013) Source: OECD DAC aid database 2015 China is not a member of OECD donors China is absent from the top ten donors because it is not a member of the OECD, so little of Chinese financing qualifies as official development assistance under the standard OECD definition (Br¨ autigam 2011). Figure 31 gives a sectoral distribution of aid to Kenya. Overall, aid to the health and population sector is the highest at US $574.3 million (31 percent) followed by economic infrastructure and services at US $407.5 million (22 percent). 11 The coefficient of variation is the ratio of the standard deviation to the mean: σ Cv = |µ| where µ is the average and σ is the standard deviation. 31 A. Sanghi and D. Johnson 600 ACTION RELATING TO DEBT ODA (USD Millions) 400 ECONOMIC INFRASTRUCTURE AND SERVICES Education Health and Population HUMANITARIAN AID MULTISECTOR Other social infrastructure and services 200 PRODUCTION SECTORS PROGRAMME ASSISTANCE UNALLOCATED/UNSPECIFIED 0 Country Figure 31: Kenya receives the most aid in health and populations services (China excluded) (2013) Source: OECD DAC aid database 2015 4.2 Most Chinese financing does not qualify as ODA The OECD Development Assistance Committee (DAC) calls export credits and loans with a less than 25 percent grant element other official flows (OOF); much of China’s financing overseas would fall under this category. The Chinese development bank, for example, created an China Africa development fund which provides equity investment capital. The ministry of finance also subsidizes the Exim bank’s preferential export credits and foreign aid loans, paying the difference between the interest rate and the actual cost of the loan. 97 percent of Chinese Exim Bank financ- ing is in suppliers’ credits and other official flows, and the remaining three percent is in foreign aid (Br¨autigam 2011). China gives most of its official financing at competitive commercial rates with a maturity of 12 to 15 years and a grace period of two to five years (Br¨autigam 2011). 4.2.1 China loans the most to ministries of energy and petroleum and transport and infrastructure China’s aid to Kenya is almost entirely loans. The government of China gave a total of KSH 20.6 billion (US $200.6 million) in loans for development expenditure for the fiscal year ending in 2015, and figure 32 shows their destinations (Ministry of Finance 2014). China gave KSH 12 billion (US $120 million) to the ministry of energy and petroleum, KSH 5.6 billion (US $56 million) to the ministry of transport and infrastructure, and KSH 2.5 billion (US $25 million) to the ministry of in- formation, communications and technology (ICT). The ministry of energy and petroleum amount includes the large geothermal loans from China EXIM bank. China’s loan breakdown matches its focus on natural resource extraction and infrastructure upgrading in many countries in Sub- Saharan Africa. Loans to the information, communication and technology sector go together with its investments in the communications sector (figure 23) and its strategy of using Kenya for access to the regional market. The small amount given to support devolution —the decentralization of roles and resources to the lowest levels of government —supports China’s non-interference for- eign policy. 32 4 OFFICIAL DEVELOPMENT ASSISTANCE FROM CHINA 12.5 10.0 Loan amount (KSH Billions) 7.5 5.0 2.5 0.0 Devolution Energy and Petroleum ICT Transport and Infrastructure Ministry Figure 32: Ministry of Energy and Petroleum receives most loans from China (2014) Source: Estimates of Development Expenditure Government of Kenya (2014) 4.2.2 Chinese stands out in education aid China is quite active in the education sector in Kenya and provides a number of training programs, university scholarships and language programs. Recent data is hard to gather, but China spent roughly US $120,000 on volunteers and US $1.2 million on trainings in 2010 (UNESCO 2015). Education aid is a small percentage of China’s bilateral loans and aid, but education is also a small part of the overall aid budget for traditional donors (figure 31). In 2011, China built St. Francis Sivo Primary School and four other primary schools in 2012 at an estimated cost of US $400,000, supported in part by China Youth Development Foundation (UNESCO 2015). China focuses on scholarships, vocational and professional programs, less on the formal edu- cation sector China’s Nanjing Agricultural University (NAU) and Egerton University began a partnership in 1994 that created fellowships for Egerton students to study and work in Nanjing. The program has trained 200 agricultural technicians to date from ten countries in East Africa (UNESCO 2015). From 2004 to 2008, China trained 697 Kenyans in short courses, focusing on agricultural technology. China offers more than 50 short courses on topics such as hydropower technology, malaria control, and hybrid rice technology (King 2010). Since 1984, China provided ten annual scholarhips to Kenyans and 40 annual scholarships in 2007. By 2013, China had more than 300 Kenyans studying in China. China offered 34 scholarships in 2015 (Ministry of Education Kenya 2015). Growing interest in Chinese language study in Kenya China also offers language scholarships: in 2009, the Confucious Institute at the University of Nairobi offered 49 language scholarships and the China information and culture communication center sent 40 self-sponsored students to China for language study (King 2010). China is becoming a more popular educational destination for Kenyans. Even without financial assistance, Kenyans find the competitve cost of Chinese universities attractive. As China’s influence in Kenya grows, the popularity of languages studies will likely rise. The Kenya Institute of Curriculum Design and Kenyatta University, for instance, 33 A. Sanghi and D. Johnson are developing plans to offer Chinese language classes in primary and secondary schools (KICD 2014). 4.3 Lack of quality Chinese aid data Since OECD and other official sources do not track China’s aid, researchers at AidData have cre- ated their own database of Chinese aid to Africa.12 The database uses media reports, government investment websites, and other official sources to calculate the total amount of Chinese develop- ment finance to various African countries by project. The data collection occurs in two stages. First, researchers look through Factiva, a Dow-Jones owned media search engine, to identify me- dia reports about aid projects in various countries. The team also uses donor and recipient country government websites to search for projects that they may have missed in the initial Factiva search (Strange et al 2013). After selecting the first group of projects, AidData re-searches using Google and other country search engines such as Baidu to retrieve the date, location, project cost, financial details, and status of the project. AidData then uses other staff members to review project details and information sources to correct mistakes. 4.3.1 Drawbacks of existing Chinese aid data: Media based data collection is problematic Compiling data from various sources to create something usable for analysis is commendable and brings more transparency to the debate on Chinese aid. Since the database is open source, anyone can suggest improvements and make corrections, allowing the database to improve with time. Re- lying on media reports, however, is problematic. Newspapers often report inaccurate information on the size and type of financial flows, or cover projects that are later cancelled. 6 of 20 Chinese deals cited in an AidData paper, for example, never actually happened, and China only financed 38 percent of the Merowe hydropower project in South Sudan. The authors, as a result, find that the total of the 20 Chinese deals was US $38 billion when the actual total is only US $9 billion autigam 2013). (Br¨ Data entry from media reports is also error-prone and classifying aid flows is hardly straight- forward. Selection bias is also a problem with media-based data collection: only projects that jour- nalists write about make it to the aid figures. Although AidData includes English and Chinese language media reports, researchers may also overlook other resources such as local newspapers in other languages or those only published in print. Given the issues with media data collection, we decide not to report aid numbers from these databases, and we will wait for a time when the open source community and researchers have had more time to correct errors. 