Dissertation
Regional Economic Growth and International Capital Flows The Case of Ukraine
Olena Bogdan
This document was submitted as a dissertation in September 2017 in partial fulfillment of the requirements of the doctoral degree in public policy analysis at the Pardee RAND Graduate School. The faculty committee that supervised and approved the dissertation consisted of Charles Wolf, Jr. (Chair), James Hosek (Chair), F. Stephen Larrabee, and Edward Keating. Svitlana Kobzar of Vesalius College in Brussels, Belgium, was the external reader for the dissertation.
PARDEE RAND GRADUATE SCHOOL For more information on this publication, visit http://www.rand.org/pubs/rgs_dissertations/RGSD403.html
Published 2018 by the RAND Corporation, Santa Monica, Calif. R® is a registered trademark
Limited Print and Electronic Distribution Rights This document and trademark(s) contained herein are protected by law. This representation of RAND intellectual property is provided for noncommercial use only. Unauthorized posting of this publication online is prohibited. Permission is given to duplicate this document for personal use only, as long as it is unaltered and complete. Permission is required from RAND to reproduce, or reuse in another form, any of its research documents for commercial use. For information on reprint and linking permissions, please visit www.rand.org/pubs/permissions.html. The RAND Corporation is a research organization that develops solutions to public policy challenges to help make communities throughout the world safer and more secure, healthier and more prosperous. RAND is nonprofit, nonpartisan, and committed to the public interest. RAND’s publications do not necessarily reflect the opinions of its research clients and sponsors.
Support RAND Make a tax-deductible charitable contribution at www.rand.org/giving/contribute
www.rand.org
In eternal memory of Dr. Charles Wolf, Jr., my mentor and the original Chair of my dissertation committee.
iii
In many countries, poorer regions are growing faster than richer regions, and provinces are converging over time. In other countries, however, regional economic differences have persisted over many years and income gaps are widening. It is important to understand why enduring regional socio-economic disparities exist within countries, what factors drive them, and what can be done to reconcile these differences. Recently, a policy community in Europe, Asia and the United States has become more interested in finding answers to these complex questions. This research examines divergent economic paths between regions in several countries. Based on the case studies of China, Italy, and the United States, I find that regional economic differences are common, even as countries have unlike history, culture, and socio-economic systems. Regional disparities are primarily attributed to local geography, human and social capital, institutional quality, infrastructure, public policies, and clustering of economic activities. Across all case studies, the most prosperous regions are also those with the largest international capital inflows. Because prosperous localities amass thriving businesses, next I focus on the role of foreign capital in regional economies. Using Ukraine as a case, I analyze data on its local socio- economic indicators and international capital inflows of various types between 2001 and 2013 and find not only substantial regional differences, but also a systematic, simultaneous, positive relationship between regional economic growth and international capital flows. This nexus is particularly distinct for Ukraine’s West where, if compared to other regions, any small increase in international capital, either public or private, would create a more than twice larger positive effect on local economic development. Based on my analysis, I recommend specific steps for Ukraine’s regional and overall economic policy.
iv
Abstract ...... v Figures...... viii Tables ...... x Acknowledgments ...... xii Abbreviations ...... xiii
Chapter 1. Introduction ...... 1 1.1. Motivation and Objectives ...... 1 1.2. Hypothesis and Research Questions ...... 3 1.3. Outline of Dissertation ...... 3 Chapter 2. Background: Ukraine ...... 4 Chapter 3. Overview of Existing Studies ...... 10 3.1. Growth Modeling Literature ...... 10 3.2. Factors Behind Economic Growth ...... 11 3.3. Literature on Differential Regional Growth ...... 16 Summary ...... 17 Chapter 4. Case Studies of Countries with Differential Regional Growth ...... 19 4.1. China ...... 19 4.2. Italy ...... 30 4.3. United States ...... 40 Summary ...... 50 Chapter 5. Ukraine's Regional Data ...... 52 5.1. Socio-economic Indicators ...... 52 5.2. International Capital Flows ...... 60 Summary ...... 73 Chapter 6. Analysis and Results ...... 75 6.1. Model of Regional Economic Growth and International Capital Flows ...... 75 6.2. Empirical Approach and Results ...... 78 6.3. Supplemental Analyses ...... 88 Summary ...... 92 Chapter 7. Conclusion and Policy Implications ...... 94 7.1 Key Findings ...... 94 7.2. Policy Implications: Ukraine ...... 97 7.3. Overall Policy Implications ...... 99 Future Research ...... 101
v Appendix A: Case Studies Maps ...... 102 Appendix B: Oblast Variations in Socio-Economic Indicators ...... 105 Appendix C: Detailed Model Estimates ...... 106
References ...... 137
vi
2.1. Administrative Map of Ukraine ...... 4 2.2. Pre-war Ukraine’s Regional Economic Growth ...... 7 2.3. Pre-war FDI Stock (Equity) in Ukrainian Regions, Percentage ...... 7 2.4. Pre-war Cumulative FDI Stock (Equity) in Ukraine, by Country of Origin ...... 8 2.5. Pre-war Cumulative FDI (Equity) in Ukraine, by Targeted Sector ...... 8 4.1.1. Map of Chinese Regions ...... 19 4.1.2. Per Capita GDP across Chinese Regions in 2001-2013, RMB ...... 20 4.1.3. Per Capita GDP across Chinese Regions and Provinces in 2013, RMB ...... 21 4.1.4. Per Capita Central Government Investment across Chinese Regions and Provinces in 2008, RMB ...... 25 4.1.5. China’s Central Government Expenditure in 2013 ...... 26 4.1.6. China’s Local Government Expenditure in 2013 ...... 26 4.1.7. Stock of Total Investment by Foreign Funded Enterprises across Chinese Regions and Provinces in 2013, million USD ...... 29 4.2.1. Map of Italian Regions ...... 31 4.2.2. Per Capita GDP across Italian Regions, 2001-09, EUR ...... 32 4.2.3. Per Capita GDP across Italian Regions in 2010, EUR ...... 33 4.2.4. Per Capita Local Government Expenditure across Italian Regions in 2011, EUR ...... 36 4.2.5. Italy’s Central Government Expenditure in 2013 ...... 37 4.2.6. Italy’s Local Government Expenditure in 2013 ...... 37 4.2.7. Net FDI Inflows across Italian Regions in 2011, Millions EUR ...... 38 4.2.8. Percentage of Companies with Foreign Participation in Total Workforce across Italian Regions in 2009 ...... 39 4.3.1. Map of U.S. Regions ...... 41 4.3.2. Per capita GDP across U.S. Regions, USD ...... 42 4.3.3. Per Capita GDP across U.S. States and Regions in 2013, USD ...... 43 4.3.4. U.S. States with ‘Right-To-Work’ Laws ...... 45 4.3.5. U.S. Federal Government Expenditure in 2013 ...... 46 4.3.6. U.S. State Government Expenditure in 2013 ...... 47 4.3.7. Per Capita Government Expenditure across U.S. States and Regions in 2013, USD ...... 47 4.3.8. FDI across U.S. States and Regions in 2006, Million USD ...... 49 4.3.9. Employment by FOEs across U.S. States and Regions in 2011 ...... 49 5.1.1. Per Capita GDP across Ukrainian Regions in 2001-2013, USD ...... 57 5.1.2. Per Capita GDP across Ukrainian Regions and Oblasts in 2013, USD ...... 58 5.1.3. GDP Growth across Ukrainian Regions and Oblasts, 2012-2013 ...... 59
vii 5.2.1. Total Stock of FDI in Equity and Debt across Ukrainian Regions, 2013, million USD .... 62 5.2.2. Stock of Per Capita FDI as Equity across Ukrainian Regions and Oblasts, 2013, USD .... 63 5.2.3. Stock of Per Capita FDI as Debt across Ukrainian Regions and Oblasts, 2013, USD ...... 64 5.2.4. Total Cumulative Foreign and Domestic M&A Investment across Ukrainian Regions, 2001-2013, million USD ...... 66 5.2.5. Cumulative Foreign M&A Investment in non-Financial Sectors Per Capita across Ukrainian Regions and Oblasts, 2001-2013, USD ...... 68 5.2.6. Cumulative EBRD and NED Investment across Ukrainian Regions, 2001-2013, million USD ...... 70 5.2.7. Cumulative EBRD Investment in non-Financial Sectors Per Capita across Ukrainian Regions and Oblasts, 2001-2013, USD ...... 72 5.2.8. Cumulative NED Investment Per Capita across Ukrainian Regions and Oblasts, 2001- 2013, USD ...... 72
A1. Map of Chinese Provinces ...... 102 A2. Map of Italian Provinces ...... 103 A3. Map of the United States of America ...... 104
viii
5.1.1. Key Socio-Economic Indicators across Ukrainian Regions, 2001-2013 ...... 55 5.2.1 Average Annual FDI Inflows across Ukrainian Regions, 2001-2013 ...... 62 5.2.2 Average Annual M&A Inflows across Ukrainian Regions, 2001-2013 ...... 67 5.2.3 Average Annual International Public Capital Inflows across Ukrainian Regions, 2001-2013 ...... 71 6.2.1. Elasticities of Ukraine’s Regional GDP Input Factors ...... 79 6.2.2. Ukraine’s Regional Economic Growth Factors ...... 82 6.2.3. Effect of International Capital Inflows on Ukraine’s Regional Economic Growth (by Type) ...... 86 6.2.4. Effect of Ukraine’s Regional Economic Growth on International Capital Inflows (by Type) ...... 86 6.3.1. Effect of International Capital Inflows on Ukraine’s Regional Economic Growth (by Type) over Medium Term ...... 90 6.3.2. Effect of Ukraine’s Regional Economic Growth on International Capital Inflows (by Type) over Medium Term ...... 91
B1. Ukraine’s Oblast Variations in Socio-Economic Indicators Overall and by Region, 2001- 2013...... 105 C1. Pairwise Comparisons of Marginal Linear Predictions for Regional Variables in Model (4) in Table 6.2.1 (Elasticities of Regional GDP Input Factors) ...... 106 C2a. Pairwise Comparisons of Marginal Linear Predictions for Regional Variables in Model (1) in Table 6.2.2 (Regional Economic Growth Factors) ...... 106 C2b. Pairwise Comparisons of Marginal Linear Predictions for Regional Variables in Model (2) in Table 6.2.2 (Regional Economic Growth Factors) ...... 106 C3. Regional Economic Growth and International Capital Inflows (by Type) – SEM Results . 107 C3a. Pairwise Comparisons of Marginal Linear Predictions for Regional Variables – FDI (Equity) Model ...... 109 C3b. Pairwise Comparisons of Marginal Linear Predictions for Regional Variables – FDI (Debt) Model ...... 109 C3c. Pairwise Comparisons of Marginal Linear Predictions for Regional Variables – FDI (Equity & Debt) Model ...... 110 C3d. Pairwise Comparisons of Marginal Linear Predictions for Regional Variables – M&A Capital Model ...... 110 C3e. Pairwise Comparisons of Marginal Linear Predictions for Regional Variables – EBRD Capital Model ...... 111
ix C3f. Pairwise Comparisons of Marginal Linear Predictions for Regional Variables – NED Capital Model ...... 111 C4. Regional Economic Growth and FDI (Equity) ...... 112 C5. Regional Economic Growth and FDI (Debt) ...... 114 C6. Regional Economic Growth and FDI (Equity and Debt) ...... 116 C7. Regional Economic Growth and M&A Capital ...... 118 C8. Regional Economic Growth and EBRD Capital ...... 120 C9. Regional Economic Growth and NED Capital ...... 122 C10. Regional Economic Growth and International Capital Inflows (by Type) – Medium Term ...... 124 C10a. Pairwise Comparisons of Marginal Linear Predictions for Regional Variables – FDI (Equity) Model ...... 127 C10b. Pairwise Comparisons of Marginal Linear Predictions for Regional Variables – FDI (Debt) Model ...... 128 C10c. Pairwise Comparisons of Marginal Linear Predictions for Regional Variables – FDI (Equity & Debt) Model ...... 129 C10d. Pairwise Comparisons of Marginal Linear Predictions for Regional Variables – M&A Capital Model ...... 131 C10e. Pairwise Comparisons of Marginal Linear Predictions for Regional Variables – EBRD Capital Model ...... 132 C10f. Pairwise Comparisons of Marginal Linear Predictions for Regional Variables – NED Capital Model ...... 133 C11. Regional Economic Growth and International Capital Inflows (by Type) – Urbanization 135
x
I would like to thank for the support and guidance of many people who helped me successfully complete this document. My special appreciation is for my dissertation committee – James Hosek (Chair), F. Stephen Larrabee and Edward Keating. I am deeply grateful for their mentorship, valuable feedback, kindness and continuous encouragement. Their thoughtful reflections on this study constantly pushed me to think more deeply and broadly about my research questions. In particular, I appreciate the mentorship of my committee Chair, James Hosek, who guided me throughout this endeavor, provided his intellectual insights, and supported in the most challenging moments. Additionally, my outside reader, Svitlana Kobzar, provided her valuable comments and perspective. I will forever be grateful for the chance to learn from Charles Wolf, Jr, the original Chair of my dissertation committee, and for his mentorship. This research was inspired by our lengthy conversations on Ukraine’s 2014 unrest and consequent economic and security crises. Charles, an outstanding scholar in the fields of international economics and policy analysis, taught me many professional and personal lessons. He anticipated my completion of this monograph, and it motivated me to continue working long hours on this research, even as I tremendously missed his precious guidance and encouraging smile. This study is dedicated to his eternal memory. Many individuals supported me during my doctorate program at Pardee RAND whom I would like to thank –Susan Marquis, Rachel Swanger, Fatima Ford, Nicole Eberhart, Jodi Liu, and Julia Pollak. There were several occasions when their support meant the world to me. I will also be always thankful to late Kakha Bendukidze for his care and for our fascinating Ukraine- related conversations, of which I was frequently reminded while working on this document. This research would not have been possible without the Jeremy R. Azrael and Andrew Marshall Scholarships, and generous funding from Anne and James Rothenberg. I also thank Oleksii Oleynikov from the Ukraine’s InVenture Investment Group for sharing the “Mergers and Acquisitions in Ukraine” dataset. Additionally, Barry Kolodkin (EBRD) and Brooke Stearns Lawson (USAID) provided valuable insights about their organizations’ operations in Ukraine, including directing me to publicly available data sources that I used in this study. I also very much appreciate helpful comments by Richard Neu and Rafiq Dossani. From a personal perspective, I am grateful to my parents and my brother who have continuously supported me in this endeavor. But most of all, I cannot express enough gratitude to my beloved husband, Dmytro, who has always been there for me throughout all these years. I thank God for the day when we met.
xi
xii
EBRD European Bank for Reconstruction and Development EU European Union FDI Foreign Direct Investment FOE Foreign-Owned Enterprise GDP Gross Domestic Product GRP Gross Regional Product HDI Human Development Index M&A Mergers and Acquisitions MSA Metropolitan Statistical Area NED National Endowment for Democracy SOE State-Owned Enterprise SSSU State Statistical Service of Ukraine USAID United States Agency for International Development
xiii
In many countries, poorer regions are growing faster than richer regions, and provinces are converging over time. In other countries, however, regional economic differences have persisted over many years and income gaps are widening. What drives these enduring regional differences? What can be done to reconcile these differences? Recently, a policy community has become more interested in finding answers to these complex questions and understanding fundamental blocks of regional differences. This includes policy-makers in Europe, the United States, and Asia. Many realize that the issue requires a multi-faceted approach and multi- disciplinary analysis. Policy discussion about regional differences regularly refers to historical, sociological, geographic, and identity-related factors, and less often to economic differences. But regional disparities in economic opportunities are important because any person living in any locality should have an opportunity to achieve prosperity without moving away from his/her home. Additionally, it is essential to identify the role of private sector in regional divergent paths because prosperous localities tend to accumulate thriving businesses and because in most cases business operations and capital flows reflect real economic circumstances. In Ukraine, a relatively homogenous country with few linguistic or cultural barriers, economists might expect to see regional convergence over time. Yet Ukraine is an example of a country with stark and persistent regional disparities. The four Ukrainian regions (West, Center, East, and South) have dissimilar geography and natural resources, historical experiences, local populations and voting patterns. They also differ economically in terms of the major industries they support, the types of investment they attract, and—consequently—their growth levels. These disparities are frequently highlighted and exploited by local politicians, but are seldom appreciated by observers abroad. As a result, when the Maidan Revolution ignited in late November 2013, many Western stakeholders were caught by surprise by the mass protests on the Kyiv streets. Externally it seemed like the movement was about the Association Agreement with the EU rather than a fundamental difference in developmental visions between the residents of divergent Ukrainian regions.1 Moreover, the Western policy-makers were caught off guard in spring 2014 when Crimea and other parts of the core electorate of former president Yanukovych in far eastern Ukraine became a fruitful ground for a Russian destabilization effort and easily fell into hands of Russia-orchestrated insurgency.
1 Another crucial factor, also often understated, was a popular outrage at the use of force by the government against peaceful protestors.
1 The goal of this dissertation is two-fold. First, I summarize existing knowledge on the topic of regional economic disparities within countries and identify key factors responsible for these gaps, including those that are related to the movement of international capital. Second, I explore how international capital flows are connected to differential regional economic development. I empirically investigate (1) why some regions attract more foreign investment than other regions; and (2) how regional differences in received investment influence regional economic development (and vice versa). First, I investigate three countries—China, Italy, and the United States—to examine the underlying factors behind the divergent paths of their richer and poorer regions. Second, I postulate that there is a bidirectional nexus between foreign investment and regional economies, and I assess this hypothesis. I categorize international capital flows by: (a) form of capital (e.g., direct investment as equity or debt, portfolio investment, mergers and acquisitions); (b) funding source (private capital or public capital in the form of foreign aid or funds from public financial institutions), (c) target industry/sector (e.g., agriculture, manufacturing, financial services); and (d) country of origin. I test my hypothesis empirically using data from Ukraine. My dissertation aims to highlight the importance of persistent regional economic divergences, and identify the reasons for them. Severe regional differences in socio-economic development may create vulnerabilities that could potentially be exploited either by domestic or international actors. Ultimately, by recognizing and taking into account such regional disparities, policy-makers can craft policies that promote efficient, inclusive regional growth and prevent extreme social divides. My research conclusions are particularly relevant to Ukraine. As Ukraine’s political situation starts to stabilize, the country, with its educated labor force and vibrant emerging metropolitan cities, will once again become attractive to international capital.2 However, Ukraine’s distinct regions could continue to differ substantially in their economic experiences and in the nature of the capital flows they attract. Stakeholders, such as Ukraine’s central government and international development organizations, should take these differences into account and take active steps to help the poorer regions catch up, not only to avoid regional unrest, but also to develop approaches for the future re-integration of Crimea and the occupied parts of the Eastern region into Ukraine’s economy. Since pre-existing social and political preferences may change based on economic experiences (including types of investment attracted), an effective re- integration strategy should carefully consider all the facets of the existing regional disparities. My findings suggest several key steps for Ukraine’s regional and overall economic policy.
2 James Brooke, “No Longer a Secret: Ukraine is Europe’s New Frontier Market”, Atlantic Council UkraineAlert, July 11, 2016.
2 The hypothesis of my research is that international capital flows, depending on their types and amounts, disproportionately affect intra-national regional economic growth. Additionally, I hypothesize that this relationship is bidirectional, i.e., that regional economic growth also affects international capital inflows. I examine this hypothesis by answering the following research questions:
1. What are the key factors driving a country’s economic growth (including policies specifically targeted at promoting economic growth and development), and the mechanisms through which they influence growth? What is the role of different types of capital flows? 2. What are the specific examples of countries that experience differential regional, intra- state economic growth? How does foreign investment affect the emergence or persistence of regional economic differences within these countries? 3. What has the pattern of Ukraine’s regional development been over the past decade? How do the key factors (including different types and amounts of international capital flows) that typically affect a destination country’s economic growth differ for Ukraine’s regions? 4. How can existing growth theories be used to develop a model of differential regional economic growth based on different types of international capital flows? 5. What are the policy implications (economic and political) of the differential effects of international capital flows on growth in different regions (and the simultaneous inverse effects) both for Ukraine and for other countries?
This dissertation is organized as follows. Chapter 2 provides background about Ukraine, its economy and basic regional differences. Chapter 3 discusses the existing literature on: (1) economic growth modeling; (2) factors behind economic growth; and (3) differential regional economic development. Chapter 4 investigates three case studies of countries with differential regional growth – China, Italy, and the United States. It highlights factors common to all cases and explores those related to the movement of international capital. Chapter 5 describes the data on Ukraine’s regional socio-economic development and international capital that I use for my empirical analysis. Chapter 6 details the methodological approach I use in testing my hypothesis, which is informed both by the literature review in Chapter 3 and the case studies in Chapter 4. It also reports the results of my econometric analyses. Finally, I conclude with Chapter 7 where I consider the policy implications of my research.
3
Ukraine is an Eastern-European country and the biggest geographic country entirely within Europe. It borders the EU (Romania, Hungary, Slovakia, and Poland) and Moldova to the West, Belarus and Russia to the North and East, and two seas (the Black Sea and the Azov Sea) to the South, which places it on the crossroads between Europe and Asia, as well as between the Nordic countries and Mediterranean region. Throughout its history, Ukraine’s geographic location and proximity to different neighbors has influenced the experiences of its varied regions, sometimes bringing suffering and other times prosperity, and ultimately creating the blend of cultures and ethnicities found in the country today. Administratively, Ukraine consists of 24 oblasts, an autonomous republic of Crimea, and two cities of national subordination—Kyiv and Sevastopol (see Figure 2.1). Throughout my study, I group Ukrainian oblasts into four regions: West, Center, East, and South, with Crimea considered as part of the Southern region. Occasionally, observers divide Ukraine into two spheres: East and West. However, I argue that four regions better explain Ukraine’s differences in economic growth and investment.
""" " !
4
Although Ukraine may seem to be a rather homogeneous country overall, with Ukrainian being the major ethnicity and Orthodox Christianity being the major religion, the four regions have stark differences. Even historically, after losing their independence in 13th century (as a result of Mongol invasion), Ukrainian regions had divergent experiences, with the Western part being under Polish, Austro-Hungarian and Czechoslovakian rule for many decades, while the Eastern and Central parts were under Russian rule for several centuries. Moreover, Crimea had a totally different historic track as it was populated by the ancient Greeks, the Persians, the ancient Romans, the Byzantine Empire, the Genoese and the Ottoman Empire before 19th century when it fell under Russian Empire. Later, in 20th century, the Soviet period was marked by a governmental effort to bring different regions of the Ukrainian SSR (as well as the Soviet Union as a whole) to a common system of values, behavior and way of life in general, but many regional differences persisted. These different historical experiences and their effects on national identity have been explored in detail by scholars such as Wilson (1995), Kuzio (1998), and Yekelchyk (2007). Ukraine is also a big country in terms of its population size—about 45 million3 (including Crimea4), which makes it a potentially attractive market. Additionally, Ukrainian people are well educated (its gross enrollment ratio in tertiary education for both sexes was 79 percent in 20135), with a majority living in urban areas (69%6). These factors contribute to the development of such high-tech industries as spacecraft and information technology (IT). Moreover, Ukraine is ranked fourth in the world in terms of certified IT professionals (after the United States, India, and Russia7). But even in Ukraine’s IT industry, regional disparities are evident, with the number of IT professionals varying widely across regions. 40% are in Kyiv, 19% are in Lviv (West), 16% in Kharkiv (East), and 15% in Odesa, Dnipropetrovsk, Kherson, Mykolayiv and Crimea combined (South).8 Moreover, Ukraine is rich in natural resources such as black soil (30% of world’s reserves), coal, iron ore, oil and gas, titanium, magnesium, and nickel.9 These natural resources contribute to the fact that Ukraine is ranked first in the world in sunflower oil production, second in grain exports, and eighth in steel production. It is also estimated to have the fourth largest deposits of shale gas in Europe behind Poland, France, and Norway. But Ukraine’s resources and corresponding economic sectors are also not equally distributed across the country. For example,
3 State Statistic Service of Ukraine, 2013 4 Crimean population is about 2 million (based on the last data from the State Statistics Survey of Ukraine in December 2013). 5 World DataBank, Education Statistics: Core Indicators. 6 World Bank, 2010-2014. 7 Ukrainian Hi-Tech Initiative. Exploring Ukraine IT Outsourcing Industry, 2012. 8 Ibid. 9 CIA. The World Factbook.
5 the Central region has abundant agricultural resources while the Eastern region is more oriented towards coal, steel, and other processing industries and the West hardly has any natural resource endowment besides oil and gas buried deeply in the Carpathian Mountains (with an unexplored possibility for extraction using unconventional drilling technologies such as fracking perhaps lying in the future). All of the above-mentioned factors contribute to substantial variation in economic development across Ukrainian regions. Figure 2.2 shows that the regional difference in economic growth was particularly stark in pre-war 2013. While the Central and the Southern regions had the highest growth, the Western part of Ukraine lagged behind, though still with about 2 percent growth. In contrast to the rest of the country, the Eastern region declined by about 2 percent in 2013. A possible explanation for this can be a trade war between Russia and Ukraine which started in summer 2013 when Russia blocked Ukrainian exports.10 Since many heavy industries from the Eastern Ukraine have close ties with the Russian economy, they had experienced financial losses and some reduction in output that resulted in overall regional economic decline. But this explanation falls short of explaining why the Central region still grew as many of its agricultural products were also banned by Russia. Looking further into the factors (including the role of foreign capital) behind these regional differences in development is a key piece of this dissertation. A related issue is Ukraine’s openness to and dependency on world markets. According to the economic literature (see Chapter 3), openness to trade substantially improves a country’s attractiveness to foreign investors. However, based on Ukraine’s official statistical data (see Figure 2.3), the Ukrainian regions had very different levels of foreign direct investment (FDI) at the end of 2013, pre-war. Even though FDI is only one type of investment discussed, Ukraine’s regional differences in terms of FDI stock were substantial in 2013: while less than 7 percent went to the Western region, a quarter of all FDI went to the country’s South and more than half flowed to the Center, particularly to Kyiv. Moreover, by comparing Figures 2.2 and 2.3, one can notice a regional pattern in both regional economic growth and levels of FDI received. I examine this relationship in my empirical analysis in Chapter 6.
10 Roman Olearchyk, “Russia accused of triggering trade war with Ukraine”, The Financial Times, August 15, 2013.
6 # $ &! $"
# $ " #"% ! " "
In terms of the countries of origin, Figure 2.4 shows that the FDI in Ukraine came mainly from Europe (with 31% from Germany, the Netherlands, Poland, Austria, and Great Britain), Russia (7%), and offshore countries such as Cyprus and the British Virgin Islands (37%). If one considers the fact that Ukrainian and Russian businesspeople frequently keep their capital in offshore zones and reinvest them back home, the share of Russian FDI is likely much higher than 7% and part of what is considered FDI is more correctly understood as Ukrainian domestic
7 investment. Interestingly, the origin of FDI into Ukraine has changed over time. For example, FDI from the United States has decreased substantially for the past eight years, from 13.2% percent of all FDI in 2005 to less than 2 percent at the end of 2013. Recent shifts are examined in Chapter 5 of this dissertation based on Mergers and Acquisitions (M&A) data that allow researchers to track the specific channels of domestic and international capital flows.
$""&" $ $ #% # !$#' " ' $ #"' "
$" "&" $ $ #% !$#' " '"## "
Figure 2.5 shows a sectorial structure of FDI in Ukraine in 2013. In the pre-war period, investors were mainly attracted by industry (such as processing and mining) which made up 31 percent, financial and insurance services, trade, and real estate, followed by the IT and telecommunications sector, which contributed about four percent to overall FDI flows into Ukraine. Even a basic overview of one type of investment, FDI, provides a sense of Ukraine’s regional diversity. For example, most of the investment in Eastern Ukraine is targeted towards heavy industry, since many factories and mines are concentrated there, and the major source of
8 FDI in this region is likely domestic Ukrainian or Russian capital (sometimes invested through offshore hubs, such as Cyprus). By contrast, since the IT industry is highly concentrated in Kyiv and other major cities such as Kharkiv (East), Lviv (West), Dnipro, Odesa and Sevastopol (South), those areas are the major targets of FDI in this sector. These and other patterns in international capital flows are explored in detail in Chapter 5, using the acquired M&A data and data from the European Bank for Reconstruction and Development (EBRD), a public institution investing in Ukrainian private businesses and public projects.
9
Scholars have long been puzzled about regional differences in development within a country, but the body of existing research is still sparse. Research on economic growth and development, which rests on relatively recent innovations (since 1950s), usually focuses on cross-country analysis, mainly due to the lack of subnational level data. Hence, there are still many opportunities for researchers to advance the existing knowledge. My study is informed by several bodies of research. The first is the literature on economic growth methodology. The second is research factors known to influence economic development and growth. And the third is a more limited set of studies that analyzes theories and evidence of differential regional, intra-state economic growth.