5 Review and future directions FDI is usually beneficial for small countries when they have the capacity to absorb the investment. Chinese investment can become more transparent if Kenya can reduce corruption and promote good governance. Reforms in the rule of law and the regulatory environment will also help com- panies run more efficiently. Just as Jamaica promoted itself as an entry point to markets in the Caribbean, so too can Kenya continue to encourage Chinese companies to use Kenya as an entry 12 See china.aiddata.org 34 5 REVIEW AND FUTURE DIRECTIONS point to markets in Eastern and Southern Africa, particularly the East African community (EAC) and the Common Market for Eastern and Southern Africa (COMESA) (UNCTAD 2010). More so- phisticated Chinese companies in Kenya may enhance spillover effects and improve technology transfer. 5.1 Consider the long term growth of local industry While there are certainly benefits to more competition, policy makers must also consider the long- term growth of labor-intensive local industries. One of the major failings of the SGR project is the near-exclusion of local suppliers. Local manufacturers were only able to supply cement and a neglibible amount of steel to SGR; contracters imported the rest of the materials from China. Local manufacturers missed a valuable opportunity for capacity building, knowledge sharing and long- term job creation. A positive outcome of SGR however, was that Chinese competition forced some local cement manufacturers to upgrade their standards. Policy makers can design policies to help form linkages between small Kenyan firms, technical institutes, and Chinese contracters, treating the SGR project as a learning experience. With greater linkages, future infrastructure projects can draw from a larger pool of skilled labor and cut its costs. Inefficient Kenyan firms will close; firms that upgrade technology can enter the market, produce efficiently, and improve consumer welfare (UNCTAD 2010). 5.1.1 SGR spotlight: Non-existent capacity building On 19 September 2015, the president chaired a Cabinet Steering committee meeting with the China Road and Bridge Corporation (CRBC) on the progress of the SGR construction from Mombasa to Nairobi. Nearly half of civil work construction is complete As of 31 August 2015, 49 percent of the construction of civil work such as bridges, culverts and sub-grade is complete. Employment • 25000 skilled and unskilled local laborers • 17000 directly employed • 7000 indirectly employed • 2000 Chinese expatriates Local content: dropping the ball on Phase 1, Mombasa to Nairobi The panel said that purchases of local content was 30 billion KSH (US $304.6 million) by 31 August 2015. The involvement of local suppliers, however, was low. Prior to private sector engagement, CRBC imported 6000 tons of cement; afterwards, cement companies managed to supply the remain- der. Construction used negligible local steel and imported all other materials (Private sector, SGR consortium 2015). Since they do not have VAT exemption, local traders and suppliers 35 A. Sanghi and D. Johnson cannot compete with Chinese contractors; they also pay many other taxes and fees that Chi- nese firm bypass. Local traders and suppliers are still waiting for VAT reimbursements to date; without VAT reimburesments, local private sector cannot supply goods on time and are essentially shut out. Industrial parks along SGR The ministry of industrialization wants to place industrial zones along the SGR route and has already targeted land in Mariakani, Emali, Voi, Naivasha, and Athi River. At the moment, the ministry is seeking funding and commercial contracts are awaiting approval. Construction of Phase 2 from Nairobi to Malaba: Local content still an afterthought The president requested for skills building and local content manufacturing on the next portion of the railway between Nairobi and Malaba. Attendees discussed including a specific require- ment for local provision, but the extension of Phase 1 to Naivasha has already started. CRBC claims to have a capacity building program in place and mentioned a plan to rebuild the rail- way technical institute near Nairobi’s Wilson airport. No concrete plans for the program or the institute, though, are in place. Without a formal plan, capacity building will be forgot- ten. During phase one from Mombasa to Nairobi, CRBC also said they would consider local content and skills, but they did not have any obligation to buy from local private sector or build local capacity: they ended up importing the bulk of the materials from China. A repeat of Phase 1 will deny the next generation of local industry a valuable opportunity to learn, upgrade technologies, and gain efficiency. New developments: Engagement with private sector The private sector has established the Kenya Industry Sector Board Working Group and Linkage of Industry with Academia (LIWA) to promote dialogue and action across private sector, academia, and government. LIWA is working to help vocational institutes tailor curriculums that match the skills that industries demand. Cross-cutting groups can work around barriers to improving skills and standards; they also ensure that stakeholders voice concerns before the start of large infras- tructure projects. An agreement among all stakeholders will result in a conscious effort on creating jobs and advancing training and research (Private Sector, SGR consortium 2015). 5.2 Diversify FDI sources to avoid overreliance on China China is a rapidly rising source of FDI, but Kenya must continue to attract FDI from other sources to avoid a dependency on Chinese FDI. Kenya performs poorly in attracting FDI relative to its po- tential, and it must improve its efforts at marketing investment opportunities to Chinese as well as other Asian firms to increase FDI inflows. Kenya can attract FDI sources for much needed in- frastructure investment that will lead to more FDI in other sectors, lessening the dependence on China. In other words, improving infrastructure, reforming taxes, and cutting labor costs should help Kenya attract more FDI in general; diverse FDI sources can serve as insurance against finan- cial shocks and China’s future slowed growth (UNCTAD 2010). 36 5 REVIEW AND FUTURE DIRECTIONS 5.3 Monitor debt levels from China Kenya still has a heavy debt burden and China’s loans can bring debt to unsustainable levels. Some of China’s loans are nonconcessional, which can raise debt to GDP levels quickly. As of 30 June 2015, Kenya’s gross public debt to GDP ratio was 49.7 percent, and external loans contributed to the KSH 420.9 billion (US $4.3 billion) increase in public debt from June 2014 (National Treasury 2015). China is already Kenya’s top source of external financing. Figure 34 shows Kenya’s top external creditors as of June 2015. China holds the largest amount of debt at US $2.6 billion, 57 percent of Kenya’s total debt of US $4.51 billion. China’s large proportion of public debt may be part of the trend of partnerships between developing countries. 5.3.1 Debt to China is growing quickly China’s loans for energy and infrastructure investments such as SGR and geothermal power are accumulating quickly. Figure 33 shows Kenya’s external debt to China from 2010 to 201413 . Debt grew at an annual rate of 54 percent between 2010 and 2014, reaching US $821 million in 2014 from just US $146 million in 2010. Figure 35 shows the growth of debt to China from the third quarter of 2014 to the second quarter of 2015. China’s stock of Kenyan debt rate grew at 16.5 percent per quarter from US $ 753.3 million to US $2.56 billion; Kenya’s other top sources of debt stagnated or even declined. Kenya’s debt to Japan fell 2.98 percent per quarter and France declined 0.58 percent per quarter. Traditional donors must coordinate efforts with China to avoid undermining governance and debt sustainability programs. 