The growth modeling literature, which is neatly summarized in a review by Mankiw, et al (1995), consists of theoretical and empirical studies, as well as studies with some mixture of both approaches. Theoretically, growth modeling broadly consists of two “camps”: neoclassical growth models and endogenous growth models. Empirically, most of the research uses international cross-country data to test either different aspects of existing theoretical models or the effect of various policies on economic growth. The neoclassical growth model, developed by Solow (1956), is the foundation of existing growth methodologies. It rests on the idea that there exists a steady-state rate of economic growth to which countries converge over time. Since richer countries are closer to the steady state than poor countries, they tend to grow more slowly than less developed countries. A variation of this concept is that there might not be a single steady state for all the countries, but rather a steady-state growth rate for each individual nation to which it converges (so called conditional convergence). In the Solow model, capital input accumulation or savings (a corresponding mirror concept) is seen as a crucial factor behind economic growth, while advances in technology are assumed to be exogenous. There is also a “golden” level of savings, i.e. an optimal level of savings which maximizes a country’s consumption at the steady-state growth rate (Phelps, 1961). Over time, the neoclassical model was expanded by bringing micro foundations to its basic structure using either the overlapping generations approach (Samuelson, 1958; Diamond, 1965) or the method of an infinitely lived representative consumer (Ramsey, 1928; Cass, 1965; Koopmans, 1965). However, there are still some underlying issues with these model extensions. First, they allow higher capital accumulation than the golden savings rate. Second, they assume
10 that individuals smooth their consumption, which lacks empirical evidence (Campbell and Mankiw, 1989). In order to address some practical difficulties involved in Solow model estimation, as well as difficulties in explaining differences in economic growth between countries and the very small share of physical capital in overall country’s production, the view of capital input was expanded to include both physical and human capital (Romer, 1986, 1987; Barro et al., 1995). Testing this broadened Solow model resulted in a better data fit and higher explanatory power (Barro, 1991; Mankiw et al., 1992; Levine and Renelt, 1992). A different theoretical view of economic growth is presented by endogenous growth models, which do not postulate some abstract steady-state rate of growth. In contrast, it suggests that capital accumulation (comprehensively defined) could lead to growth forever. Specifically, if capital includes physical machines and equipment, human capital, and social knowledge in the form of innovation, then its accumulation means endogenous advances in technology that consequently lead to infinite growth (Romer, 1986; Kremer, 1993). The endogenous growth model was also expanded to better characterize “innovation production” in the form of research and development (Uzawa, 1965; Lucas, 1988; Mulligan and Sala-i-Martin, 1993) or monopolistic competition among firms (Romer, 1990; Grossman and Helpman, 1991). The endogenous growth model is considered to better explain growth differences between countries, though many practical issues arise with measuring this model’s broad concept of capital.
Another distinct part of research concerns known factors influencing economic growth. Barro (1996a) summarizes this group of studies and draws a number of growth factors. For instance, he determines that, for a given country’s initial real GDP per capita, its growth rate is “enhanced by higher initial schooling and life expectancy [labor input], lower fertility [population growth input], lower government consumption, better maintenance of the rule of law, lower inflation, and improvements in the terms of trade [government policies inputs]” (p. 2). In another essay, Barro (1996b) establishes the relationship between economic development and democracy as measured by the indicators of political rights and civil liberties (p. 34-37). He also discusses growth models in terms of their view of technology, which can either be treated as an exogenous factor or be endogenized. In his view, a remaining question is whether there is some threshold which determines whether a given capital flow into a country includes the technology transfer with it (i.e. technology is endogenous) or not (i.e., technology is exogenous). Other studies provide insights about the role of different types of capital input in a country’s GDP, such as private or public capital. For example, Romp and De Haan (2007) review the literature on existing theoretical and empirical links between public capital (e.g., infrastructure) and economic growth. As a result, they generalize that in existing cross-country analyses, public capital is usually considered using either a cost-function approach, a vector autoregression
11 approach, or a crowding-out approach. The authors conclude that there is an overall consensus in the literature about a positive influence of public spending on economic growth, but note that the magnitude of this effect varies. They also note that public capital is a complex factor since it is a function of government institutions and policies. Consequently, public capital, when analyzed, is usually treated as domestic capital which is not the case, for example, when government gets foreign aid and uses it for infrastructure enhancement. Looking at private capital, Borensztein et al. (1998) examine an interaction between FDI and economic growth through human capital. Their hypothesis is that FDI brings along knowledge, i.e. technology transfer through available human capital. The authors develop a theoretical endogenous growth model and use it to test their hypothesis based on a sample of 69 developing countries that receive FDI from high income countries. Their results indicate that FDI boosts economic growth, though the magnitude of this process depends on the human capital capacity of receiving country; the higher the human capital is, the greater is the absorption of the positive effects of FDI. Mishra, Mody and Murshid (2001) look at the role of private foreign capital in a country’s economic growth. They argue that there are three main channels through which private capital flows affect economic growth: (a) by increasing domestic investment (with the “relationship being strongest for foreign direct investment and international bank lending and weaker for portfolio flows”); (b) by improving productivity (with the effect being higher for those countries which have a more educated labor force and better R&D capabilities, such as Malaysia, Taiwan, and the southeastern provinces of China); and (c) by affecting volatility, which is inversely related to growth but could be mitigated by good government policies and regulations. Choong et al. (2010) expand these findings by looking closely at that the stock markets as the “channel[s] through which foreign [private] capital flows could promote growth.” The authors show that “both foreign debt and portfolio investment have a negative impact on growth” (p. 112). An article by Easterly and Levine (2001) presents a summary of the existing evidence about the role of total factor productivity (TFP) in economic growth. The authors derive five major conclusions: (1) the factor accumulation is not as important as the TFP which is a residual in econometric estimations; (2) in the long run, economies diverge (i.e. conditional convergence does not hold); (3) economic growth is not steady over time as compared to capital accumulation (which is why the authors argue the existing growth models are very country-specific); (4) economic factors of production such as physical and human capital, usually concentrate in some areas, especially in richer areas; and (5) public policy can influence economic growth in the long run if it is focused on the TFP and technology transfer rather than factor accumulation. Shepotylo (2010) examines spatial determinants of FDI location in 27 transition countries (in Eastern Europe and Central Asia) between 1993 and 2007. Using a spatial autoregressive model and a gravity model approach, he finds a statistically significant spatial effect of FDI in transition countries, especially in early stages of their transition to the market economy and in their underdeveloped sectors (e.g., services). Furthermore, the author argues that, as a market matures,
12 its size starts to play a more important role in FDI allocation. Shepotylo’s paper is partially related to my dissertation topic. It highlights the fact that the regional spillover effects are important factors in understanding local economies. Numerous studies analyze the effects of public policy (e.g., fiscal, monetary, and regulatory policy) and economic growth. For example, Barro (1990) develops a detailed theoretical model of the nexus between fiscal policy and economic growth. He concludes that government expenditure, if financed by a flat income tax, first enhances growth and then reduces it. He emphasizes that government policies are not exogenous (random) but rather endogenous, and shows that if government were to set policies randomly, the relationship would be non- monotonic. However, as government tries to optimize the utility attained by a representative household, the relationship between fiscal policy and economic growth (and savings) is inverse. Barro’s findings have a number of limitations. For example, he does not differentiate between types of capital in a private production function (e.g., domestic, foreign, human capital) or specify a labor input. Also, he does not examine such issues as the scale or nature of fiscal policy (i.e. how large government spending is, whether it is increasing or stable; whether debt-financed or tax financed) and different types of private capital. In another study, Barro (1996) looks at a link between monetary policy, as measured by inflation, and economic growth, and concludes that the higher inflation is, the lower is economic growth. King and Rebelo (1990) summarize the existing models of fiscal policy and growth. They also expand a simple endogenous growth model with one aggregate capital input to a two-sector production function with both physical and human capital. The latter model enables the authors to draw conclusions about impact of fiscal policy on economic growth through its impact on human capital accumulation. Their results show that this relationship is inverse. Moreover, by calibrating parameters for small open economies (i.e., in terms of access to international capital markets), the authors conclude that government taxation results either in a rapid growth “miracle” (low taxes) or in economic stagnation (high taxes). Yet they leave the mechanism of influence involved in both cases unexplained. The authors also do not differentiate between different taxes (such as flat or graduated, on income or on consumption, VAT) to estimate their possible divergent effect. Instead, they assume an average tax rate applied equally to all sectors of the economy. There is some empirical evidence that there can be both an optimal taxation regime and an optimal income distribution (as measured by Gini coefficient) that maximize country’s economic growth (Scully, 2003). Moreover, any deviation from the optimal level of taxation, for the purpose of reducing income inequality, would result in reduction of economic growth. Scully estimates the US optimal annual growth rate at 6.97% with an optimal taxation level at 19.3% of GDP and a growth-maximizing Gini coefficient of .359. However, he doesn’t specify why different inequality levels would adversely affect growth. Another result of Scully’s study is that changes in government fiscal policies influence a country’s growth rate more than its income distribution. In Ukraine’s case, even though there is hardly any difference in the tax regime
13 across regions, government expenditure (such as subsidies to industries) is differentiable, and more government spending flows to the Eastern industrial region than to the Western region. Another comprehensive analysis of the impact of monetary and fiscal policy on long-run economic growth is provided by Fischer (1991). The author defines these policies to include government determination of a targeted inflation rate, the budget deficit, and external debt levels, focusing on the former two in his paper. Overall, both inflation and the deficit are viewed to be negatively correlated with growth. High inflation induces inefficient capital accumulation whereas budget deficit, through taxation, crowds out private investment. A part of Fischer’s study concerns his review of the existing literature on macroeconomic factors that affect investment. He outlines several key influencing factors: (a) stability of output, (b) exchange rate, (c) cost of capital, and (d) a country’s overall economic strategy (i.e. market openness, size of government, infrastructure and human capital development). But he does not discuss an implied mechanism of how output stability affects investment and growth. For instance, macroeconomic instability, as measured by a black market premium, has a negative effect on investment and ambiguous association with growth. Ultimately, Fischer concludes that the existing evidence is weak, especially when it concerns understanding how macroeconomic policies impact growth or underlying factors of investment. He also suggests that the research should move from cross-country analysis to time-series studies of one particular country at a time. This is where my research fills a gap in the literature. Following Fischer’s work, Stiglitz (2000) considers a possible cyclical pattern of investment that might both boost growth and impart economic instability. His overview of the existing studies shows that capital liberalization on average leads to a higher economic growth rate, but it can also cause instability which, in turn, adversely impacts growth. As a result, he suggests a policy intervention in the capital market—taxation of short-term capital inflows. To understand an effect of international public capital (i.e. foreign aid) on a country’s economic growth, Boone (1996) looks at different political regimes (elitist, egalitarian and laissez-faire) and shows that whatever the regime is, foreign aid will ultimately go to consumption rather than investment. It will also lead to an increase in government consumption (e.g. two thirds of the aid would be consumed by an elitist government which optimizes wealth of political elites). The author also tests any statistically significant impact of foreign aid on human development and concludes that, no matter what political regime is considered, there is no effect. Boone’s analysis raises a question about any policy measures that are likely to interfere with the consumption vs. investment effect of foreign aid. For example, the known differential policies are OECD philanthropic aid vs. China’s “quid-pro-quo” type of aid. Hansen and Tarp (2000) conduct a systematic review of 29 existing cross-country studies of foreign aid’s effect on economic growth (via savings, investment and macroeconomic policies). They conclude that there is a consistent pattern of results in the reduced-form models: foreign aid increases both savings and investment and positively affects growth. The authors also state that a perceived negative relationship between aid and growth is based on several studies that dominate
14 a scholar discussion, rather than an overall body of existing research on the topic. Additionally, they analyze influence of macroeconomic policies on marginal productivity of aid and find that this effect is very ‘noisy’. In the authors’ view, future research should focus on understanding how foreign aid affects growth and what aid instruments are more effective based on various local preexisting conditions. To summarize, the existing studies highlight an overall positive effect of foreign aid on a country’s economic development—a relationship I test in the case of foreign aid to each of Ukraine’s four regions. The literature also emphasizes the importance of policies that could reduce the likelihood of negative effects of foreign public funding, and increase likelihood of positive ones. Given the current security situation in Ukraine, I also explored the literature about the relationship between economic development and security, on top of other known economic and social factors. Even though this field of research is not as developed as those mentioned above, there is research on such issues as the impact of military spending on economic growth and vice versa, and the role of national security in a country’s economic development in general. Below I summarize some selected studies on the topic. Conventional Keynesian economic reasoning indicates that in times of crises, increased government spending can stimulate economic growth. Indeed, some research (Benoit, 1978; Whynes, 1979) shows that there is a significant positive correlation between military spending and economic growth. However, economic theory and empirical evidence suggest that large government investment “crowds out” private investment (Lim, 1983; Deger and Sen, 1983; Deger and Smith, 1983). For example, one study finds that a ten-percentage-point increase in defense spending leads to a reduction of country’s annual growth by .13 percentage points (Faini, et al., 1984). Similarly, a study of military spending and economic growth in Middle Eastern countries by Lebovic and Ishaq (1987) indicates that higher military spending suppresses these countries’ economic growth. Yet, economic growth can also affect national security, so the relationship is bidirectional. During an economic slowdown, a country can lack resources to finance its defense sector. For example, as the EU introduced austerity measures in the aftermath of the global financial crisis 2008-2009, its member nations’ capacity to project military power became much constrained. Budget cuts and reductions in military spending substantially affected the manpower, force structure, and capabilities of the key EU (and NATO) members, including the United Kingdom, France, Germany, Italy, Spain, the Netherlands, and Poland (Larrabee et al., 2012). The current European security crises involving Ukraine started exactly at the time that defense budgets in EU and NATO member countries were being pared down. Many observers have suggested there might be a relationship. Lastly, Murphy and Topel (2013) present a theoretical concept of national security economics. The authors define national security as “the set of public policies that protect the safety or welfare of a nation’s citizens from substantial threats” (p. 508) and regard expenditure
15 on national security broadly. They show that public investments in national security measures depend largely on the perceived probability and severity of various threats, which determine their expected value and pay-offs. Consequently, the authors underscore that it is important to make national security investment flexible and scalable because of associated uncertainties.
A third group of existing literature relevant to this study focuses on theories and evidence of differential regional, intra-state economic growth. For instance, Gennaioli et al. (2014) use a sample of 1,528 regions from 83 countries, including Ukraine, to estimate the speed of regional convergence within countries as well as cross-country convergence. The authors conclude that the regions converge on average at about two percent, a rate comparable to convergence across countries. Also, their findings suggest that factor mobility and institutions (e.g. capital market development) play an important role in regional economic differences. Based on these findings I include a measure of local capital market development in my regional analysis. Ali and Ahmad (2010) explore the impact of FDI on regional economic growth by looking at a Malaysian case. The authors explore a similar hypothesis to that in my dissertation—namely, that foreign private capital disproportionately affects regions within a country and leads to disparities between developed and poor regions. However, they use a less comprehensive and robust approach, only focusing on a production function approach and how output is correlated with FDI. Their models also include only a sparse set of explanatory variables beyond basic capital and labor inputs measures in the country’s output. Thus, even though the authors ask a question about the effects of FDI on Malaysian regional development, they do not closely analyze the channels of FDI impact. They mention that natural resources, urbanization, infrastructure and the location of industry may play a role in differential FDI in regions and establish the existence of such connection, but do not explore it in great depth. Resende (2011) investigates differential regional development in Brazil. In his paper, he looks at a spatial analysis of local Brazilian economic growth on three levels: municipalities, spatial clusters and states. The author runs the same single regression for 1991-2000 data on three levels and concludes that there are several variables that are simultaneously significant at all the spatial scales: higher education, health and local infrastructure. These factors are usually targeted by government policies across the whole country. Finally, the author finds that the Brazilian regions gradually converge over time. The author’s key innovation is in inclusion of regional spillovers into the analysis. For example, there might be some externalities that affect regional development besides local specific factors, such as local public policies and health. To take into account these spillovers I consider areas with similar background factors together and group 24 Ukrainian oblasts into four macro regions. Using principal component analysis, Henning et al. (2011) study Swedish regional growth since 1860 and conclude that there are several clear influencers in Swedish regional economic
16 growth. These include the First Industrial Revolution and expansion of export markets, the Second Industrial Revolution and development of the welfare state, and, finally, the expansion of the metropolitan regions. The authors find that metropolitan regions develop in a counter- cyclical way if compared to the country as a whole. These localities also demonstrate higher economic growth than regions with abandoned resources, though no hypothesis is advanced about reasons behind this counter-cyclical pattern of metropolitan development. Finally, Akulava and Vakhitova (2010) examine differential impact of FDI on economy’s sectorial performance between 2001 and 2007, using a dataset of Ukrainian firms. They consider three different economic sectors: primary (agriculture and mining), secondary (manufacturing), and services. Based on the statistical analysis, the authors show that the non-equal FDI impact is due to (a) weak intra- and inter-sectorial links (horizontal and vertical FDI spillovers) and (b) sector-specific entry restrictions. Other key findings include: (a) that there is a positive horizontal FDI spillover effect, especially in the manufacturing sector due to its high competitiveness; (b) that there is a negative vertical FDI spillover effect in the primary sector for domestic firms and in the service sector for foreign firms (defined as those with more than 50% foreign ownership) due to productivity differences between domestic and foreign firms; (c) that the effect of FDI is largest in the primary sector, which is also the most restricted one; and (d) that foreign firms, on average, perform better and benefit more from the FDI spillover effects than Ukrainian domestic firms. The study does not elaborate on the mechanism underlying the last relationship. The paper by Akulava and Vakhitova (2010) is important for my dissertation as it not only focuses on Ukrainian domestic and foreign firms, but also investigates the disproportionate effects of FDI on different economic sectors. Given that Ukrainian regions are non-homogeneous in their sector-specific development, my study expands this research by looking at a broader regional dimension.
My literature review informed my selection of variables for inclusion in my empirical models, which are reported in Chapter 6. Across the papers, the most commonly considered categories of factors thought to influence growth were labor, capital, institutions, and policies. Labor-related factors include population growth, life expectancy, fertility, mobility, and urbanization. Capital-related factors are typically divided into three categories—private, public, and human—which are seen as having distinct effects. Private capital can be domestic capital stock and foreign investment (e.g., FDI or private foreign debt). Public capital is usually viewed as either development aid (e.g., provided by international organizations such as the World Bank and USAID), financial aid (e.g., provided by international financial organizations such as IMF), or public infrastructure (roads, bridges, utilities, transportation, cultural/sports), which can also be financed through development or financial aid. Human capital is usually measured in terms of schooling, R&D, universities, managerial skills, and health.
17 Institutional development, another crucial factor, is usually measured in terms of the strength of the rule of law; social capital, political rights, and democracy (sometime measured using data on voter participation, protests, and internet penetration), market development, economic activity, the nature of interactions between people and firms, and ease of doing business (e.g., the World Bank ranking). Monetary and fiscal policies thought to contribute to economic growth include targeted inflation, cost of capital/interest rate, exchange rate regimes, government subsidies (either debt- financed or tax financed), and taxes collected/government consumption. In addition, existing studies point to a host of other factors that play a role in economic growth. These factors include openness to trade, the extent of exports and imports, tourism, volatility or output instability, geography, natural resources, living standards, and the location of industry.
18
This chapter explores three case studies of countries in which distinct regions have experienced different growth trajectories and still have stark income gaps today: China, Italy, and the United States. The case studies shed light on the nature and extent of regional economic disparities that can emerge within countries, and on the role that international capital flows can play.
Administratively, China consists of 23 provinces (including Taiwan province), four municipalities (Beijing, Tianjin, Shanghai, Chongqing), five autonomous regions (Guangxi, Inner Mongolia, Tibet, Ningxia, Xinjiang), and two special administrative regions (Hong Kong, Macau). Based on the classification system used by China’s National Bureau of Statistics, I group Chinese provinces into four regions: Eastern, Central, Western and Northeastern. (See Figures 4.1.1. and A1 in Appendix A for more details.) Specifically, these regions include: • Eastern region: Beijing, Tianjin, Hebei, Shanghai, Jiangsu, Zhejiang, Fujian, Shandong, Guangdong and Hainan; • Central region: Shanxi, Anhui, Jiangxi, Henan, Hubei and Hunan; • Western region: Inner Mongolia, Guangxi, Chongqing, Sichuan, Guizhou, Yunnan, Tibet, Shaanxi, Gansu, Qinghai, Ningxia and Xinjiang; • Northeastern region: Liaoning, Jilin and Heilongjiang.11
11 China Statistical Yearbook, National Bureau of Statistics of China, 2005
19 China has developed considerably since 1978, the start of its market-oriented reforms that gradually opened some parts of the economy to foreign trade and investment. These reforms also introduced free market principles to limited economic activities in some regions, leading to fewer restrictions on labor and capital markets (Cai et al., 2002). Since then, Chinese GDP has increased more than ten times and its annual GDP growth was in double digits for many years until country’s economic slowdown in mid 2000s. In the 2010s, China has maintained real GDP growth of about 6-8%.12 Despite this quick growth and the overall national wealth, its per capita GDP is still quite low ($14,600 in 2016, PPP13). In other words, its GDP per person is comparable to that of Brazil and Macedonia, and is ranked 112th among similar figures for 230 countries and dependent territories. China’s rapid economic growth was associated with increasing disparities in development between the coastal and inland regions and between urban and rural areas (Chen and Feng, 2000; Pedroni and Yao, 2006; Goh et al, 2009; Cheong and Wu, 2014). Figures 4.1.2 and 4.1.3 show the country’s regional differences in per capita GDP. Between 2001 and 2013, the Eastern coastal region was about two times more prosperous than the Central and Western regions. In the Northeastern industrial region, the average per capita income was about two thirds of that in the Eastern region and more than 25 percent higher than the average for the inland provinces. These regional disparities have long been of interest to researchers and policy-makers and, hence, there are numerous studies of the factors responsible, including those linked to foreign investment and specific public policies. I consider them in the following sections.
12 CIA Factbook, 2016 13 Ibid.
20
21 The body of research on the factors underlying China’s regional disparities generally confirms the common findings of the economic growth literature (Chen and Feng, 2000). Specifically, the factors are those described in Chapter 3 of this monograph and associated with: labor (human capital and fertility), capital (industrialization, infrastructure and other government expenditures, private investment), technology, and policies (fiscal and monetary), among other factors (institutions, geography, and openness to trade). While the list of the influencers is widely accepted, their importance and ranking are not. For the labor factors, China’s demographics (e.g. the doubling of its population size since the 1960s) are thought to be connected with its regional development in two ways. On the one hand, the evidence suggests that the overall rapid population growth (as measured by fertility or migration) is associated with smaller economic growth (Chen and Feng, 2000; Phillips and Chen, 2011). On the other hand, the population age structure is critical. Particularly, whereas younger cohorts contribute to a larger labor force and higher human capital development (as measured by university enrollment or years of schooling) that enhance growth (Henderson et al., 2007), the aging population has a negative effect on regional growth as the local economy spends more resources on their support (Chen and Feng, 2000; Liu and Li, 2006; Zhang et al., 2015). Additionally, China’s labor mobility, another important growth factor which can offset negative demographic effects, is relatively low. The latter fact may be explained by the rigid labor policies, such as the existing household registration system (hukou), that make it hard for people relocate far from their native provinces (Cai et al., 2002). Along with local population and its human capital, geography is another factor widely known to cause regional economic differences (Sachs, 1997; Gallup and Sachs, 1998; Bao et al., 2002; Batisse, 2002; Demurger, 2001). In China’s case, the Western region is the hilliest (with the Qinghai–Tibet Plateau) while the Northeastern and Eastern regions along the coastline are mostly flat. This creates more favorable conditions for agriculture and trade in coastal provinces. For example, the seaports in the East, such as Hong Kong and Shanghai, have long been the hubs of trade with Europe and America and played an important role since the 19th century. One study even finds that, if investment and labor inputs are endogenously defined by the locality, then up to 60% of the Chinese regional variation in growth could be explained by its geography (Bao et al., 2002). Moreover, since 60% of China’s territory is mountainous, and only about 10% is plain, its “topographic feature implies an unfavorable natural condition for economic development because of the higher costs of transportation and infrastructure construction.” Besides demographics and geography, the literature suggests that government policies, both national and local, are crucial for Chinese regional development. The economic reforms of the late 1970s resulted in the divergent growth between the Eastern region and the rest of the country, magnifying the pre-existing economic differences between the Eastern and Western parts. For example, the initial stages of the reforms benefited local agriculture and resulted in
22 more agricultural output in the Western and Central provinces, where agriculture was an important economic sector (Cai et al., 2002). In contrast, the ‘open door’ policy particularly concerned the Eastern region where the first special economic zones and open cities were established to promote international business. It led to an increase in international trade, improved regional productivity and enhanced economic development in those areas (Jiang, 2011). As a result, the economic gap between the Eastern region and the rest of the country became substantial and remained even when, in 1990s, the opening-up policy was expanded to inland regions. Several studies recognize the crucial role that the government policies of special economic zones (SEZs) and industrial clusters have played in China’s rapid economic growth (Batisse, 2002; Zeng, 2011; Frattini and Prodi, 2013). Industrial clusters are “geographic concentrations of interconnected companies, specialized suppliers, service providers, firms in related industries, and related institutions in a particular field” (Porter, 1998). These include financial and educational institutions, and various levels of government (Zeng, 2011). SEZs are rather of a commercial nature. The World Bank defines SEZ in the following way: “The basic concept of a special economic zone includes several specific characteristics: (a) it is a geographically delimited area, usually physically secured; (b) it has a single management or administration; (c) it offers benefits based on physical location within the zone; and (d) it has a separate customs area (duty-free benefits) and streamlined procedures (World Bank 2009). In addition, an SEZ normally operates under more liberal economic laws than those typically prevailing in the country. SEZs confer two main types of benefit, which explain in part their popularity: “direct” economic benefits such as employment generation and foreign exchange earnings; and the more elusive “indirect” economic benefits [such as technology transfers and export diversification].” (Zeng, 2011, p.3) In China, industrial clusters are usually in labor-intensive manufacturing sectors in cities along the Pearl River Delta. In some rare cases, these industrial clusters perform higher-end services, especially when they develop from the existing SEZs. The examples of the latter include the information technology clusters in Zhongguancun (Beijing) and Shenzhen, and the electronics and biotech clusters in Pudong (Shanghai) (Zeng, 2011). The research also suggests that SEZs and industrial clusters contribute to a country’s product value added, increase exports and attract foreign investment (Wang, 2013). The literature also suggests that the more a region is integrated into the global economy, the higher its domestic benefits are. These benefits include productivity and technology spillovers that result from export sophistication,14 especially in the processing trade,15 and contribute to
14 Depending on definition and measurement, export sophistication is either productivity or technology associated with the country’s exported goods. 15 Processing trade occurs when imported raw materials and parts are processed and assembled domestically, but the final products are exported for sale in the global market.
23 (endogenous) economic growth (Jarreau and Poncet, 2009). The latter implies that the gap between the Eastern and the inland regions would probably remain in future because, while the Central and Western regions were isolated from world markets, the Eastern region benefited from high trade, investment and interaction with foreign firms for an extended time period. Finally, the direct interference of the central government in local economies, through state- owned enterprises (SOEs) and government consumption and lending, have also played a role in China’s regional development. Liu and Li (2006) show that “domestic bank loans and foreign- owned enterprises are important in coastal provinces, while state appropriation and state-owned enterprises are important in inner provinces.” This finding echoes that of Chen and Feng (2000), who show that SOEs impede regional growth in China. Intra-regional gaps are further aggravated by the fact that local SOEs not only receive more government lending but also trade less with other regions due to the local protectionist policies (Demurger et al., 2002).
Public investment (such as airports, railways, roads, pipelines, education, health etc.) is an important piece of the Chinese government policy towards regional development. The evidence suggests that public investment in physical and social infrastructure enhance economic growth (Demurger, 2001; Fan and Zhang, 2004) by not only improving the productivity of private investment (Aschauer 1989a, 1989b, 1993) but also by potentially reducing disparities between regions (Costa-i-Font and Rodriguez-Oreggia, 2005; Fan and Zhang, 2002). The latter notion is supported by the fact that governments tend to balance equity and efficiency in their decisions of where to invest locally (Lambrinidis et al, 2005; Castells and Solé-Ollé, 2005). However, in China, political factors play a significant role in allocation of public funds for local projects (Zheng et al., 2013). Specifically, its centralized socialist system uses a top-down approach for investment allocation, and China Communist Party members’ personal connections play an important role in this process. According to Zheng et al. (2013), between 1995 and 2008, the Chinese central government invested up to 773.66 billion RMB on average (or 226.88 billion USD, using the average PPP exchange rate for the period) in local infrastructure projects, about 80% of the central government total budget revenues or about 5% of the averaged GDP during the period. But the distribution of this public investment was uneven. The least developed Western region obtained several times more resources than the Central or Eastern regions (except for Beijing and Tianjin provinces; see Figure 4.1.4 for more details), and the Northeastern region received the second largest public investment.
24 ! ! !
! " " !! ! !# !
Figures 4.1.5 and 4.1.6 compare more recent central and local government expenditure. In 2013 the major central government spending mainly concerned national policies—security, education, science and technology, welfare, healthcare and environment. In contrast, priorities of local governments were rather different. The same 2013 data show that local governments spent most of their money on local public goods such as urban and rural development, housing, transport and local infrastructure. Hence, regional authorities invested primarily in physical infrastructure while public capital in social infrastructure (such as education) was provided mainly by the central government. Additionally, much of the local public spending was financed by debt, local real estate tax revenue and public-private-partnerships, whereas national public spending was more tax financed.
25 " !
$ ' &'$, &$ $ $ %"!$&& ! !$%&$+ &$ 28 ' $( % 38 ( $! & ! %$( + 48 $!&& ! 18 /8 !$ $% 08
&$ .8
! &+ & " !+ & ! !+ !$& /38 & ! % 28 238
' '& ! '$ &+ 58 68
" !