800 Debt (USD Millions) 600 400 200 2010 2011 2012 2013 2014 Year Figure 33: Kenya’s debt to China is growing quickly (2010-2014) Source: Kenya National Bureau of Statistics Economic Survey 2015 13 Here 1 USD = 98.48 KSH. We take this from the UN Operational Rates of Exchange Effective date 30 Jun 2015 37 A. Sanghi and D. Johnson 2000 Debt (USD Millions) 1000 0 BELGIUM CHINA FRANCE GERMANY JAPAN Country Figure 34: Kenya owes most debt to China (2015) Source: The National Treasury of Kenya (2015) 2500 2000 China Debt (USD Millions) Japan 1500 France 1000 Germany Belgium 500 0 2013Q3 2014Q1 2014Q3 2015Q1 Year Figure 35: Kenya’s debt to China outpaces the rest (2015) Source: The National Treasury of Kenya (2015) 5.4 Supply-side Shortages: Reducing labor costs Chinese firms may see Kenya’s lower labor cost as an advantage, but the unit labor cost is still quite high. The unit cost of labor in Kenya is 25 percent of the value added, but is only 15 percent of value added in China. The combination of low wages and high productivity results in a low unit cost of production for Chinese workers; low productivity, however, raises the unit labor cost of Kenyan labor. Reducing the unit cost of labor —wages increases should at most match produc- tivity increases —through employment subsidies or reduction of formal sector wages can attract more Chinese firms aiming to take advantage of the low cost Kenyan labor force, especially firms that are developing export businesses to the United States after the renewal of AGOA. It may also help export competitiveness generally and help to reduce the overall trade deficit. 38 5 REVIEW AND FUTURE DIRECTIONS 5.4.1 Encourage technology transfer and capacity building with infrastructure projects Chinese firms involved in labor-intensive activities identified the lack of a skilled workforce as a constraint to doing business in Kenya (SACE 2014). The language barrier is a constraint for skills transfer, but Kenya may continue to promote Chinese language training. Many Kenyans are al- ready learning Chinese and the government of China offered 34 full scholarships to study in China in 2015; greater numbers of Kenyans learning Chinese will lower language barriers on both sides. Policy makers can also help initiatives such as LIWA to create links between Chinese firms and Kenyan universities and training institutes to address key skills gaps. At the China-Africa summit in Johannesburg on 4 December 2015, China pledged to support capacity building in Africa and build five “Jiao Tong”, or transportation universities that train scientists, engineers, and techni- cians; leaders should negotiate for a “Jiao Tong” in Kenya, a step that will ensure more technology transfer and skills development. The literature suggests that FDI creates an opportunity for do- mestic firms to enter and supply inputs to foreign firms, increasing labor demand. Since requiring local content is prohibited under the WTO Trade Related Investment Measures (TRIM) agreement, Kenya can seek alternatives to promote local content. For example, Estonia offered funds to large companies to develop linkages with local suppliers, and Jamaica took a “cluster” approach that attracted FDI to sectors that could exploit synergies (UNCTAD 2010). Similarly, Kenya can build capacity in small and medium entreprises (SME) to supply Chinese businesses by connecting local suppliers to large Chinese companies. But Kenya should not push too hard for domestic content Though many countries have lo- cal content laws, Kenya should be careful to avoid overpushing for domestic content. Grossman (1981) finds that requiring local content is equivalent to taxing the foreign firm, hurting efforts to increase and diversify FDI, and in some cases even lowering domestic value added. In fact, firms will fail to meet the local content requirement if the cost of the foreign intermediate good is less than the cost of buying enough of the domestic intermediate good to meet the content require- ment. Chinese companies import 59 percent of inputs from China because they are cheaper than local goods; restricting foreign inputs may be unprofitable for Chinese firms. When promoting local content, Kenya can also look to improve the competitiveness of local industry so firms nat- urally move towards local products. It can also offer support to companies that develop linkages with local firms. 5.4.2 Bring more transparency to loans and infrastructure projects China’s loans come with attractive interest rates and without strings attached for good gover- nance. Though Kenya faces few conditions on its aid, other countries in Sub-Saharan Africa lever- age China’s financing to gain bargaining power with traditional donors.14 The loans, however, could harm Kenya in the long-run because of their lack of transparency and failure to tie aid to key governance reforms. Traditional donors can work to help China follow the environmen- tal protection standards, and civil society groups can demand increased transparency in aid and finance, pushing Chinese and non-Chinese firms in the natural resources sector to involve them- selves in the Extractive Industries Transparency Initiative (EITI) and report payments made to the government. 14 Angola turned down financing from the IMF that required measures to improve transparency in the oil sector, and accepted US $2 billion in loans from the Chinese EXIM bank (Davies 2007) 39 A. Sanghi and D. Johnson 5.4.3 Special Economic Zones Kenya’s export processing zones have acheived some diversification in manufacturing FDI, but garment manufacturing is still the majority of employment and investment. Table 4 provides the distribution of FDI by sector in the export processing zones (EPZ). The garment sector is the largest within the EPZ: it is 79.7 percent of employment, 53.1 percent of exports, 48.9 percent of total sales, 26.9 percent of firms, and attracts 27.9 percent of investment. Notably, Kenya has managed some diversification in the EPZs; the agro-processing sector contributes 11 percent of employment, 15.9 percent of export sales, 14.7 percent of total sales, and 16.1 percent of investment. Services FDI has also increased within the EPZ. Farole (2011) finds that horticultural and food-processing, call centers, and human and veterinary pharmaceuticals have entered the zones. Despite the diver- sification, Kenya’s EPZs have remained stagnant at just over US $400 million in exports since its inception during the 1990s. In addition, the EPZ program’s exports relative to national levels are still small (Farole 2011). Firms operating inside the EPZ cannot sell products to the domestic mar- ket. Kenya can open sales to the domestic and regional market to stimulate Chinese and other foreign investment in the SEZs. The positive effects from the higher investment and improved infrastructure may help sales performance and employment growth in the SEZs (Farole 2011). 