$ &$ $%! &! 28 48 $ %$(% $%! /8 &$ 18 $ ' !%&-$&#' !($+ $( % ! %&$'& ! 008 .8
$%! *" !$& ! , !)$ ' '$ &+ !$ & ! //8 68
$ ' &'$,"!$& '$ ! ' &+ $ %"!$&& ! $% 28 /38 $%! /78 !'% '$ &+ 58
26 On the policy side, in 2013 Chinese government introduced the Silk Road Economic Belt and the 21st-century Maritime Silk Road (also called the Belt and Road Initiative or the One Belt One Road) with a goal to not only promote economic development and open up the Western region to global trade, but also improve China’s relations with countries in Central Asia, the Middle East, and South Asia. The initiative focuses on infrastructure investment in the Chinese underdeveloped inland regions and connecting them via transport and economic corridors to foreign markets (National Development and Reform Commission, 2015). The policy rests on a major assumption that regional economic development is primarily related to infrastructure investment. The effectiveness of this approach in ‘building a bridge’ between the country’s poor Western and richer Eastern regions has yet to be evaluated. Another policy pursued by the government to reduce regional disparities is fiscal decentralization. The effect of this policy remains uncertain (Zhang and Zou, 1998; Lin and Liu, 2000; Jin and Zou, 2005; Brehm, 2013). The proposed explanation lies in the fact that (a) ultimately poor regional economic development impedes local fiscal opportunities, and (b) local fiscal decisions are usually made on the provincial levels, which leaves little discretion for policy-making on sub-provincial rural level (Brehm, 2013). Also, Wang (2013) argues that the central and local governments have conflicting motivations regarding levied taxes: local authorities are interested in a broader tax base while the central government is more interested in tax revenues from higher corporate profits (that become smaller as more firms enter the local markets and drive profits down).
The role of foreign direct investment (or FDI) in China’s economic growth has long been of research interest. Overall, the literature suggests that there is a positive correlation between the two (Liu and Li, 2001, 2006; Mullen and Williams, 2005; Yao, 2006; Ran et. al, 2007; Whalley and Xin, 2010). Moreover, the evidence also indicates that FDI is one of the reasons behind the country’s regional disparities (Liu and Li, 2001; Ran et al., 2007; Yu et al., 2011; Chen and Groenewold, 2013; Lessmann, 2013), though the findings about the magnitude of this effect are mixed. Isolating the causal relationship between FDI and economic growth is also complicated because of the variables’ likely endogeneity and the interactions between them. Figure 4.1.7 maps Chinese provinces with their corresponding 2013 FDI stock. A stark difference between the coastal and inland regions is evident. Three Eastern provinces alone – Shanghai, Jiangsu, and Guangdong—received half of all FDI. 25 percent was invested in other Eastern provinces. The rest of the country got only a quarter of total foreign investment.16 The major sectors attracting FDI were manufacturing, real estate and trade, although manufacturing’s
16 National Bureau of Statistics of China, 2013.
27 share has declined since China’s accession to the World Trade Organization (WTO) in 2001 (Yu et al., 2011).17 There are numerous factors that contribute to regional differences in both FDI attractiveness and absorptive capacity. The known aspects include, but are not limited to, geography (Bao et al., 2002), local labor markets, differences in regional productivity (measured by TFP or human capital; Borensztein et al., 1998; Bengoa and Sanchez-Robles, 2003; Shiu and Heshmati, 2006), financial development (Hermes and Lensink, 2003), overall market development (Balasubramanyam et al., 1996; Basu et al., 2003), infrastructure, and government policies (e.g., Open Door Policy, Special Economic Zones, Development Zones, or Economic and Technology Development Zones, Coastal Priority Development Strategy, the Resurgence of the Old Northeast Industrial Base). In China’s context, there are several reasons why the coastal region is particularly attractive to foreign capital. First, as previously mentioned, the Eastern provinces not only had a pre- existing competitive advantage but also were the first to benefit from the ‘opening-up’ policy via their ports and newly established special economic zones. As a result, this region prospered from the very launch of the economic reforms. Second, Long et al. (2014) find that Chinese policies of lower tax rates and fee burdens, paired with “less arbitrariness in such burdens, as well as a higher level of legal protection for domestic firms in the same region,” are significantly correlated with FDI inflows into these provinces. FDI also seems to be negatively correlated with local taxes, such as those on natural resources, land and housing. Third, the literature suggests that a region’s ability to absorb FDI matters for how much FDI is eventually attracted. The latter depends on regional institutional development, including local public policies and fiscal decentralization (Wang, 2013). Finally, in addition to producing positive technology spillovers (Coe and Helpman, 1995; Liu and Wang, 2003; Tuan et al., 2009), foreign capital also augments and reinforces local institutions (Long et al., 2014). For example, foreign and domestic firms can join their efforts in lobbying for a simpler regulatory environment. An important developmental mechanism, then, is that the local initial endowment and public policies induces FDI inflows, and this FDI then strengthens local public institutions.
17 National Bureau of Statistics of China, 2013.
28 &# %! !!% '$% %( !# & %#"# $$#!$$ $ ! $ #!' $ !
29 Italy is a medium-sized country with a relatively large population (about 62 million18). Its territory is half the size of Ukraine’s, while its population is about 40% bigger. However, similar to Ukraine, Italian regional development is very uneven. Its Northern and Southern parts are very different economically, much like Ukraine’s Western and Eastern parts. Administratively, Italy consists of 20 regions, 5 of which have autonomous status (Sardinia, Sicily, Trentino-Alto Adige, Aosta Valley and Friuli-Venezia Giulia). Each region consists of provinces that total to 110. However, since 2014, Italy has been going through a constitutional reform, aimed at reducing the number of provinces to 51 and creating 10 special metropolitan areas (Rome, Turin, Milan, Venice, Genoa, Bologna, Florence, Bari, Naples and Reggio Calabria). Proposed legal changes would have implied a change in local governance structure: provincial governments were to be abolished and local decisions (about territorial planning, transportation, and local education) were to be transferred to the central government in Rome, while other political decisions were to be made by three appointed councilors, instead of local political governments.19 The proposed changes were rejected by the popular referendum in December 2016. For the purposes of this case study, I group Italian administrative regions into five macro- regions: Northeastern, Northwestern, Central, Southern and islands. (See Figure 4.2.1. and A2 in Appendix A for more details.) These regions include the following administrative entities20: • North-West: Piemonte, Aosta Valley, Lombardy, Liguria • North-East: Trentino Alto Adige, Veneto, Friuli Venezia Giulia, Emilia Romagna • Centre: Toscana, Umbria, Marche, Lazio • South: Abruzzo, Molise, Campania, Apulia, Basilicata, Calabria • Islands: Sicilia and Sardegna Italy is economically highly developed, with the third largest economy in the euro area and the eighth largest in the world. Its economy is very diversified. 99.9 percent of Italian businesses are private small and medium-sized enterprises, usually family-owned. Tourism plays an important economic role, along with pharmaceuticals, industrial machinery, electrical appliances, automobiles and auto parts, textiles/fashion, and food and wine. Another Italian advantage, which facilitates trade, is its proximity to emerging economies in the Middle East and North Africa (U.S. Department of State, 2015).
18 CIA Factbook, 2016. 19 OECD Regional Outlook 2014: Regions and Cities: Where Policies and People Meet. OECD 2014. Lisabetta Povoledo, “Italy’s Constitutional Referendum: What You Need to Know”, The New York Times, December 2, 2016. 20 This classification of geographical areas is used by the National Institute of Statistics of Italy, ISTAT.
30
Italy’s economic growth was relatively slow (on average about 1.5% between 1990 and 2007)21 until the country was hit by a global financial crisis in 2008. In 2009, Italian GDP dropped by 5.5%22 and due to its high public debt (more than 130% of GDP23) the country experienced severe fiscal problems. However, the Italian economy finally started to recover, with 0.8% growth in 2015-16.24 Despite being one of the leading global economies, Italy’s GDP per capita (at $36,300 in 2016, PPP) is below the European Union average ($39,200 in 2016, PPP) and causes the country to be ranked 37th of the 196 countries of the world.25 Moreover, Italy’s national wealth is distributed very unequally across its regions: there is a relatively big gap between its Northern and Southern regions. Figures 4.2.2 and 4.2.3 show the Italian regional differences in per capita
21 OECD.Stat, 1990-2015 22 Ibid. 23 CIA Factbook, 2016. 24 Ibid. 25 CIA Factbook, 2016.
31 GDP between 2001 and 2009.26 While the country’s Northeast and Northwest have about the same per capita income, their figures are almost twice as large as those for the Southern provinces and Islands. The Central region (which includes the capital) is a bit poorer than the Northern provinces but it is still more than 50 percent wealthier than the Southern part of the country. The developmental gap between the Italian regions also reflects their regional economic structure where the Northern and Central provinces are industrially and technologically developed and the Southern regions and Islands are predominantly agricultural, greatly subsidized and with high unemployment rates.
!
Studies on Italian regional disparities generally follow the research findings summarized in Chapter 3 of this monograph. Specifically, its regional economic differences are associated with local differences in infrastructure and other public expenditure, human and social capital, TFP, R&D, technological and agglomeration spillovers (Coppola et al., 1998; Picci, 1999; Aiello and Scoppa, 2000, 2008; Petraglia, 2002; Ascari and Di Cosmo, 2005; Di Liberto et al., 2008).
26 Note that 2009 is the most recent year available for this statistic on Istat.it
32
However, in the Italian case, geography and physical “assets” have not always guaranteed good economic performance (in contrast to China’s case). Previous studies have found that the country’s Northwest benefited mostly from its physical assets, whereas the Northeast and Center advanced from non-material social factors. In this respect, the Italy’s South is different because it lacks both physical and non-physical factors that exist in other three Italian regions. A substantial body of research about Italian development focuses on the notion of social capital introduced by R. Putnam in 1993. By studying Italian regional governments for two decades, Putnam concluded that the key regional difference was in different “civicness”, i.e. “norms and networks of civic engagement” (Putnam, 1993). Moreover, the author argues that regional economic disparities can also be explained by these differences in civicness: the Northern Italian provinces have greater horizontal civic networks, while social relations in the Southern provinces are more vertically structured. Putnam measures civicness by a “Civic
33 Community Index” which consists of the following components: preference voting, referendum turnout, newspaper readership, and scarcity of sports and cultural associations. Ten years later, Putnam’s theory was tested again using more recent Italian data. Ballarino and Schadee show that the relationship between civicness and economic performance still holds, though mainly in the longer run (20 years effect). Moreover, the authors also find a simultaneous shorter-term effect of the economy on civicness (10 years effect), which they explained by cohort changes. While Putnam’s theory seems to be true for the Northern Italian provinces, it does not fit the recent Southern Italian provincial data as well (Ballarino and Schadee, 2005). Some studies on Italian regions also look at the notion of territorial capital, defined as a “system of territorial assets of economic, cultural, social and environmental nature that ensures the development potential of places” and includes both tangible (such as physical capital) and intangible (such as human capital and institutions) private and public goods (Perucca, 2014). Specifically, there are several key parts of territorial capital that influence differential regional GDP growth the most—tourism attractiveness, “soft assets” such as social capital (i.e. institutions and human capital), and relational capital (i.e. high economic and human interactions, including exports). Physical capital is also an important, but not a sufficient factor for regional success (its separate effect is not statistically significant; Perucca, 2014). Several research investigations claim that the North-South differences in Italian development could be explained by such extreme factors as intelligence level (Lynn, 2006) and historic racial roots (Lynn, 2010, 2012; Piffer and Lynn, 2014) of the people living in these regions. Specifically, some authors argue that Italian regional data on people’s IQs show that there is a statistically significant correlation between the average regional IQ and regional per capita income, infant mortality, life expectancy, literacy, and years of education (Piffer and Lynn, 2014). However, the most recent research shows that these correlations hold for the current data rather than past data and instead serve as an indication of relationship between educational attainment (i.e. human capital) and socio-economic factors (Daniele, 2015). Finally, echoing findings from the China case study, the literature on Italian regional growth supports the hypothesis that financial development is an important contributor to enduring disparities. Local financial development, as measured by local bank competition or bank market power (Jayaratne and Strahan, 1996; Lucchetti et al., 2001; Guiso et al., 2004; Vaona, 2008; Coccorese, 2009), the role of cooperative banks, and bank credit volume granted to non-financial institutions, all significantly affect Italian regional GDP (Destefanis et al., 2014). Similarly, in terms of regional integration into global trade, studies show that the more a region is endowed with complementary sectors and industries (rather than specialized ones), the higher its local economic growth is (De Benedicts, 2007; Guerrieri and lammarino, 2007; Boschma and Iammarino, 2009). A mechanism behind this lies in the fact that, because related variety stimulates learning, regions with complementary economic sectors better absorb new extra-regional and external (international) knowledge than areas with highly specialized firms.
34 These insights mirror the literature on Chinese special economic zones and industrial clusters as channels for enhanced growth.
According to the Italian Constitution, its regions enjoy considerate autonomy from the central government. Italian regions, provinces and municipalities set their own legal statutes and define many functions of their public agencies (Article 114). Moreover, three Italian regions (Aosta Valley, Trentino Alto Adige and Friuli Venezia Giulia) and two islands enjoy even greater autonomy than other regions. Each regional statute defines the form of the local executive bodies, the Cabinet and the President, and the local legislative body, the Council (Article 121). Since the 1990s, Italy’s regional governance structure has been under reform with introduced changes to the local electoral system, government functions and overall jurisdictions. This reform is still under discussion.27 Italy’s public investments, made both on the national and local levels, have also played an important and illustrative role. In fact, the Italian Constitution states that local public investment decisions must include public administration, health care (the largest share of spending28), education, social services, housing, transportation infrastructure, and local economic development (article 117 and 119). The most recent data confirms these observations (see Figure 4.2.5 and Figure 4.2.6). In 2013, 47 percent of all local public expenditure went to health care, 7 percent to education, and social protection along with the environment got 5 percent each of all spending. In contrast, in the same year, the central government spent mostly on its general services (37 percent) and social services like social protection (22 percent), and less on education (11 percent) and health (11 percent). Interestingly, 10 percent of all national expenditure went to defense, national security and law enforcement while local governments contributed only 2 percent to these purposes. In contrast, China spent more than half of its national spending on defense and public safety in the same period. Italian regional distribution of public spending is far from even. If one accounts for the population size of each region, in 2011 regional governments spent between 4 and 8 thousand euro per person, except for Aosta Valle and Trentino Alto Adige autonomous regions. The latter two regions not only had very high per capita regional GDP, but also spent about 50 percent more per person than other areas, roughly 12-13 thousand euro per capita (see Figure 4.2.4). On the fiscal side, most of the regions do not have much legal power to set tax rates and 90 percent of the fiscal revenues come from the central government disbursements (Fiorino and Ricciuti, 2007). The distribution of fiscal authorities has also been under reform since 1990s. Based on Italian statistical data, Fiorino and Ricciuti (2007) write that: “in 1991, only 2.5% of
27 In fact, the proposed changes to the Constitution, that would reform the distribution of power between the central and regional governments, was rejected so far by the popular referendum in December 2016. 28 This reflection is consistent over time.