40 5 REVIEW AND FUTURE DIRECTIONS Table 4: Export Processing Zones sector contribution in 2012 (%) Sector Number of Employment Exports Total Sales Investment Firms Agro pro- 22 11 15.9 14.7 16.1 cessing Beverages 4 0.5 0.9 0.8 0.6 Chemicals 1 0.2 0.1 0.1 2.9 Dartboard 1 0.7 1.5 1.4 1.8 Electricals 3.7 0.1 6.9 6.3 1.2 Food Pro- 3.7 0.9 1.0 1.5 5.6 cessing Garments 26.9 79.7 53.1 48.9 27.9 Garment 6.1 0.1 0 0.1 0.2 supported services Minerals, 4.9 1.2 8.2 7.5 20.3 Metals, and Gemstones Medicine 2.4 0.8 1 1.4 2.7 and medical services Plastics 6.1 1.3 1 1.3 3 Printing 1.2 0.7 1.5 7.5 6.2 Relief sup- 2.4 0.3 2.5 2.5 1.4 plies Services 12.2 2.5 6.4 6.0 10 Other 2.4 0.02 0.01 0.01 0.01 Total 100 100 100 100 100 Source: Adapted from Export Processing Zones Annual Report 2012 5.4.4 Conclusion Greater competition from imports lowers consumer prices and gives some producers access to cheaper inputs and capital goods. Chinese imports do not necessarily displace domestic produc- tion, as they may replace imports from other countries, lowering the cost of imports. The imports could also boost labor productivity within certain sectors and increase employment in services. But Kenya must still work to improve its manufacturing sector to compete with China’s low cost manufacturing; Kenyan producers have to upgrade their skills or specialize in areas where they have a comparative advantage. To improve exports, Kenya could negotiate for duty free access for cut flowers as part of the 404 primary products that China already allows to enter duty free. Exports to China, especially of services, may increase once China transitions to a consumption- driven economy closer to 2030. When focusing on trade, policy makers should shift attention to improving export competitiveness and the overall balance of trade. Kenya’s exports need to im- prove overall, not just to China. 41 A. Sanghi and D. Johnson Chinese companies create a large number of local jobs, and they are the fifth largest employer from foreign direct investment in Kenya. Unlike in other places in Sub-Saharan Africa, China invests in more than natural resources in Kenya. Although metals investment is large, Chinese companies have also invested US$150.9 million in the communications sector, the second high- est amount. China has also provided considerable financing for infrastructure, infrastructure that will help growth by lowering the cost of doing business. For future infrastructure projects, deci- sion makers can create programs to encourage technology transfer with local firms and vocational institutes. Local industry will grow as more firms meet the standards to supply inputs for mega infrastructure projects, providing much needed jobs and skills. China is a business partner and is uninterested in the internal affairs of other countries. When considering opportunities, Kenya must remember local capacity development to absorb the most from investment and infrastruc- ture projects. Our paper is not the final word, but the start of a discussion on China’s economic influence in Kenya. We hope that future research will address data limitations in aid flows and investment, and examine what Chinese involvement means for growth and poverty reduction. 42 REFERENCES References Ajaikaye, O., “China and Africa-Opportunities and Challenges”, Presentation at the African Union Task Force on Strategic Partnership Between Africa and Emerging Countries of the South, Addis Abba, Ethiopia, 11-13 September 2006 Arkolakis, C., “Notes on Graduate International Trade” Notes prepared for Graduate International Trade at Yale, May 2012 Anderson, J. E., & Van Wincoop, E. (2003). “Gravity with gravitas: a solution to the border puzzle”, American Economic Review 93 autigam, D. (2011). “Aid With Chinese Characteristics: Chinese Foreign Aid and Development Br¨ Finance Meet the OECD-DAC Aid Regime”. Journal of international development, 23(5), 752-764 autigam, D. “Rubbery Numbers for Chinese Aid to Africa”, China in Africa: The Real Story April Br¨ 2013 autigam, D., & Gallagher, K. P. (2014). “Bartering Globalization: China’s Commodity-backed Fi- Br¨ nance in Africa and Latin America.” Global Policy, 5(3), 346-352. Chege, M., “Economic relations between Kenya and China, 1963-2007.”, Washington D.C.: CSIS 2008 Chen, G., Geiger, M., and Mingui Fui, “Manufacturing FDI in Sub-Saharan Africa: Trends, Determi- nants, and Impact” World Bank Working Paper, June 2015 Chen, W., Dollar, D., and Tang, H. (2015). “Why is China investing in Africa? Evidence from the firm level.” Constantaras, E. (2014) “Tullow Oil sub-contractor employee data reveal contracting driven by market forces, not economic policy in Turkana”, http://landquest.internewskenya.org/ tullow-oil-sub-contractor-employee-data-reveal-contracting-driven-by-market-forces-not-econom Accessed 6 November 2015 Davies, P. (2007). China and the end of poverty in Africa: towards mutual benefit?. Sundbyberg: Diakonia. Dupasquier, C. and Patrick N. Osakwe, “Foreign Direct Investment in Africa: Performance, Chal- lenges, and Responsiblities”, Journal of Aisan Economics 17 (2): 241-60 Executive Research Associate (Pty) Ltd (2009), “China in Africa: A strategic overview”, http:// www.ide.go.jp/English/Data/Africa_file/Manualreport/pdf/china_all.pdf Farole, T. Special economic zones in Africa: comparing performance and learning from global experiences. World Bank Publications, 2011 Farole, Thomas and Mukim, M, “Manufacturing Export Competitiveness in Kenya : policy note.”, Washington, D.C.: World Bank Group http://documents.worldbank.org/curated/en/2013/ 12/18746089/manufacturing-export-competitiveness-kenya-policy-note Feinberg, Robert M. “Do international shocks affect small wholesalers and retailers?” Review of World Economics 146, no. 2 (2010): 323-338 43 A. Sanghi and D. Johnson Fiott, D., “The EU and China in Africa: the case of Kenya”, Madariaga Paper, Vol. 3 No. 5, July 2010 Grossman, G. M. (1981). “The theory of domestic content protection and content preference”. The Quarterly Journal of Economics, 583-603 Goldstein, A., Pinaud, N., Helmet, R. and Xiaobao Chen, “China and India: What’s in it for Africa?”, OECD Development Centre, Feb 2006 Musembi, R., “Gdc¡U+0080¿¡U+0099¿S Geothermal Development Strategy For Kenya: Progress & Opportunities”, Presented for Power Africa-Africa Union Commission Geothermal Roadshow, September-October 2014 Hansen, Eric Rolf; Moon, Young; Mogollon, Maria Paulina. 2015. Kenya - Leather industry : diagnosis, strategy, and action plan Washington, D.C. : World Bank Group. http://documents.worldbank.org/curated/en/2015/09/25022153/ kenya-leather-industry-diagnosis-strategy-action-plan Head, K., Mayer, T., & Ries, J. (2010). The erosion of colonial trade linkages after independence. Journal of International Economics, 81(1), 1-14 Herbling, D., “Investor targets Chinese with Westlands hotel”, Business Daily, August 2014 Kalinaki, D. “Kenya’s ambitious infrastructure projects will improve links with neighbours”, Finan- cial Times, 02 December 2014 Kaplinsky, R., Morris, M. “Dangling by a thread: How sharp are the Chinese scissors?”, Institute of Development studies at University of Sussex. Report prepared for DFID trade division, Feb 2006, http: //asiandrivers.open.ac.uk/documents/Dangling_by_a_thread_Feb_06_%20final.pdf Kaplinsky, R., McCormick, D., and Mike Morris, “The Impact of China on Sub-Saharan Africa”, http://opendocs.ids.ac.uk/opendocs/handle/123456789/4142 Kaplinsky, R., “What does the rise of China do for industrialisation in Sub-Saharan Africa?”, Review of African Political Economy 35 (115): 7-22 Kaplinsky, R. and Mike Morris, “Chinese FDI in Sub-Saharan Africa: Engaging with Large Drag- ons”, http://www.palgrave-journals.com/ejdr/journal/v21/n4/pdf/ejdr200924a.pdf Fengler, Wolfgang and Kiringai, Jane (2010). Kenya economic update : running on one engine - Kenya’s uneven economic performance with a special focus on the port of Mombasa Kenya Economic Update; edition no. 2. Washington D.C.: World Bank Group http://documents.worldbank.org/curated/en/2010/06/12384899/ running-one-engine-kenyas-uneven-economic-performance-special-focus-port-mombasa World Bank. (2015). Kenya economic update : storm clouds gathering ¡U+0080¿¡U+0093¿ the economy facing strong headwinds with a special focus on public participa- tion. Kenya economic update; edition no. 12. Washington D.C.: World Bank Group. http://documents.worldbank.org/curated/en/2015/10/25153454/ kenya-economic-update-storm-clouds-gathering-economy-facing-strong-headwinds-special-focus-pu 44 REFERENCES Kenya Institute of Curriculum Development (2014). “KICD and KU partner to intro- duce Chinese curricular in schools”. http://kicd.ac.ke/newsletter/index.php/news/ item/405-kicd-and-ku-partner-to-introduce-chinese-curricular-in-schools. Accessed September 25, 2015 King, K. (2010). “China’s cooperation in education and training with Kenya: A different model?” International Journal of Educational Development, 30(5), 488-496 Kenya National Bureau of Statistics, Foreign Investment Survey 2013 (2013) Kenya National Bureau of Statistics, Economic Survey 2015 (2015) KPMG, Kenya Country Profile: KPMG Africa Region 2012/2013 http://www.kpmg.com/Africa/ en/KPMG-in-Africa/Documents/2012-2013%20Country%20Profiles/Kenya%20Country% 20Profile_2012-2013.pdf Accessed May 15, 2015 Kremer, Michael, Jean N. Lee, Jonathan M. Robinson, and Olga Rostapshova. “The return to capital for small retailers in Kenya: evidence from inventories.”, Working paper, Harvard and UCSC. 2010. Lakatos, Csilla, Maryla Maliszeska, and Israel Osorio-Rodarte (2015). “China’s slowdown and re- balancing: Potential Growth and Poverty Impacts on Sub-Saharan Africa”. Mimeo Lyman, P. 2005. “China’s Rising Role in Africa.” Presented to the U.S.