35 regional revenues derived from regional taxes and share of national taxes, while 97% was … from central government [grants]. In 1994, revenues from regional taxes and share of national taxes increased, representing 6.3% of the total, whereas grants from central governments decreased to 46% (Istat, 2000).” As with the local government structure reform, a fiscal decentralization reform is also still underway in Italy. The research on factors influencing disproportions in regional public investment is scarce. The literature shows that the regions that increased a number of their legislators based on the legal powers in their Statutes also experienced an increase in their local public expenditures (Fiorino and Ricciuti, 2007). Also, corruption may undermine the effect of public investment by reducing the regional GDP growth by 4 percentage points (Fiorino et al., 2012).
!
36 "# !
!!' " '
' #' ' ' %' ' '
"# !
$' &' ' "' !' "'
#'
#&' $' #'
37 It is estimated that, in the 1990s, the EU countries attracted about 40% of the world’s foreign private investment (or FDI). The major destination countries were the United Kingdom, France, Germany, and Spain. Italy, in contrast, experienced very few FDI inflows during that period, except for one region, Lombardy, which attracted about half of all Italy’s FDI inflows.29 Today there are still stark disparities in FDI inflows across Italian regions, but Lombardy is no longer a major recipient of FDI (see Figure 4.2.7). Lombardy and Piemonte are the most attractive regions for international capital in Italy’s North-West, but other regions in the North- East (Veneto and Emilia Romagna) and Center (Lazio) have caught up substantially. In fact, in 2011, Lazio region was the leading recipient of inflows across all regions.
29 ICE, Rapporto sul commercio estero 2004, Roma, 2004.
38 Recent data on companies with foreign participation in their workforces across Italian regions confirm these findings: even though Lombardy has the largest share of foreign firms, Lazio is not far behind (see Figure 4.2.8). These phenomena have puzzled researchers for many years and there is some literature on possible explanations, though the latest research on this topic dates back to early and mid-2000s.
!
Several authors show that a reason behind the difference in FDI inflows across Italian regions, in contrast to similarly developed EU regions, lies in the “country effect” (Basile et al., 2003). This effect mainly implies inefficiencies in Italian bureaucracy, weaknesses in its institutions and national public policies (such as tax regime, labor market regulations, legal system and property rights protection) as compared to other EU members (Nicoletti, 2002; Basile et al., 2005). Moreover, Basile et al. (2005) argue that Italy’s country-specific effect accounts for about a 40 percent loss in attractiveness to FDI. The literature also suggests other factors behind Italy’s regional differences in received FDI. In particular, foreign enterprises are more likely to be located in regions with a larger market size and market potential (or market access), bigger agglomerations, lower wage, higher R&D intensity and schooling, and better transportation infrastructure (Basile et al., 2005; Bronzini, 2004). All these findings are in line with factors behind foreign investment attraction outlined in Chapter 2 of this monograph.
39 The United States is a very large and diverse country. It is the third largest by land area and the third most populous country in the world. It is also, by many measures, the most developed economy on the globe. However, the U.S. regions differ substantially in their development. Some scholars even call the United States a “coastal nation,” as its most developed parts are predominantly located along the two oceans and Great Lakes coasts (Rappaport and Sachs, 2003), similar to China’s Eastern region along the Pacific coast. Administratively the country consists of 50 states and a federal District of Columbia (DC). It also has a number of dependent territories such as Puerto Rico and Guam. Each state contains counties and metropolitan statistical areas (MSA). According to the U.S. Bureau of Economic Analysis (BEA) definition, MSA is an “area consisting of a core county or counties in which lies an urban area having a population of at least 50,000, plus adjacent counties having a high degree of social and economic integration with the core counties as measured through commuting ties”.30 There are several classifications of regions within the U.S. This study uses the BEA approach and groups all the states and DC into eight regions. These groupings were “developed in the mid- 1950s, are based on the homogeneity of the states in terms of economic characteristics, such as the industrial composition of the labor force, and … demographic, social, and cultural characteristics.”31 (See Figures 4.3.1 and A3 in Appendix A for more details.) The regions are: • New England: Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, Vermont • Mideast: Delaware, District of Columbia, Maryland, New Jersey, New York, Pennsylvania • Great Lakes: Illinois, Indiana, Michigan, Ohio, Wisconsin • Plains: Iowa, Kansas, Minnesota, Missouri, Nebraska, North Dakota, South Dakota • Southeast: Alabama, Arkansas, Florida, Georgia, Kentucky, Louisiana, Mississippi, North Carolina, South Carolina, Tennessee, Virginia, West Virginia • Southwest: Arizona, New Mexico, Oklahoma, Texas • Rocky Mountains: Colorado, Idaho, Montana, Utah, Wyoming • Far West: Alaska, California, Hawaii, Nevada, Oregon, Washington The U.S. economy is one of the largest on the globe (the second after China by PPP approach32 and 40 percent larger than China based on market prices and exchange rates33), and is a leader in terms of technological and financial power (its high-tech industries and financial market are the most influential in the world). For example, the Great Recession, that started in
30 BEA Regional Definitions http://www.bea.gov/regional/definitions/ 31 Ibid. 32 CIA Factbook, 2016 33 World Bank, 2016 http://data.worldbank.org/indicator/NY.GDP.MKTP.CD?locations=CN-US&name_desc=true
40 the United States in 2008, also hit almost every country in the world. In fact, the 2008 recession was the deepest economic crisis in the U.S. since the Great Depression. However, the introduced monetary and fiscal policies enabled the country to start recovering in 2010 while the global economy felt the shock up until 201234.
The United States is a free market economy. It is a global exporter of telecommunication, transportation and energy products, computers, pharmaceuticals, and agricultural goods such as wheat, corn, soy, fruits and vegetables35. Its major trading partners are NAFTA members (Canada and Mexico), China, Japan and South Korea in the Asia-Pacific, and Germany, the United Kingdom and France in Europe36. While the U.S. economy is one of the most advanced in the world, its GDP per capita, $57,300 (2016, PPP) is the 13th in the countries rank.37 It also has considerable income inequality (as measured by Gini coefficient of 0.401, disposable income, post taxes and transfers38) and differences in its regional development (see Figures 4.3.2 and 4.3.3 for more details). For instance, in 2013 the states in New England and Mideast had about 40 percent
34 World Bank, 2015 http://data.worldbank.org/indicator/NY.GDP.MKTP.KD.ZG?end=2016&indicators=NY.GDP.MKTP.KD.ZG&locat ions=US-EU-1W&name_desc=false&start=2000 35 CIA Factbook, 2016 36 Ibid. 37 Ibid. 38 OECD Data, 2013
41 higher real per capita GDP than the Southeastern states and about 15 percent higher than the states in the Rocky Mountain and Great Lakes regions.39 The Far West region was consistently behind New England and the Mideast in terms of its economic development, but wealthier than the states in the Southwest and Plains. Moreover, a gap between U.S. regions has been rather stable over time, despite the fact that the estimated rate of convergence across states is about 2 percent per year (Barro et al., 1991; Garofalo and Yamarik, 2002; Turner et al., 2013).
$
#
"
!
" % !
39 Wyoming and North Dakota have the highest income across all states in these two regions.
42
There is a rich body of the literature about factors behind U.S. regional economic differences. For the most part, it highlights the regional growth variables summarized in Chapter 3 of this monograph and other case studies. Specifically, geography, TFP, human capital, migration, institutional quality, economic freedom and local public policies (such as facilitating business climate) are among the most widely studied factors (Rappaport and Sachs, 2003; Bologna et al., 2015; Neumark and Muz, 2016). Geography and history have long been hypothesized to be responsible for U.S. regional differences in development. For example, only 559 out of 3,144 U.S. counties,40 or 13 percent of the mainland U.S. area, lie within 80 kilometers from the oceans or Great Lakes coast, but these counties account for more than 50 percent of the U.S. total population and income (Rappaport and Sachs, 2003). Scholars find that coastal proximity augments local productivity and quality of life (as measured by population density), which in turn contributes to higher growth in these areas. However, this same research also indicates that controlling for location and prior conditions does not fully explain economic differences between U.S. coastal and inland counties (Rappaport and Sachs, 2003).
40 U.S. Census Bureau, 2013
43 Another geography-related factor includes local clustering of economic activities and consequent agglomeration.41 Urban economists argue that agglomeration benefits local economies by means of an increased scale and larger networks (e.g., a bigger market and tighter competition), information spillovers and positive social capital externalities (Goldstein and Gronberg, 1984; Helsley and Strange, 1990; Rosenthal and Strange, 2001; Latzko, 2013). As a result, economic interactions become more frequent and the region grows. In the U.S. context, there are two highly geographically concentrated sectors—manufacturing, and agriculture. For example, some counties in the Great Lakes, Mideast and New England regions (Baltimore, Philadelphia, New York, Boston, Cleveland, Detroit, Milwaukee, Cincinnati, and Chicago) are not only very prosperous but also have the highest manufacturing densities (Ellison and Glaeser, 1997; Rappaport and Sachs, 2003; Latzko, 2013).42 Other aspects of geography-induced differences between U.S. regions include the spillover effects of technological clusters and ‘knowledge hubs’, i.e. urban areas around universities and research institutions (Henderson, 1988; Krugman, 1991; Fujita and Mori, 1996; Fujita et al., 1999; Youtie, and Shapira, 2008). Technological clusters, along with geography, taxation, and supporting infrastructure, may be viewed as factors generally related to human capital. As in the other case studies in this chapter, human capital is an important contributor to the U.S. regional economic growth (Garofalo and Yamarik, 2002; Glaeser and Gottlieb, 2008; Glaeser and Redlick, 2009; Turner et al., 2013). However, a distinct U.S. feature is that the U.S. population is highly mobile. Consequently, human capital accumulation plays an even bigger role in divergent paths of poor and rich U.S. states as people with more human capital easily migrate to wealthier regions. Some authors note that high human capital mobility also effects social capital as poor places lack both and rich localities have abundance of both (Glaeser and Redlick, 2009). Silicon Valley represents a unique example of a technological concentration in the Far West region. The extensive literature examines the origins and evolution of Silicon Valley, including influencers on its successful economic performance (Saxenian, 1996; Porter, 1998; Swann, 1998; Kenney, 2000; Bresnahan and Gambardella, 2004). These studies raise several important insights into the disproportionately high growth of this region: (1) labor mobility as people constantly move from one firm to another firm to start-ups in pursuit of bringing their ideas to life; (2) an institutional setting that makes it relatively easy to open or close a business, facilitated by specialized legal practices; (3) venture capital that chases promising ideas, generated by mobile labor; and (4) knowledge spillovers or close interactions between researchers at Stanford University, University of California Berkley and Silicon Valley firms (i.e., a back-and-forth flow
41 Geographic clustering in the U.S. is also linked to its growing political polarization (see Sussell and Thomson, 2015). 42 Their manufacturing density partially declined over time (Dumais et al., 2002).
44 of ideas as local researchers and graduates become part of Silicon Valley industry and entrepreneurs share their knowledge and experience with faculty and students). Public polices, such as labor, tax and other regulatory rules, also play a significant role in states’ economic development. These policies characterize the quality of local institutions (Helpman, 1998), influence labor market conditions, and affect local attractiveness to domestic and international businesses (Neumark and Muz, 2016). Furthermore, local government size (as measured by local government employment) is negatively correlated with economic growth (Higgins et al., 2009). Finally, some research suggests that unionization laws may significantly influence regional business activities. Specifically, there may be fewer new firms established in states that have high unionization rates (Newman, 1983; Plaut and Pluta, 1983; Bartik, 1985). Figure 4.3.4 shows twenty-six U.S. states in the Southeastern, Southwestern, Rocky Mountains, Plains and Great Lakes regions where laws restrict labor unions (i.e. ‘right-to-work’ laws). However, if one compares the U.S. unionization map with the real per capita GDP and investment distribution across states, several states with high unionization rates,43 such as California, Washington and Oregon in the Far West region, Illinois in the Great Lakes region, New York and other states in the Mideast region have both high per capita GDP and high investment (see discussion in the following sections). This fact challenges the unionization argument and requires further investigation into this matter, which is beyond the scope of this case study.
43 “50 Years of Shrinking Union Membership, In One Map”, NPR Planet Money, February 23, 2015.
45 U.S. governmental system is highly decentralized with fiscal policies being set at three levels –federal, state and local.44 Each level has its own tax system and authorities. Overall, research suggests mixed findings about the effect of such fiscal decentralization on local economic growth with some studies indicating a positive correlation (Akai and Sakata, 2002; Stansel, 2005) and others finding either no correlation (Xie et al., 1999) or even a negative correlation (Higgins et al., 2009). If one looks at the federal and state budgets in 2013, there is an evident difference in spending distribution on the two levels (see Figures 4.3.5 and 4.3.6). For instance, the U.S. federal budget spending was mainly distributed between defense (similar to China’s case) and federal programs such as Medicare and Social Security (unlike China and Italy cases). The states’ budget expenditure, in contrast, was mainly focused on education (35 percent), welfare (31 percent), hospitals/health, highways and law enforcement (different from China and Italy cases). Also, a substantial share of the federal budget outlays (about 17 percent in 2011) was directed to local and state governments in the form of federal grants. These grants supported a wide range of state and local programs that were viewed to be beneficial but might have been underfinanced otherwise. For example, about half of these federal grants were allocated to state Medicaid programs.45
44 In contrast, monetary policy is set by the U.S. Federal Reserve System (on the federal level only). 45 Congressional Budget Office. Federal Grants to State and Local Governments. March 2013
46 !! #"
! &+ &+ $ *+ %+ &(+ '+
"# )+ &%+ *+ !#! # # ## $% "" $ " !
Local budget amounts clearly vary by state, as more economically developed states have more public resources than poorer states. Nonetheless, in several states, such as New Mexico, Louisiana, Mississippi, Arkansas, per capita state budget expenditure is rather high if compared to their real per capita GDP. In most other cases the overall regional distribution of public per capita investment roughly tracks the state income distribution (see Figures 4.3.3 and 4.3.7).