-China Commission. July 21. Washington D.C.: Council on Foreign Relations Lawson, S., and Gilman, D. B. (2009). The Power of the Purse: Gender Equality and Middle-Class Spend- ing. Goldman Sachs, Global Markets Inst. Leifield, P. (2013). texreg: Conversion of Statistical Model Output in R to LaTeX and HTML Tables. Journal of Statistical Software, 55(8), 1-24. http://www.jstatsoft.org/v55/i08/ Manson, K. “Walmart lines up for slice of Kenyan retail.” Financial Times, June 1, 2015 Ministry of Finance. “Estimates of Development Expenditure of the Government of Kenya for the year ending 30th June 2015”, April 2014. http://www.cofek.co.ke/GoK%20Development% 20Budget%202014-2015.pdf Moss, T and Sarah Rose, “China Exim Bank and Africa: New Lending, New Challenges” CGD Notes (2006) Mutegi, M. “Kenya Power signs deal to connect five key towns” Business Daily Africa January 2015 Mwega, Francis M., “A case study of aid effectiveness in Kenya” Wolfensohn Center for Development Working Paper 8 2009 Nord, R., Harris E., Giugale, M., and Jeffrey Lewis. “Kenya Fifth Review Under The Three-Year Arrangement Under The Extended Credit Facility And Request For A Waiver And Modification Of Performance Criteria-Debt Sustainability Analysis” International Monetary Fund April 2013 45 A. Sanghi and D. Johnson Onjala, J. “A Scoping Study on China-Africa Economic Relations: The Case of Kenya”, AERC Scoping Studies on China-Africa Relations. AERC, Nairobi, March 2008 http: //www.researchgate.net/profile/Joseph_Onjala/publication/239732019_A_Scoping_ Study_on_China-_Africa_Economic_Relations_The_Case_of_Kenya_Revised_Final_ Report/links/0deec51c27bde2ac11000000.pdf R Core Team (2015). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. http://www.R-project.org/ Reilly, J (2015). “The role of China as an education aid donor”, Background paper prepared for the Edu- cation for All Global Monitoring Report 2015 http://unesdoc.unesco.org/images/0023/002324/ 232475e.pdf SinoAfrica Centre of Excellence Foundation. Business Percpetion Index Kenya 2014: Chinese companies’ perception of doing business in Kenya 2014 2014. Santos Silva, J. and Tenreyo, S. (2006). “The Log of Gravity”, Review of Economics and Statistics, 88(4), 641-658. Situma, E. “Chinese firm to rebuild Nginyo Kariuki’s Lenana Mount hotel”, Business Daily May 2015 Strange, M. A., O’Donnell, B., Gamboa, D., Parks, B., and Charles Perla, “AidData’s Methodology for Tracking Underreported Financial Flows” AidData: Open Data for Development. 2010 Subramanian, U. and Matthias Matthijs, “Can Sub-Saharan Africa Leap into Global Network Trade?” World Bank Publications Vol. 4112 https://openknowledge.worldbank.org/bitstream/ handle/10986/6888/wps4112.pdf Toyoshima, Toshihiro, Yutaka Yoshino, and Chad Leechor. “Patterns of Africa-Asia Trade and In- vestment: Potential for Ownership and Partnership.” (2004) United Nations Conference on Trade and Development, How to Attract and Benefit from FDI in Small Countries: Lessons from Estonia and Jamaica (2011) United Nations Conference on Trade and Development, Report on the implementation of the investment policy review (2013) United Nations Conference on Trade and Development, A Practical Guide to Trade Policy Analysis (2012) Wang, J., Abdoulaye, B., “Africa’s Burgeoning Ties with China”, Finance & Development, 45 Wickham, Hadley. ggplot2: elegant graphics for data analysis. Springer New York, 2009 World Bank. Commodity Markets Outlook Accessed June 20, 2015 www.worldbank.org/content/dam/ Worldbank/GEP/GEPcommodities/GEP2015b_commodity_Apr2015.pdf World Bank. Africa’s Pulse (October 2015). Washington, D.C. World Bank Group http://documents. worldbank.org/curated/en/2015/10/25116683/africas-pulse-october-2015 46 A GRAVITY MODEL OF TRADE Handjiski, Borko; Sanghi, Apurva; Bogoev, Jane; Larbi, George Addo; Angelique, Umutesi; Randa, John; Kiringai, Jane Wangui; Chege, Patrick Nderitu; Whimp, Kathy; Gubbins, Paul Michael; Mistiaen, Johan A.; Farole, Tom; Nishiuchi, Toru; Battaile, William G.; Van Doorn, Ralph; Saez, Juan Sebastian; Hollweg, Claire Honore; Cirera, Xavier; Mogollon, Maria Paulina; Dowdall, Georgia Frances Isabelle; Onder, Harun. 2016. Kenya - Country eco- nomic memorandum : from economic growth to jobs and shared prosperity. Washington, D.C. : World Bank Group. http://documents.worldbank.org/curated/en/2016/03/26029597/ kenya-country-economic-memorandum-economic-growth-jobs-shared-prosperity Zafar, A. “The Growing Relationship between China and Sub-Saharan Africa: Macroeconomic, Trade, Investment, and Aid Links”, The World Bank Research Observer 22 (1): 103-30, Oxford Uni- versity Press, 2007 Appendix A Gravity model of trade Anderson (1979) provides the first attempt at a model of gravity under the assumptions that con- sumers have a desire for foreign goods and each country produces a unique good that it exports to the rest of the world. In the model, all consumers purchase at least some goods from all countries (UNCTAD/WTO 2012). Let Xij be the bilateral trade flows between country i and country j, G be a constant term, Yi and Yj be country i and country j’s GDPs, and Dij be the distance between i and j. Then the simplified version of the gravity model is Yi Yj Xij = G (1) Dij . Here bilateral trade increases with country i and j’s GDPs and decreases with the distance between i and j. In general, a gravity equation can take the form of β β β Xij = eα Yi 1 Yj 2 Dij3 (2) where β 1 and β 2 equal one and β 3 is close to -1 which gives us (1). We can take advantage of the multiplicative form and express (2) in log form with an unobservable term uij , which gives ln Xij = α + β 1 ln Yi + β 2 ln Yj + β 3 ln Dij + uij (3) which allows us to estimate trade flows between i and j using a least squares approach with an easy interpretation of the coefficients as elasticities. Here we have added the unobservable term uij . Although (1) resembles the gravity equation from physics, it also has a grounding in economic theory. In fact, a wide range of trade theories can produce a gravity equation. Bergstrand (1985 and 1989) show that one can derive gravity from a monopolistic competition model from Paul Krugman (1980) (UNCTAD/WTO 2012). The monopolistic competition model does not assume that location of production matters for good differentation, and countries and firms specialize in the production of goods. Gravity can even arise in a perfect competition model. Eaton and Kor- tum (2002) show in the perfect competition model that the lowest cost producer across countries that produce a good will supply that particular good (Arkolakis 2012). 