47 ! "
Because of the significant fiscal decentralization, local and state tax incentives are regarded to be important in economic development. For instance, many states offer tax abatements to businesses that relocate to them, as incentives. These incentives are efficient in theory (Glaeser, 2001) but have mixed empirical evidence (Hansen and Kalambokidis, 2010; Buss, 2001). If there is a positive effect of such policies on local economic development and employment, it is usually small and indirect (Wasylenko, 1997; Peters and Fisher, 2004). Moreover, one study of business executives finds that, even though local policy-makers often believe that tax credits spur private activity, in fact, executives argue that reduction in overall tax burden is what really matters (Gerking and Morgan, 1998; Jolley et al., 2015). In many instances businessmen do not even know about the existing tax abatements (Jolley et al., 2015). Finally, some studies look at the relationship between federal defense spending and local economic development and find a positive correlation between them (Atesoglu, 2004). For instance, military spending played an important documented role in the creation of Silicon Valley (Kenney, 2000). Yet, the literature also suggests that non-military public investment has a larger effect on local growth and investment (Atesoglu, 2004; Aschauer, 1989a).
The role of foreign investment in the United States has significantly grown over the past decades. The country shifted from being a net creditor (with net investment position equal to +11% of GDP) to being a net debtor (with net investment position equal to -20% of GDP). Moreover, foreigners, on average, earn lower returns on their U.S. assets than the U.S. investors earn abroad (Bosworth et al., 2007; Forbes, 2010). The former notion is puzzling, but might be explained by foreign investors’ motives to hedge against heightened uncertainties abroad, such
48 as those in Europe, Asia, Middle East, and Russia. In fact, more than 50 percent of FDI stock in the U.S. originates from the EU (including the Great Britain) and a bit less than 10 percent from Canada.46 Additionally, foreign-owned enterprises (FOEs) are not uniformly located across the U.S. (see Figure 4.3.8). As a matter of fact, in absolute terms, FDI is concentrated in just two states, California and Texas, and two regions, the Great Lakes and Mideast. In terms of employment, California, Texas, and New York also have the largest number of jobs in FOEs, but FOEs employ relatively more in New England (e.g., Massachusetts, Vermont) and the Southeastern states (e.g., North and South Carolina). (See Figure 4.3.9.) The major industries of these FOEs are manufacturing (about 38 percent) and services (about 57 percent). The latter include mainly trade and retail, banking, professional, scientific and technical services (Saha et al., 2014).
$!$ &""$!%& $!! # $" "" $$"#""$ ! !# %"# # # ###" !# " #" ! " $
46 BEA, 2015. 47 The last year when data is publicly available. Data includes gross book value of property, plant, and equipment.
49
There are many hypothesized factors that influence the distribution of foreign investment across U.S. regions. A number of studies find that agglomeration or clustering of economic activities (in addition to being a factor in regional economic differences in general, as in the cases of China and Italy) play a significant role in foreign investment location decisions (Bobonis and Shatz, 2007; Shaver and Flyer, 2000; Head et al., 1999, 1995). Indeed, FDI regional distribution echoes the geography of U.S. industry clusters, such as manufacturing- heavy areas in the Great Lakes and Northeastern regions and technological clusters in California and Texas. Research also indicates that the largest U.S. MSAs have about 75 percent of jobs in FOEs (Saha et al., 2014). Additionally, since U.S. regions are more or less homogeneous institutionally, the literature suggests that, besides institutional factors (Abdioglu et al., 2013), decisions to invest in the US regions are based chiefly on a business approach. Specifically, the related factors include considerations of local market potential (including local infrastructure and market size), the quality of local labor force and corresponding labor regulations (Friedman et al., 1996; Bartik, 1985; McConnell, 1980; Little, 1978). This may explain the attractiveness of MSAs to foreign investors, as large metropolitan areas have also large markets and draw more educated labor force. Finally, local tax policies also affect decisions to invest into a particular locality within the U.S. For example, targeted tax policies and foreign office representation (in the investor's home country) appear to have a positive effect on FDI attraction (Bobonis and Shatz, 2007). Studies
50 also find that a high state corporate tax rate discourages foreign investment (De Mooij and Ederveen, 2003), whereas labor and capital subsidies have almost no effect on bringing more foreign capital (Bobonis and Shatz, 2007).
China, Italy, and the United States are three examples of countries with pronounced regional differences where similar combinations of factors have influenced the movement of international capital and development. Despite having very different histories, cultures, and socio-economic systems, all three countries have substantial intra-national variation. Their regional economic disparities are commonly attributed to several key factors, including local geography, human and social capital, institutional quality, infrastructure, public policies, and clustering of economic activities. However, the relative importance of these factors considerably varies across the countries. In China, local geography and specific government policies, such as creation of SEZs and industrial clusters, played an important role in divergent paths of its Eastern developed region and the rest of the country. The same is true for the inflow of foreign capital that was motivated by the preferential regulatory regimes of SEZs and industrial clusters and presence of the large sea ports in proximity. In the Italian case, the quality of social and human capital, local institutions, and the level of corruption are seen as instrumental in driving the substantial regional disparities between the Northern and Southern provinces, both in economic development and attractiveness to foreign investors. The United States has also highly unequal regional development. Its pre-existing regional disparities are reinforced by high labor mobility, clustering of economic activities and agglomeration: rich states along the coasts draw more human capital and get richer, and poor inland states become poorer. Similarly, foreign capital is also attracted by big markets that are concentrated in the country’s Mideast and Great Lakes, as well as in California in the Far West and Texas in the Southwest. The three case studies inform my analysis of the Ukrainian case, reported in Chapter 6. First, I discuss the data I used to construct a detailed picture of the Ukrainian economy, with all its regional variety.
51
This dissertation uses secondary data for the analysis of international capital flows and regional economic growth. I have constructed a unique dataset that includes both socio-economic indicators and foreign investment data for Ukrainian regions from a number of sources. The data are the most representative and reliable available, yet they have the shortcoming that they do not reflect Ukraine’s shadow economy.48 The constructed dataset spans 15 years, from 2001 to 2015. However, for the model estimates in Chapter 6, I use the data between 2001 and 2013 because the unusual events that started in December 2013 (the Maidan Revolution and subsequent annexation of Crimea and war in the Eastern Ukraine) would likely obscure any underlying economic relationship between Ukraine’s regional economic growth and international capital flows.
The socio-economic statistical data for Ukrainian regions mainly come from the “Regions of Ukraine” and “Ukraine” statistical yearbooks assembled by the State Statistical Service of Ukraine (SSSU). The data cover all 24 oblasts of Ukraine, the autonomous republic of Crimea, and two cities of national subordination, Kyiv and Sevastopol, and span the period between 2001 and 2015. The data are obtained from enterprises, organizations, and individuals using surveys and administrative data, and compiled by the ministries and departments. The “Regions of Ukraine” yearbook has been available annually since 2004 and covers the period since 2001. The “Ukraine” yearbook has been assembled annually since 1993. The latter yearbook provides a more comprehensive account of Ukraine by various social and economic dimensions but includes mostly national data. The “Regions of Ukraine” collection provides information on the same indicators at the regional (oblast) level and, hence, is used as a source for most of the socio-economic indicators in this analysis. In some rare cases when there were slight discrepancies in the data between the two yearbooks, I gave preference to the regional data. For the purposes of this analysis, I use the following key regional socio-economic indicators: • Regional GDP or Gross Regional Product (GRP) is the measure of regional (oblast) economic development. The GRP is calculated as the sum of regional value added of all
48 The estimates of Ukraine’s shadow economy vary from 30% to 50% across years and sectors. For more information see: “General Trends in Shadow Economy in Ukraine”. Ministry of Economic Development and Trade of Ukraine, 2016; and Vinnychuk, Igor, and Serhii Ziukov. “Shadow Economy in Ukraine: Modelling and Analysis”. Business Systems and Economics. Vol. 3 (2), 2013.
52 goods and services, including taxes and subsidies.49 Since the GRP is counted in Ukrainian local currency, hryvnia, the indicator was converted to US dollars based on the average exchange rate in a given year as reported by the National Bank of Ukraine. The nominal GRP was adjusted for inflation from the previous year based on the GDP deflator, as reported by the World Bank for the period50. • Domestic Capital Investment includes the annual total capital investment flow in tangible and intangible assets (equipment, buildings, transport, land, agricultural stock etc.) by businesses, national and local governments, and population for housing and construction. Since this indicator is accounted in Ukrainian local currency, hryvnia, it was converted to US dollars based on the average exchange rate in a given year as reported by the National Bank of Ukraine. • Domestic Mergers and Acquisitions (M&A) include all the domestic M&A deals for the period and come from the InVenture Investment Group dataset (described in more detail below). • Population is measured as the number of people living in an oblast in a given period. It is based on a Ukraine’s 2001 census, adjusted for natural population growth/decline and migration records from administrative data. • Labor force is calculated according to the International Labor Organization’s methodology as the number of persons aged 15-70 who are on the labor market and includes employed and unemployed people. • Education is a key indicator of human capital and is measured as the number of students enrolled in higher educational institutions per 10,000 people.51 • Regional Human Development is measured by the Human Development Index (HDI) from the SSSU “Regional Human Development” statistical bulletin. The regional HDI between 1999 and 2011 is essentially based on the United Nations Development Program’s (UNDP) methodology. It includes nine blocks of indicators—demography, labor market, income and welfare, living conditions, education, health and healthcare, social environment, environment, and public finance. HDI is between 0 and 1 with higher value indicating enhanced human development. • Crime is a measure of local institutional development (including law enforcement).52 The indicator covers the total number of all detected crimes according to the Ukraine’s
49 The GRP was introduced with the regional accounts in 2004. Before 2004, the regional GDP was measured as a difference between the sum of regional value added of all goods and services and intermediate consumption. 50 In order to compare international capital flows with Ukraine’s regional GDP, the two indicators should be measured in similar units. Since the investment data, as discussed in other sections of this chapter, are in current US dollars, I consider mainly regional GDP in current US dollars rather than in local Ukrainian currency. Additionally, in order to reflect the annual price change in the economy, regional GDP is adjusted for inflation (i.e. year-to-year change in price level). At the same time, I assume that foreign capital inflows in a given year are not subject to local inflation for that year. 51 Note that the literacy rate, which is a standard measure of educational achievement, is not relevant for Ukraine as its literacy is 99.8% (CIA Factbook, 2015) due to the mandatory primary and secondary education. 52 Note that Ukraine is a unitary republic with little policy discretion at the local level. However, the effectiveness of local law enforcement varies substantially across oblasts and regions.
53 Criminal Code, based on the records of local police offices.53 Additionally, I calculate regional crime rates as the number of cases per 100,000 population. • Number of financial institutions (entities) in an oblast is a measure of the regional financial market development and includes the total number of registered business entities in the financial sector (e.g., banks, insurance, credit unions). The indicator is available between 2001 and 2012.54 These data are also used to compute the oblasts’ shares in the national financial market for the investment allocations in the banking and insurance sectors. Additionally, I calculate the number of financial institutions per 100,000 population • Number of public libraries is used as a measure of regional cultural engagement in an oblast.55 Additionally, I calculate the number of public libraries per 100,000 population. • Voter turnout is used as a measure of regional civil engagement (per Putnam, 1993). The data are from the Ukraine’s Central Voting Commission and cover the years of presidential and parliamentary elections (both regular term and snap elections) – 2002, 2004, 2006, 2007, 2010, 2012, and 2014. Table 5.1.1. summarizes each of the regional indicators described above, except for those that are available at the oblast level only such as HDI and voter turnout (see Table B1 in Appendix B). Both tables (5.1.1 and B1) show that Ukrainian regions are economically and socially heterogeneous. For example, during the study period their population sizes varied from about 9 million (West, 2013) to 14 million (Center, 2001) and their labor forces differed by about 1-2 million. Also, across all regions, the labor force was, on average, less than a half of the population, which indicates that, for every Ukrainian employed, there was usually more than one person out of the labor force. The average regional GDP for the period varied from 14 (West) to 37 (Center) billion USD in total or from 1,337.21 (West) to 2,740.32 (Center) USD per capita. At the same time, the Ukraine’s overall per capita GDP was 2,166.66 USD. However, between 2001 and 2013 the four Ukrainian regions grew at a mean double-digit rate of about 16-18 percent in nominal terms and 4-6 percent in real terms. This rapid economic growth signifies both the regions’ low starting point and their potential for growth.56 In this analysis, the Central region includes eight middle-sized oblasts of Ukraine, rich in black soil reserves, oil and gas (in Poltava oblast), and provides high-end services in its major city, Kyiv (such as R&D, pharma, IT, and space aircraft). The Eastern region includes three densely populated oblasts, all sharing their borders with Russia, that were a primary target for the
53 Since 2005 the data also includes the tax evasion crimes and since 2013 the records from the Prosecutor General of Ukraine. 54 The reporting method had changed since 2013. Hence for my model estimation I assume there were at least as many financial institutions in oblasts in 2013 as in 2012. 55 Mass adoption of digital technologies and Internet in Ukraine began in late 2000s. As reported by Ukraine’s Internet Association, the internet penetration (users older than 15 years of age) was 28% in 2010 and 50% in 2013. Before 2010s public libraries played a much more important role in local cultural development and engagement than has been recently observed. 56 This observation is consistent with the economic growth literature, see Chapter 3 for more details.
54 Russian destabilization efforts in the spring of 2014. The Southern region, which includes Crimea, is similar to the Central region in terms of size, natural resources and high-quality service sector in its two major cities, Dnipro and Odesa. Finally, the Western region is the least populous region, even though it spans over eight oblasts. This region is a Ukraine’s major tourist destination (along with Crimea), provides services in its urban areas (e.g., software development), has some oil and gas reserves in the Carpathian Mountains, and some manufacturing of parts and appliances.
$ " ! !! !
" !" " #" !" " # "" %$ $ $ ") %$ !*() # "" %$ $ '& ) $ ") %$ !*() $ '%,) - ')%- ' $ '%,) $ &' ( $ %# () & )"$+ ()# $)# "" %$ $
55 1/)45,' 07)45/)05/,..,10" )%0 ,0%0',%.045,565,1042)3 )%0 126.%5,10/,..,10 )%0 %&1313')/,..,10 )%0 (6'%5,10456()0542)3 )%0 3,/)3%5)'%4)42)3 )%0 ,&3%3,)42)3 )%0 &.%4540'.6()( #1.80 #,00854,% 10)54- 3,/)% &.%454 !3%04'%32%5+, $+851/83 6+%04- 0,231 3,/)% %7%01 8,7 +%3-,7 $%213,9+,% 3%0-,74- ,3171+3%( 8-1.%,7 7,7 ,70) 1.5%7% ()4% !)3012,. 6/8 +)3410 +/).0854-8, +)3-%448 +)30,754, +)30,+,7 " 65+13:4'%.'6.%5,104&%4)(10(%5%*31/ 5%5) 5%5,45,'4 )37,')1*"-3%,0)0#)0563)07)45/)053162
56 In 2001-2013 the Central region was the most economically developed in both aggregate and per capita terms (see Table 5.1.1, Figures 5.1.1 and 5.1.2). It is also the most populous region, though its oblasts are, on average, smaller than those in the Eastern region (see Appendix B). During the study period, it had the most developed financial market, as measured by the local financial institutions, and received most of the domestic capital investment (almost half of all the national investment) and domestic M&A investment (more than a half of all the country’s M&A capital). Its domestic M&A deals were focused primarily on agriculture and services and unevenly distributed across its oblasts. This region also experienced, on average, the highest economic growth with the Kyiv, Vinnytsia and Kirovograd oblasts being at the lead (see Figure 5.1.3). It also had the highest level of human capital as measured by the number of university students (and HDI, see Appendix B). The crime rate, a measure of regions’ institutional strength, was below the national average crime rate, although it was higher than that in the West.