47 A. Sanghi and D. Johnson One typically estimates trade flows between i and j using a least squares approach with a simple interpretation of the coefficients as elasticities, but the OLS model is biased because it fails to account for zero reported trade; the Poisson model is better suited to handle a large number of observations with zero trade (Santos Silva and Tenreyo 2006). Let the exporter i be Kenya for all time t. Let Xijt be the export volume, β 0 be a constant term, Yit be the GDP of the exporter, Yjt be the importer GDP, Nit be the exporter population, Njt be the importer population, Dijt be the bilateral distance between i and j, and Z ijt be a vector of characteristics that include information on colonial histories, regional trade agreements, year dummies, a China dummy variable and China×year interaction dummies. Xijt = exp( β 0 + β 1 Yit + β 2 Yjt + β 3 Nit + β 4 Njt + δ Dijt + Z ijt γ) + uijt (4) A.1 Model Anderson (2003) develop a theoretical model for gravity equations. Let wi be the unit price of a good produced in country i and τij be the transport cost from country i to j. Let N be the number of countries. Here we follow the convention that costs are “iceberg” meaning that some fraction of the goods shipped from i to j will not arrive. Let wi τij be the price the importer in country j pays at the port. The constant elasticity of substitution (CES) price index is given by N Pj1−σ = ∑ (wi τij )1−σ (5) i =1 where Pj1−σ is the price index for country j and σ is the elasticity of substitution, or the measure of substitutability between goods. The representative consumer utility in country j is σ N 1− σ 1− σ σ −1 Uj = ∑ αi σ cijσ (6) i =1 =1 wi τij cij . Here 0 < α < ∞ is a given preference parameter subject to the budget constraint y j = ∑iN and cij is the consumption level of good i in country j. Then from the CES utility function, we can get the demand function of 1− σ αi wi τij xij = yj. (7) Pj Now let the value of aggregate production be yi = ∑ N j=1 xij , meaning that output matches the total demand of good i. Substituting (7) we get N 1− σ τij y i = ( α i wi )1− σ ∑ y j , (i = 1, . . . , N ). (8) j =1 Pj Now let yw = ∑ N j=1 y j be world income. If we solve (8) for ( αi wi ) 1−σ and then substitute into (7) we obtain 1− σ 1− σ yi y j τij yi y j τij xij = 1− σ = 1− σ y . (9) y j Pj1−σ 1− σ τij τij ∑N yw ∑ N j j =1 Pj j =1 Pj yw Pj 48 B GRAVITY DATA Let 1 N 1− σ 1− σ τij yj Πi = ∑ Pj yw (10) j =1 be the exporter ease of access from country i to j. Similarly, we can define the importer ease of 1− σ access by substituting (8) solved for αi wi into (5) to get   1− σ N yi τij Pj1−σ = ∑   τij 1− σ  i =1 ∑N j =1 Pj yj   1− σ N  yi τij  (11) = ∑   1− σ  i =1 τij yj yw ∑ N   j =1 yw Pj N 1− σ yi τij = ∑ yw Πi . i =1 A key simplifiying assumption of symmetric trade costs or τij = τji allows us to write Πi = Pi . So we now have N 1− σ τij yi Pj1−σ = ∑ Pi yw (12) i =1 and we can recover a familiar form of the gravity equation 1− σ yi y j τij xij = . (13) yw Pi Pj Appendix B Gravity Data We conduct our analysis using a data set covering bilateral trade between all pairs of countries from 1948 to 2014 used in Head et al (2010)15 ; we extend the data to include bilateral trade up to 2014 with information from UN Comtrade. A detailed explanation and codebook of the data are found in the appendix of Head et al (2010). However, one key aspect of the data is the treat- ment of missing trade flows. A trade flow of zero will cause problems for the log transformations. The authors therefore include only non-zero trade flows in the data. While this leads to selection bias, other forms of dealing zero flows are also unsatisfactory. For example, a common practice is to add one to zero trade flows which yields incorrect estimates because it does not truly reflect the underlying values (UNCTAD/WTO 2012). Another common technique is to use a Poisson maximum likelihood estimator that deals directly with zeros. However estimates will be biased if there is a large number of zeros in the data. Tobit estimation, which can also handle zero trade, imposes assumptions of the unobservable term uijt that are too strong, namely log normality and 15 http://www.cepii.fr/cepii/en/bdd_modele/download.asp?id=8 49 A. Sanghi and D. Johnson homoskedasticity. Choosing a left censor or lower bound also presents problems because esti- mates are highly sensitive to the choice of left censor value, or the value assigned to zero trade. Hence, in using this data set we have also dropped zero trade flows. Appendix C Results Table C5: Estimates of Kenya’s Trade Flows (1948-2014). Dependent variable: Kenya’s exports Pooled OLS Fixed Effects Random Effects ∗∗∗ ∗∗∗ Constant 5.26 (0.61) 5.26 (0.61) Kenya GDP per capita (Log) 0.95 (0.13)∗∗∗ 0.95 (0.13)∗∗∗ Importer GDP per capita (Log) 0.64 (0.02)∗∗∗ 0.69 (0.03)∗∗∗ 0.64 (0.02)∗∗∗ Population Kenya (Log) −1.47 (0.14)∗∗∗ −1.47 (0.14)∗∗∗ Importer Population (Log) 0.42 (0.01)∗∗∗ 0.64 (0.03)∗∗∗ 0.42 (0.01)∗∗∗ ∗∗∗ ∗∗∗ Distance (log) −1.25 (0.05) −1.31 (0.10) −1.25 (0.05)∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ Regional Trade Agreement 1.73 (0.13) 1.80 (0.15) 1.73 (0.13) ∗∗∗ ∗∗∗ ∗∗∗ Contiguous 1.14 (0.14) 0.77 (0.17) 1.14 (0.14) Common Language 0.67 (0.06)∗∗∗ 0.92 (0.08)∗∗∗ 0.67 (0.06)∗∗∗ GATT Kenya −1.36 (0.12)∗∗∗ −1.33 (0.11)∗∗∗ −1.36 (0.12)∗∗∗ GATT importer 0.09 (0.06) −0.05 (0.05) 0.09 (0.06) ∗∗∗ ∗∗∗ Colonial History 2.88 (0.14) 2.16 (0.11) 2.88 (0.14)∗∗∗ ∗∗∗ ∗∗∗ China dummy 1.11 (0.19) 1.11 (0.19) Year dummies Year1949 −0.04 (0.37) 0.04 (0.40) −0.04 (0.37) Year1950 0.01 (0.37) 0.03 (0.38) 0.01 (0.37) Year1951 −0.25 (0.37) −0.25 (0.35) −0.25 (0.37) Year1952 0.39 (0.40) 0.46 (0.40) 0.39 (0.40) ∗ Year1953 0.68 (0.35) 0.81 (0.39) 0.68 (0.35) Year1954 0.14 (0.34) 0.31 (0.30) 0.14 (0.34) Year1955 −0.02 (0.35) 0.13 (0.34) −0.02 (0.35) Year1956 −0.43 (0.37) −0.34 (0.36) −0.43 (0.37) Year1957 −0.32 (0.38) −0.24 (0.38) −0.32 (0.38) Year1958 0.01 (0.34) 0.20 (0.36) 0.01 (0.34) Year1959 −0.49 (0.35) −0.37 (0.34) −0.49 (0.35) Year1960 0.21 (0.34) 0.36 (0.34) 0.21 (0.34) Year1961 0.08 (0.33) 0.30 (0.29) 0.08 (0.33) Year1962 0.29 (0.31) 0.41 (0.30) 0.29 (0.31) Year1963 0.11 (0.33) 0.22 (0.31) 0.11 (0.33) Year1964 0.11 (0.36) 0.14 (0.29) 0.11 (0.36) Year1965 0.01 (0.34) 0.04 (0.33) 0.01 (0.34) Year1966 0.19 (0.35) 0.16 (0.35) 0.19 (0.35) Year1967 −0.16 (0.33) −0.14 (0.30) −0.16 (0.33) Year1968 0.40 (0.32) 0.54 (0.32) 0.40 (0.32) Year1969 0.18 (0.32) 0.27 (0.29) 0.18 (0.32) Year1970 −0.12 (0.33) −0.05 (0.35) −0.12 (0.33) 50 C RESULTS Pooled OLS Fixed Effects Random Effects Year1971 0.16 (0.34) 0.18 (0.31) 0.16 (0.34) Year1972 0.41 (0.32) 0.47 (0.34) 0.41 (0.32) Year1973 0.08 (0.33) 0.19 (0.30) 0.08 (0.33) Year1974 −0.13 (0.34) −0.04 (0.35) −0.13 (0.34) Year1975 0.08 (0.32) 0.15 (0.30) 0.08 (0.32) Year1976 0.03 (0.34) 0.05 (0.32) 0.03 (0.34) Year1977 0.41 (0.32) 0.46 (0.32) 0.41 (0.32) Year1978 −0.04 (0.34) 0.06 (0.34) −0.04 (0.34) Year1979 −0.01 (0.33) 0.03 (0.34) −0.01 (0.33) Year1980 0.15 (0.33) 0.16 (0.31) 0.15 (0.33) Year1981 −0.09 (0.33) 0.06 (0.32) −0.09 (0.33) Year1982 0.37 (0.33) 0.42 (0.29) 0.37 (0.33) Year1983 0.18 (0.34) 0.19 (0.37) 0.18 (0.34) Year1984 −0.22 (0.33) −0.11 (0.31) −0.22 (0.33) Year1985 0.03 (0.34) 0.03 (0.29) 0.03 (0.34) Year1986 −0.19 (0.33) −0.24 (0.32) −0.19 (0.33) Year1987 0.33 (0.33) 0.31 (0.29) 0.33 (0.33) Year1988 −0.02 (0.34) 0.11 (0.33) −0.02 (0.34) Year1989 0.34 (0.31) 0.38 (0.29) 0.34 (0.31) Year1990 0.01 (0.31) 0.07 (0.29) 0.01 (0.31) Year1991 −0.16 (0.35) −0.18 (0.35) −0.16 (0.35) Year1992 0.24 (0.31) 0.26 (0.30) 0.24 (0.31) Year1993 0.14 (0.31) 0.15 (0.31) 0.14 (0.31) Year1994 −0.03 (0.33) −0.03 (0.31) −0.03 (0.33) Year1995 0.14 (0.33) 0.16 (0.33) 0.14 (0.33) Year1996 0.12 (0.32) 0.18 (0.32) 0.12 (0.32) Year1997 0.17 (0.32) 0.22 (0.32) 0.17 (0.32) Year1998 −0.09 (0.32) −0.09 (0.26) −0.09 (0.32) Year1999 0.03 (0.31) 0.04 (0.31) 0.03 (0.31) Year2000 0.11 (0.31) 0.12 (0.27) 0.11 (0.31) Year2001 0.27 (0.31) 0.31 (0.31) 0.27 (0.31) Year2002 0.11 (0.31) 0.15 (0.28) 0.11 (0.31) Year2003 −0.01 (0.32) −0.02 (0.34) −0.01 (0.32) Year2004 0.21 (0.31) 0.33 (0.32) 0.21 (0.31) Year2005 0.24 (0.30) 0.32 (0.30) 0.24 (0.30) Year2006 0.03 (0.31) 0.11 (0.29) 0.03 (0.31) Year2007 −0.10 (0.33) 0.06 (0.32) −0.10 (0.33) Year2008 0.07 (0.36) 0.11 (0.36) 0.07 (0.36) Year2009 0.10 (0.34) 0.19 (0.31) 0.10 (0.34) Year2010 −0.13 (0.34) −0.05 (0.35) −0.13 (0.34) Year2011 0.28 (0.34) 0.24 (0.33) 0.28 (0.34) Year2012 0.03 (0.32) 0.10 (0.32) 0.03 (0.32) Year2013 0.25 (0.33) 0.32 (0.31) 0.25 (0.33) 51 A. Sanghi and D. Johnson Pooled OLS Fixed Effects Random Effects Year2014 0.18 (0.34) 0.16 (0.33) 0.18 (0.34) Year China dummies China Year 1948 −0.45 (0.26) China Year 1956 0.62 (0.26)∗ China Year 1957 −0.34 (0.24) ∗ China Year 1959 0.55 (0.25) China Year 1960 −0.75 (0.25)∗∗ China Year 1961 0.68 (0.24)∗∗ ∗∗∗ China Year 1962 0.58 (0.17) China Year 1964 2.09 (0.25)∗∗∗ China Year 1966 −1.91 (0.26)∗∗∗ ∗∗∗ China Year 1967 0.93 (0.19) China Year 1968 −4.49 (0.22)∗∗∗ China Year 1969 0.71 (0.19)∗∗∗ China Year 1970 −0.31 (0.20) China Year 1971 −0.59 (0.22)∗∗ China Year 1972 −1.93 (0.24)∗∗∗ ∗∗∗ China Year 1973 2.03 (0.23) ∗∗∗ China Year 1974 1.04 (0.26) China Year 1975 −0.62 (0.20)∗∗ ∗∗∗ China Year 1976 1.67 (0.22) China Year 1977 −1.53 (0.19)∗∗∗ China Year 1978 0.08 (0.20) ∗∗∗ China Year 1979 2.19 (0.22) China Year 1981 −1.07 (0.23)∗∗∗ China Year 1982 −2.46 (0.24)∗∗∗ China Year 1983 −1.60 (0.26)∗∗∗ China Year 1984 0.00 (0.22) China Year 1985 1.14 (0.17)∗∗∗ China Year 1986 0.87 (0.19)∗∗∗ ∗∗∗ China Year 1987 1.84 (0.21) China Year 1988 −1.02 (0.28)∗∗∗ China Year 1989 0.23 (0.19) China Year 1990 −0.96 (0.18)∗∗∗ China Year 1991 0.75 (0.24)∗∗ China Year 1992 1.41 (0.19)∗∗∗ China Year 1993 −0.63 (0.17)∗∗∗ China Year 1994 1.02 (0.20)∗∗∗ China Year 1995 0.51 (0.20)∗∗ China Year 1996 −0.87 (0.19)∗∗∗ China Year 1997 −3.65 (0.29)∗∗∗ China Year 1998 −3.28 (0.29)∗∗∗ ∗∗∗ China Year 1999 1.14 (0.17) 52 C RESULTS Pooled OLS Fixed Effects Random Effects China Year 2000 −0.02 (0.18) China Year 2001 0.00 (0.18) China Year 2003 −0.35 (0.19) China Year 2004 −0.70 (0.21)∗∗ China Year 2005 −2.23 (0.21)∗∗∗ China Year 2006 −1.40 (0.19)∗∗∗ China Year 2009 −2.73 (0.20)∗∗∗ China Year 2010 −0.07 (0.27) China Year 2011 −0.28 (0.19) China Year 2012 0.01 (0.23) China Year 2013 −0.41 (0.18)∗ ∗∗∗ China Year 2014 1.42 (0.19) R2 0.63 0.40 0.63 Adj. R2 0.62 0.39 0.62 Num. obs. 6917 6917 6917 ∗∗∗ p < 0.001, ∗∗ p < 0.01, ∗ p < 0.05. Standard errors in parentheses. Table C6: Estimates of China’s Trade Flows (1948-2014). Dependent variable: China’s exports Pooled OLS Fixed Effects Random Effects Constant −7.78 (1.31)∗∗∗ −7.78 (1.31)∗∗∗ ∗∗∗ China GDP per capita (Log) 0.90 (0.06) 0.90 (0.06)∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ Importer GDP per capita (Log) 0.75 (0.01) 0.74 (0.01) 0.75 (0.01) Population China (Log) 0.40 (0.23) 0.40 (0.23) Importer Population (Log) 0.80 (0.01)∗∗∗ 0.79 (0.01)∗∗∗ 0.80 (0.01)∗∗∗ Distance (log) −0.42 (0.04)∗∗∗ −0.42 (0.04)∗∗∗ −0.42 (0.04)∗∗∗ Regional Trade Agreement 0.76 (0.20)∗∗∗ 0.81 (0.21)∗∗∗ 0.76 (0.20)∗∗∗ Contiguous 0.31 (0.10)∗∗ 0.29 (0.08)∗∗∗ 0.31 (0.10)∗∗ ∗∗∗ ∗∗∗ ∗∗∗ Common Language 1.40 (0.10) 1.40 (0.09) 1.40 (0.10) GATT China 0.44 (0.06)∗∗∗ −0.47 (0.28) 0.44 (0.06)∗∗∗ ∗∗∗ ∗∗∗ GATT importer 0.14 (0.04) 0.15 (0.04) 0.14 (0.04)∗∗∗ Colonial History 0.22 (0.19) 0.19 (0.27) 0.22 (0.19) ∗∗ Kenya dummy 0.26 (0.10) 0.26 (0.21) 0.26 (0.10)∗∗ Year dummies Year1949 −0.30 (0.25) −0.29 (0.26) −0.30 (0.25) Year1950 −0.13 (0.25) −0.12 (0.25) −0.13 (0.25) Year1951 −0.15 (0.23) −0.15 (0.25) −0.15 (0.23) Year1952 −0.37 (0.24) −0.37 (0.25) −0.37 (0.24) Year1953 −0.28 (0.26) −0.27 (0.25) −0.28 (0.26) Year1954 −0.24 (0.23) −0.24 (0.25) −0.24 (0.23) Year1955 −0.27 (0.22) −0.27 (0.24) −0.27 (0.22) 53 A. Sanghi and D. Johnson Pooled OLS Fixed Effects Random Effects Year1956 −0.46 (0.25) −0.45 (0.24) −0.46 (0.25) Year1957 −0.19 (0.20) −0.19 (0.24) −0.19 (0.20) Year1958 −0.37 (0.25) −0.36 (0.25) −0.37 (0.25) Year1959 −0.30 (0.23) −0.28 (0.24) −0.30 (0.23) Year1960 −0.35 (0.22) −0.35 (0.24) −0.35 (0.22) Year1961 −0.33 (0.23) −0.29 (0.24) −0.33 (0.23) Year1962 −0.45 (0.22)∗ −0.43 (0.23) −0.45 (0.22)∗ Year1963 −0.74 (0.24)∗∗ −0.72 (0.23)∗∗ −0.74 (0.24)∗∗ Year1964 −0.29 (0.22) −0.28 (0.23) −0.29 (0.22) Year1965 −0.26 (0.22) −0.19 (0.23) −0.26 (0.22) Year1966 −0.35 (0.21) −0.32 (0.23) −0.35 (0.21) Year1967 −0.63 (0.22)∗∗ −0.60 (0.23)∗∗ −0.63 (0.22)∗∗ Year1968 −0.26 (0.21) −0.22 (0.22) −0.26 (0.21) Year1969 −0.34 (0.20) −0.31 (0.23) −0.34 (0.20) Year1970 −0.37 (0.21) −0.35 (0.22) −0.37 (0.21) Year1971 −0.28 (0.22) −0.27 (0.23) −0.28 (0.22) Year1972 −0.29 (0.20) −0.29 (0.22) −0.29 (0.20) Year1973 −0.36 (0.20) −0.35 (0.22) −0.36 (0.20) Year1974 −0.41 (0.22) −0.41 (0.22) −0.41 (0.22) Year1975 −0.30 (0.21) −0.28 (0.22) −0.30 (0.21) Year1976 −0.25 (0.20) −0.25 (0.22) −0.25 (0.20) Year1977 −0.44 (0.21)∗ −0.42 (0.22) −0.44 (0.21)∗ Year1978 −0.35 (0.20) −0.34 (0.22) −0.35 (0.20) Year1979 −0.48 (0.21)∗ −0.45 (0.22)∗ −0.48 (0.21)∗ Year1980 −0.15 (0.21) −0.11 (0.22) −0.15 (0.21) Year1981 −0.36 (0.21) −0.36 (0.22) −0.36 (0.21) Year1982 −0.08 (0.20) −0.09 (0.22) −0.08 (0.20) Year1983 −0.41 (0.20)∗ −0.43 (0.22)∗ −0.41 (0.20)∗ Year1984 −0.55 (0.21)∗∗ −0.55 (0.22)∗ −0.55 (0.21)∗∗ Year1985 −0.53 (0.20)∗∗ −0.55 (0.22)∗ −0.53 (0.20)∗∗ Year1986 −0.33 (0.20) −0.33 (0.22) −0.33 (0.20) Year1987 −0.54 (0.19)∗∗ −0.53 (0.22)∗ −0.54 (0.19)∗∗ Year1988 −0.23 (0.19) −0.17 (0.22) −0.23 (0.19) Year1989 −0.39 (0.20)∗ −0.38 (0.21) −0.39 (0.20)∗ Year1990 −0.42 (0.20)∗ −0.41 (0.22) −0.42 (0.20)∗ Year1991 −0.30 (0.19) −0.30 (0.21) −0.30 (0.19) Year1992 −0.47 (0.20)∗ −0.46 (0.21)∗ −0.47 (0.20)∗ Year1993 −0.61 (0.20)∗∗ −0.60 (0.21)∗∗ −0.61 (0.20)∗∗ Year1994 −0.60 (0.20)∗∗ −0.59 (0.21)∗∗ −0.60 (0.20)∗∗ Year1995 −0.39 (0.19)∗ −0.38 (0.21) −0.39 (0.19)∗ Year1996 −0.33 (0.20) −0.32 (0.21) −0.33 (0.20) Year1997 −0.32 (0.20) −0.34 (0.21) −0.32 (0.20) Year1998 −0.22 (0.19) −0.24 (0.21) −0.22 (0.19) 54 C RESULTS Pooled OLS Fixed Effects Random Effects Year1999 −0.55 (0.19)∗∗ −0.56 (0.21)∗∗ −0.55 (0.19)∗∗ Year2000 −0.33 (0.19) −0.34 (0.21) −0.33 (0.19) Year2001 −0.45 (0.21)∗ −0.46 (0.21)∗ −0.45 (0.21)∗ Year2002 −0.56 (0.20)∗∗ −0.55 (0.21)∗∗ −0.56 (0.20)∗∗ Year2003 −0.48 (0.19)∗ −0.47 (0.21)∗ −0.48 (0.19)∗ Year2004 −0.48 (0.19)∗ −0.51 (0.21)∗ −0.48 (0.19)∗ Year2005 −0.26 (0.19) −0.27 (0.21) −0.26 (0.19) Year2006 −0.32 (0.19) −0.32 (0.21) −0.32 (0.19) Year2007 −7.59 (0.25)∗∗∗ −0.00 (0.20) −7.59 (0.25)∗∗∗ Year2008 −7.48 (0.23)∗∗∗ 0.09 (0.17) −7.48 (0.23)∗∗∗ Year2009 −7.47 (0.22)∗∗∗ 0.11 (0.17) −7.47 (0.22)∗∗∗ Year2010 −7.39 (0.23)∗∗∗ 0.15 (0.16) −7.39 (0.23)∗∗∗ Year2011 −7.28 (0.23)∗∗∗ 0.27 (0.16) −7.28 (0.23)∗∗∗ Year2012 −7.52 (0.23)∗∗∗ −7.52 (0.23)∗∗∗ Year2013 −7.55 (0.23)∗∗∗ −7.55 (0.23)∗∗∗ Year2014 −7.62 (0.22)∗∗∗ −7.62 (0.22)∗∗∗ Kenya Year dummies Kenya Year 1960 0.73 (1.49) Kenya Year 1961 −0.65 (1.49) Kenya Year 1962 0.66 (1.49) Kenya Year 1963 1.87 (1.49) Kenya Year 1964 2.00 (1.48) Kenya Year 1965 −3.16 (1.49)∗ Kenya Year 1966 −1.63 (1.48) Kenya Year 1967 1.40 (1.48) Kenya Year 1968 −0.94 (1.48) Kenya Year 1969 0.23 (1.48) Kenya Year 1970 −1.55 (1.48) Kenya Year 1971 0.06 (1.48) Kenya Year 1972 0.08 (1.48) Kenya Year 1973 −0.57 (1.48) Kenya Year 1974 1.63 (1.48) Kenya Year 1975 −0.30 (1.48) Kenya Year 1976 0.84 (1.48) Kenya Year 1977 −2.91 (1.48)∗ Kenya Year 1978 0.51 (1.48) Kenya Year 1979 −1.24 (1.48) Kenya Year 1980 −2.75 (1.48) Kenya Year 1981 1.18 (1.48) Kenya Year 1982 0.55 (1.48) Kenya Year 1983 1.34 (1.48) Kenya Year 1984 0.74 (1.48) Kenya Year 1985 1.52 (1.48) 55 A. Sanghi and D. Johnson Pooled OLS Fixed Effects Random Effects Kenya Year 1986 0.37 (1.48) Kenya Year 1987 0.58 (1.48) Kenya Year 1988 −5.31 (1.48)∗∗∗ Kenya Year 1989 0.75 (1.48) Kenya Year 1990 −1.04 (1.48) Kenya Year 1991 1.29 (1.48) Kenya Year 1992 −0.19 (1.48) Kenya Year 1993 −1.07 (1.48) Kenya Year 1994 −0.54 (1.48) Kenya Year 1995 1.42 (1.48) Kenya Year 1996 0.12 (1.48) Kenya Year 1998 −0.48 (1.48) Kenya Year 1999 0.32 (1.48) Kenya Year 2000 −0.48 (1.48) Kenya Year 2002 −2.12 (1.48) Kenya Year 2004 2.70 (1.48) Kenya Year 2005 0.70 (1.48) Kenya Year 2008 1.27 (1.48) Kenya Year 2009 −0.05 (1.48) Kenya Year 2010 0.68 (1.48) Kenya Year 2013 1.04 (1.48) R2 0.74 0.62 0.74 Adj. R2 0.74 0.61 0.74 Num. obs. 8207 8207 8207 ∗∗∗ p < 0.001, ∗∗ p < 0.01, ∗ p < 0.05. Standard errors in parentheses. Table C7: Estimates of China’s Trade Flows (1948-2014). Dependent variable: China’s exports Poisson Quasi Maximum Likelihood Constant 8.82 (15.22) China GDP per capita 0.00 (0.00) ∗∗∗ Importer GDP per capita 0.00 (0.00) Population China −0.01 (0.03) Population Importer 0.00 (0.00)∗∗∗ Distance −0.00 (0.00)∗∗∗ Contiguous 0.09 (0.06) Common Language 1.94 (0.06)∗∗∗ GATT China 0.90 (18.60) ∗∗∗ GATT importer 0.79 (0.08) Colonial History −1.45 (0.91) Kenya dummy 4.42 (32.00) 56 C RESULTS Poisson Quasi Maximum Likelihood Year dummies Year 1950 0.08 (4.29) Year 1951 0.02 (4.44) Year 1952 −0.39 (5.11) Year 1953 0.11 (4.63) Year 1954 0.06 (4.94) Year 1955 1.99 (3.68) Year 1956 2.34 (3.73) Year 1957 2.54 (3.91) Year 1958 1.61 (4.56) Year 1959 1.70 (4.73) Year 1960 1.90 (4.47) Year 1961 1.70 (4.45) Year 1962 1.79 (4.51) Year 1963 2.07 (4.70) Year 1964 2.61 (4.87) Year 1965 3.06 (5.11) Year 1966 3.47 (5.47) Year 1967 3.57 (5.90) Year 1968 3.82 (6.35) Year 1969 4.18 (6.80) Year 1970 4.44 (7.28) Year 1971 4.87 (7.79) Year 1972 5.33 (8.25) Year 1973 5.97 (8.64) Year 1974 6.41 (9.08) Year 1975 6.67 (9.42) Year 1976 6.89 (9.82) Year 1977 7.12 (10.08) Year 1978 7.64 (10.47) Year 1979 8.02 (10.72) Year 1980 8.37 (10.99) Year 1981 8.68 (11.30) Year 1982 8.90 (11.66) Year 1983 9.12 (11.97) Year 1984 9.35 (12.24) Year 1985 9.62 (12.48) Year 1986 9.89 (12.92) Year 1987 10.27 (13.45) Year 1988 10.61 (13.81) Year 1989 10.91 (14.17) Year 1990 11.18 (14.58) Year 1991 11.50 (14.94) 57 A. Sanghi and D. Johnson Poisson Quasi Maximum Likelihood Year 1992 11.74 (15.21) Year 1993 12.02 (15.52) Year 1994 12.16 (15.58) Year 1995 12.23 (15.53) Year 1996 12.39 (15.57) Year 1997 12.65 (15.68) Year 1998 12.87 (15.86) Year 1999 13.05 (16.04) Year 2000 13.27 (16.02) Year 2001 12.50 (24.63) Year 2002 12.59 (24.60) Year 2003 12.60 (24.47) Year 2004 12.52 (24.19) Year 2005 12.39 (23.90) Year 2006 12.32 (23.45) Year 2007 3.92 (12.51) Year 2008 2.86 (10.44) Year 2009 3.46 (9.59) Year 2010 2.86 (7.67) Year 2011 1.39 (4.81) Year 2012 1.27 (3.08) Kenya Year dummies Kenya Year 1960 −8.04 (150.60) Kenya Year 1961 −15.92 (4873.15) Kenya Year 1962 −7.93 (150.60) Kenya Year 1963 −6.89 (90.80) Kenya Year 1964 −6.12 (45.92) Kenya Year 1965 −6.06 (42.75) Kenya Year 1966 −5.51 (37.77) Kenya Year 1967 −6.12 (43.49) Kenya Year 1968 −5.51 (38.81) Kenya Year 1969 −5.93 (41.19) Kenya Year 1970 −5.85 (40.78) Kenya Year 1971 −5.81 (39.55) Kenya Year 1972 −6.17 (40.77) Kenya Year 1973 −6.32 (38.99) Kenya Year 1974 −5.53 (34.77) Kenya Year 1975 −6.76 (40.22) Kenya Year 1976 −6.09 (36.34) Kenya Year 1977 −5.51 (34.32) Kenya Year 1978 −6.13 (35.06) Kenya Year 1979 −6.17 (34.53) Kenya Year 1980 −5.96 (33.70) 58 C RESULTS Poisson Quasi Maximum Likelihood Kenya Year 1981 −6.18 (33.83) Kenya Year 1982 −6.40 (34.22) Kenya Year 1983 −6.85 (35.29) Kenya Year 1984 −6.56 (34.30) Kenya Year 1985 −6.58 (34.11) Kenya Year 1986 −6.48 (33.81) Kenya Year 1987 −6.17 (33.18) Kenya Year 1988 −6.51 (33.45) Kenya Year 1989 −6.56 (33.38) Kenya Year 1990 −6.65 (33.43) Kenya Year 1991 −6.58 (33.18) Kenya Year 1992 −6.53 (33.04) Kenya Year 1993 −6.15 (32.64) Kenya Year 1994 −5.89 (32.47) Kenya Year 1995 −5.42 (32.29) Kenya Year 1996 −5.62 (32.33) Kenya Year 1997 −5.42 (32.23) Kenya Year 1998 −5.63 (32.26) Kenya Year 1999 −5.85 (32.30) Kenya Year 2000 −5.74 (32.23) Kenya Year 2001 −5.79 (32.22) Kenya Year 2002 −5.57 (32.17) Kenya Year 2003 −5.32 (32.13) Kenya Year 2004 −4.94 (32.09) Kenya Year 2005 −4.64 (32.07) Kenya Year 2006 −4.42 (32.05) Kenya Year 2008 −1.15 (57.87) Kenya Year 2009 −1.92 (57.18) Kenya Year 2010 −1.38 (49.47) AIC BIC Log Likelihood Deviance 13045625.79 Num. obs. 8215 ∗∗∗ p < 0.001, ∗∗ p < 0.01, ∗ p < 0.05. Standard errors in parentheses. Table C8: Estimates of Kenya’s Trade Flows (1948-2014). Dependent variable: Kenya’s exports Poisson Quasi Maximum Likelihood Constant −0.93 (0.41)∗ Kenya GDP per capita 0.00 (0.00)∗∗∗ 59 A. Sanghi and D. Johnson Poisson Quasi Maximum Likelihood Importer GDP per capita 0.00 (0.00)∗∗∗ Population Kenya −0.01 (0.01) ∗∗∗ Population Importer 0.00 (0.00) Distance −0.00 (0.00)∗∗∗ Regional Trade Agreement 1.16 (0.10)∗∗∗ ∗∗∗ Contiguous 2.30 (0.09) ∗∗ Common Language 0.19 (0.06) GATT Kenya 0.78 (0.23)∗∗∗ ∗∗∗ GATT importer 1.50 (0.09) Colonial History 2.53 (0.08)∗∗∗ China dummy −1.74 (0.42)∗∗∗ Year dummies Year 1949 −0.24 (0.48) Year 1950 0.51 (0.43) Year 1951 0.05 (0.47) Year 1952 −0.03 (0.49) Year 1953 0.56 (0.40) Year 1954 −0.06 (0.47) Year 1955 −0.31 (0.44) Year 1956 0.49 (0.40) Year 1957 0.41 (0.41) Year 1958 0.29 (0.41) Year 1959 0.45 (0.38) Year 1960 −0.21 (0.40) Year 1961 0.13 (0.40) Year 1962 0.38 (0.39) Year 1963 0.36 (0.37) Year 1964 0.17 (0.36) Year 1965 0.88 (0.38)∗ Year 1966 0.38 (0.38) Year 1967 −0.34 (0.39) Year 1968 0.17 (0.38) Year 1969 0.30 (0.40) Year 1970 0.35 (0.38) Year 1971 0.51 (0.39) Year 1972 0.41 (0.39) Year 1973 0.29 (0.37) Year 1974 0.43 (0.39) Year 1975 0.39 (0.38) Year 1976 0.54 (0.37) Year 1977 0.05 (0.39) Year 1978 0.26 (0.39) Year 1979 −0.41 (0.43) 60 C RESULTS Poisson Quasi Maximum Likelihood Year 1980 0.03 (0.38) Year 1981 0.21 (0.39) Year 1982 0.50 (0.37) Year 1983 0.12 (0.38) Year 1984 0.54 (0.37) Year 1985 0.64 (0.36) Year 1986 0.38 (0.36) Year 1987 0.39 (0.36) Year 1988 −0.06 (0.41) Year 1989 0.32 (0.39) Year 1990 0.24 (0.39) Year 1991 0.07 (0.38) Year 1992 0.15 (0.40) Year 1993 0.28 (0.36) Year 1994 −0.31 (0.40) Year 1995 0.29 (0.37) Year 1996 0.51 (0.36) Year 1997 0.30 (0.36) Year 1998 0.25 (0.36) Year 1999 0.15 (0.36) Year 2000 0.09 (0.36) Year 2001 0.36 (0.36) Year 2002 0.30 (0.37) Year 2003 0.37 (0.35) Year 2004 0.33 (0.37) Year 2005 0.24 (0.37) Year 2006 0.54 (0.37) Year 2007 −0.13 (0.44) Year 2008 0.09 (0.41) Year 2009 −0.10 (0.39) Year 2010 0.47 (0.39) Year 2011 0.86 (0.36)∗ ∗∗ Year 2012 1.07 (0.36) Year 2013 0.38 (0.40) Year 2014 0.36 (0.40) China Year dummies China Year 1948 −13.71 (1603.39) China Year 1955 −12.41 (1603.39) China Year 1956 −13.42 (1603.39) China Year 1957 −0.89 (8.50) China Year 1960 −16.02 (1603.39) China Year 1961 0.70 (4.61) China Year 1962 −0.66 (7.97) 61 A. Sanghi and D. Johnson Poisson Quasi Maximum Likelihood China Year 1963 −12.09 (1603.39) China Year 1964 −1.99 (17.81) China Year 1965 −14.40 (1603.39) China Year 1967 −11.30 (1603.39) China Year 1968 −13.54 (1603.39) China Year 1969 −12.19 (1603.39) China Year 1970 −11.42 (1603.39) China Year 1971 −13.86 (1603.39) China Year 1972 −13.98 (1603.39) China Year 1973 −12.07 (1603.39) China Year 1974 0.52 (3.80) China Year 1975 1.58 (2.79) China Year 1976 −11.95 (1603.39) China Year 1977 −1.00 (5.37) China Year 1978 −0.45 (3.06) China Year 1979 −1.85 (8.50) China Year 1980 −13.56 (1603.39) China Year 1981 −3.97 (17.81) China Year 1982 0.23 (4.91) China Year 1984 −0.00 (5.37) China Year 1985 −1.41 (6.01) China Year 1987 −0.94 (4.60) China Year 1989 −11.57 (1603.39) China Year 1990 −0.25 (3.43) China Year 1991 −13.27 (1603.39) China Year 1992 −13.60 (1603.39) China Year 1993 −11.43 (1603.39) China Year 1994 −11.52 (1603.39) China Year 1995 1.76 (1.12) China Year 1996 −1.24 (2.80) China Year 1997 −11.76 (1603.39) China Year 1998 −11.58 (1603.39) China Year 1999 0.32 (2.66) China Year 2000 −12.98 (1603.39) China Year 2001 −0.79 (6.94) China Year 2003 −0.09 (3.27) China Year 2004 −11.94 (1603.39) China Year 2006 −11.71 (1603.39) China Year 2007 0.13 (0.88) China Year 2008 −1.85 (16.99) China Year 2009 −11.00 (1603.39) China Year 2010 −11.54 (1603.39) China Year 2011 −12.23 (1603.39) 62 C RESULTS Poisson Quasi Maximum Likelihood China Year 2012 −2.53 (17.81) China Year 2013 −2.22 (16.98) AIC BIC Log Likelihood Deviance 98819.06 Num. obs. 6118 ∗∗∗ p < 0.001, ∗∗ p < 0.01, ∗ p < 0.05. Standard errors in parentheses. 0 Trade/GDP (Log) −10 −20 7 8 9 Distance (Log) Figure C36: Trade and Distance Kenya and Rest of the World (1948-2014) Source: CEPII Gravity Database 2010 and UN Comtrade 2015 0 Trade/GDP (Log) −5 −10 −15 6 7 8 9 10 Distance (Log) 63 A. Sanghi and D. Johnson 10 Trade (Log) 0 −10 −20 3 6 9 12 Importer GDP per capita (Log) Figure C37: Trade and Importer GDP per capita Kenya and Rest of the World (1948-2014) Source: CEPII Gravity Database 2010 and UN Comtrade 2015 K U 10 na hi C Trade (Log) 5 0 6 8 10 Importer GDP per capita (Log) Figure C38: Trade and Importer GDP per capita Kenya and Rest of the World (2014) Source: CEPII Gravity Database 2010 and UN Comtrade 2015 64 C RESULTS 6 3 Trade (Log) a ny Ke 0 −3 −6 6 8 10 Importer GDP per capita (Log) 65