The Southern region has been historically approximately on par with the Central region in terms of its population size and economic structure. However, in 2001-2013, its mean regional GDP, both aggregate and per capita, was considerably lower than that of the Center (see Table 6.1.1, Figures 5.1.1 and 5.1.2). It received the second largest amount of domestic capital and M&A investment after the Center, though, in per oblast terms, this investment was similar to the average across all oblasts (see Appendix B). Socially, across all regions, it experienced the highest mean crime rate, and low voter turnout, and had a smaller than average number of both libraries and university students (even though it has a large number of higher educational institutions). The Dnipro, Odesa and Mykolaiv oblasts were the principal contributors to the
57 region’s economic development. These three oblasts, with ports on the Black sea and the Dnipro river, are the leading transportation and logistics hubs, and the centers of high quality IT services and manufacturing (see Figure 5.1.2 and Figure 5.1.3).
!
Between 2001 and 2013 the Eastern region was the second most prosperous in terms of mean per capita regional GDP, after the Central region.57 The Donetsk and Kharkiv oblasts accounted for the biggest share of the region’s economic success (see Figure 5.1.2). Additionally, the region’s economic growth rate was somewhat slower than that in the Center, with an evident slowdown since 2011 (see Figure 5.1.1 and 5.1.3). A possible cause of the slowdown was that Ukraine’s mining industry, which is located predominantly in the Donetsk and Luhansk oblasts, struggled to remain competitive on the global market. In contrast, the high-end services (such as IT, pharma, and logistics) and the agricultural sector, that are located mainly in the Central and Southern oblasts, improved their global competitiveness over time and attracted substantial domestic and foreign investment. In terms of local financial market development, domestic capital and M&A investment, the Eastern region lagged considerably behind the Center and the South. Moreover, for the period its domestic investment was, on average, less than a quarter of the total national investment. Most of
57 However, across all Ukraine’s oblasts, its three oblasts were, in fact, among the most economically developed (see Appendix B).
58 the local M&A deals were done in heavy industries such as mining and chemicals production. Likewise, even though the Eastern oblasts had a slightly higher than average number of university students, the region lagged behind in terms of HDI and had the lowest number of public libraries. As for its local institutional development, it had the second highest average crime rate after the South.
!
The Western region was the least populous and the least economically developed among all the regions between 2001 and 2013.58 Its mean regional GDP was less than the national average and more than 50 percent less than that in the Center, even though its economic growth was comparable to that in other regions (see Tables 5.1.1, Figure 5.1.1 and 5.1.2, Appendix B). Figure 5.1.2 shows that, in per capita terms, the Lviv (bordering Poland) and Ivan-Frankivsk oblasts contributed the most to regional prosperity, and Figure 5.1.3 indicates that the Chernivtsi (next to Romania and Moldova) and Rivne (next to Poland) oblasts grew the most in this region. The Western region received the smallest amount of domestic capital investment and very little domestic M&A capital. But in terms of social development and human capital, the region had, on
58 It is noteworthy that the Western region has the highest number of labor migrants abroad across all Ukraine regions. It is also estimated that the Ukrainian labor emigrants annually transfer about 5 billion USD in remittances to their families back home (International Organization for Migration, Mission in Ukraine, 2011).
59 average, the largest number of public libraries, the second highest HDI (after the Center), the highest voter turnout and was the safest region. Thus, in 2001-2013 the Ukrainian regions differed significantly in their socio-economic development. The Central region was the most advanced. The Southern region was not far behind the Center with respect to its economic prosperity, but its institutions and human capital were considerably weaker. The Western region was the least economically developed, but it had a high level of human capital, institutional capacity and social engagement. The Eastern oblasts were the wealthiest among all the Ukrainian oblasts, but they grew more slowly over time than either the Central or the Southern regions, and stagnated since early 2010s.
My study analyzed both private and public international capital flow data for Ukraine, recorded at the local level. The private foreign investment data includes FDI in the form of equity and debt and the M&A deals, and come from two sources: the “Investments and Foreign Economic Activity” statistical yearbook compiled by the SSSU and the “Mergers and Acquisitions in Ukraine” dataset purchased from Ukraine’s InVenture Investment Group. The public foreign investment data cover projects in Ukraine by two public financial institution—the European Bank for Reconstruction and Development (EBRD) and the National Endowment for Democracy (NED). All capital data span the period between 2001 and 2015. If the investment amount was specified in euros, it was converted to US dollars using an average exchange rate in a given year.
Foreign direct investment or FDI is an investment by a resident of one country in a business entity in another country with a managerial and/or ownership control of at least 10%. Per the SSSU’s methodology, “all other transactions between a direct investor and a direct investment enterprise are [also] classified as direct investment (for example, lending).” However, “debt liabilities between separate affiliated financial corporations, including banks, are not classified as direct investment.”59 FDI is tracked by the SSSU in the “Investments and Foreign Economic Activity” since 1995. The yearbook includes annual FDI stock (both as equity and debt) in Ukraine by target region, country of origin, and economic sector, accounted in US dollars. To calculate the annual FDI flow, I use a difference in FDI stock between January 1st of the two consecutive years.
59 These are included in ‘other investment category’ in the SSSU reporting. State Statistics Service of Ukraine. Methodological Commentary on External Sector of Ukraine Statistics. 2017
60 The data are based on the reporting from enterprises corrected for the difference between real and nominal prices of shares and other types of capital (based on the IMF methodology since 2003). The data are used by the National Bank of Ukraine for country’s balance-of-payments and international investment position calculations. Based on these calculation, during the core study period, FDI constituted about 16-37 percent of all Ukraine’s financial liabilities.60 It is worth noting that the FDI data in “Investments and Foreign Economic Activity” are available on the oblast level only and, hence, do not permit tracking down any specific investment. This caveat introduces challenges in FDI data analysis and interpretation because there is no way to infer for sure ‘true’ FDI from abroad, i.e. FDI from foreign investors rather than Ukrainian oligarchs’ capital reinvested via offshore zones. As noted in Chapter 2 about 37 percent of FDI stock in Ukraine in 2013 was from offshore zones, Cyprus and British Virgin Islands. In contrast, the M&A data (discussed below) do not have this disadvantage. Figure 5.2.1 shows the stock of FDI as equity and debt across the Ukrainian regions in 2013. The Center received the largest FDI amount during the study period by both types of financing – about 32 billion USD as equity and 5 billion USD as debt. The South was the second destination for the FDI, but its investment stock was less than a half of that in the Center – 13 billion USD as equity and 2.5 billion USD as debt. The East received about 8 billion USD in total FDI, a quarter of the attracted capital by the Center. Lastly, about 5 billion USD were invested in the West, similar to the Eastern region, even though the Western region was significantly poorer than the Eastern region during the period. Table 5.2.1 shows a mean and a standard deviation of regional FDI inflows as debt and equity for the thirteen-year period. On average, Ukrainian regions received about a billion dollars of FDI in equity and about 200 million dollars FDI in debt annually. In per capita terms, across all regions, on average about 80 dollars were invested in equity and 17 dollars in debt. Also, about 48 percent of all foreign investments in Ukrainian regions (in my dataset) were FDI as equity and 9 percent were FDI as debt. Table 5.2.1 also illustrates that between 2001 and 2013 the Central region, on average, attracted the largest annual FDI flows – about 2.3 billion USD as equity (or 170 USD per capita) and 400 million USD in debt (or 30 USD per capita). Also, this investment constituted more than 60% of the region’s total foreign capital. The Kyiv, Cherkasy, Poltava and Zhytomyr oblasts received a bulk of these flows (see Figures 5.2.2 and 5.2.3).
60 Other financial liabilities included portfolio investment (10-17 percent) and debt liabilities (50-70 percent). The latter cover predominantly the Ukraine’s central government debt and bank loans. The former is not directly related to the regional economic differences because it likely settled almost entirely in Kyiv, the capital city (where the central government is). The latter, bank credits and loans from abroad, may influence regional economic differences. Since all the banks have their headquarters in Kyiv, this capital probably remained to a large extent in there. However, if it did reached regions, the number of local financial institutions such as bank branches (a measure of the local financial market development I use in my modeling) might be a good proxy to understand how this money affected local economic development.
61 '$ "& &" " ! #' &*! &$"%% $ ! ! "!% "!
32'-081" *"3* 2(-,1! 1$#-,# 2 %0-+ 2 2$ 2 2(12("1 $04("$-% )0 (,$
($ !!' ! ")%$"%% $ ! ! "!%
! &"$ %& !&$ "'& %& "&
/3(26+(**(-, $0 .(2 /3(26 $0"$,22-**-0$(&,,4$12+$,2 $!2+(**(-, $0 .(2 $!2 $0"$,22-**-0$(&,,4$12+$,2 $ , %-02'$"-0$123#6.$0(-# -06$ 015(2',$2-32%*-5(,%*-5 +-3,25 1 113+$#2-!$7$0- 32'-081" *"3* 2(-,1! 1$#-,# 2 %0-+ 2 2$ 2 2(12("1 $04("$-% )0 (,$
The Southern region attracted much less FDI than the Central region (about a half of the Central region’s amount), but its FDI were about the Ukrainian regional average. Figures 5.2.2 and 5.2.3 show that the Dnipro, Crimea, Odesa and Zaporizhia oblasts absorbed most of these capital inflows. Crimea is particularly remarkable here because it did not attract a lot of other international capital flows (see discussion below) and may serve as an example of the above- mentioned data caveat. Additional analysis by a destination economic sector and a country of
62 origin revealed that the FDI stock in tourism and recreation industry, which is a major one in Crimea, came primarily from the offshore countries (Cyprus and British Virgin Islands) and Russia. Consequently, one could assume that most of the FDI in Crimea were either domestic Ukrainian or Russian capital flows. The Eastern region received less than a half as much FDI as equity as the South and a similar amount of FDI as debt. The Kharkiv oblast led the region’s FDI attraction. Curiously, during the period, much of the FDI stock in heavy industry came from Luxemburg, the origin of the investment activities by System Capital Management (or SCM) company. This company is owned by a Ukrainian oligarch Rinat Akhmetov and controls assets in metals, mining and power generation, located primarily in the Eastern region. As a result, one could assume that FDI in the Eastern region partially represented Ukrainian domestic capital.
$!# ! # " $#%! ""! " "#"
63 " ! ! ! !
Finally, the Western region lagged considerably behind other regions in FDI inflows. On average, in a given year only about 25 USD per person were invested as equity and 7 USD as debt in this region (see Table 5.2.1). The Lviv oblast received most of this capital. It is also noteworthy that overall, in the study period, much more FDI inflows were in the form of equity than debt for all regions. But if one looks only at the capital flows as debt, four Ukrainian oblasts stand out – Kyiv in the Center, Dnipro in the South, and Donetsk and Luhansk in the East (see Figure 5.2.3). The Central region had acquired the largest debt, but, across all regions, its debt presented the smallest share of its total foreign investment (see Table 5.2.1).
“Mergers and Acquisitions in Ukraine” dataset was acquired from the Ukraine’s InVenture Investment Group. It lists foreign and domestic M&A deals between 2001 and 2016 and includes the following information: date, target company, seller company, bidder company, bidder headquarters location, and, where available, deal monetary value and target ownership share.61 Based on this available information, I expanded the dataset to include the location of the target company key operations in Ukrainian oblasts. In most cases I also searched the web for the
61 InVenture Investment Group gathers this information from various sources, including official press releases and media articles about the announced deals. I also performed a selective cross-validation check of these data based on the Thompson Reuters information on M&A activity globally.
64 information on the bidder companies to identify their true country of origin. For example, in most cases when Cyprus was listed as the bidder headquarters location, it turned out that in reality these companies were either Russian or Ukrainian (predominantly Ukrainian). As a result, many bidder company locations were updated to reflect the reality. Also, some duplicate deals were removed from the dataset. If one compares the M&A dataset to the FDI data described above, the two datasets are most likely overlap. The FDI data probably include Ukraine’s domestic capital investments (through offshore zones) and certainly foreign investment in companies with at least 10% control. The M&A data include company purchases of other businesses with at least majority control (50%+)62. As a result, the M&A data usually include larger capital investment than FDI (though less frequent) and, arguably, better reflect the reality with respect to the capital countries of origin. The final M&A dataset includes 356 M&A deals: 297 between 2001 and 2013 (study core period), and 59 in 2014 and 2015. Of the 297 deals in the core study period 159 transactions were made by foreign investors (54 percent) and 138 by Ukrainian investors (36 percent). Out of all foreign M&A deals between 2001 and 2013, 46 or 29 percent were made in financial sector (banking and insurance) and the rest, 113 or 71 percent – in other economic sectors such as agriculture, pharmaceuticals, IT, manufacturing, retail, real estate. Out of 46 foreign M&A deals in the financial sector, four were made with regard to local banks—two in Odesa (the Southern region), one in Kharkiv and one in Donetsk (the Eastern region)—and the rest concerned banks or insurance companies operating nationwide. Similarly, out of 138 domestic M&A transactions, 18 deals or 13 percent were made in the financial sector (two with regard to local banks in Dnipro and Kharkiv) and 120 deals or 87 percent—in other economic sectors such as agriculture, heavy industry, mining, power generation, retail, media, real estate and recreation/tourism. For analytic purposes, since nationwide banks and insurance companies are all headquartered in the Kyiv city, the general M&A investment in these sectors were allocated to oblasts based on their financial market development measure. Figure 5.2.4 compares the total cumulative foreign and domestic M&A investment across the regions between 2001 and 2013. The Center received about 14 billion USD in foreign M&A capital, the largest amount across all the regions. The South was the second destination for foreign M&A investment and accumulated more than 9 billion USD of this capital. The Eastern region attracted less than a half of the amount invested by foreign businesses in the Center, about 6 billion USD. But the biggest contrast was the Western region, which received the least foreign M&A capital—about 2 billion USD. Also, M&A investment by Ukrainian businesses was significantly smaller than that by foreigners, and the regional pattern for domestic M&A was similar to that for foreign M&A: the Center received the most, 8 billion USD, the South and the
62 In several rare cases the acquired ownership share was less than 50%, e.g. when a controlling company purchased additional shares of the target business.
65 East lagged behind with 3 and 2 billion USD respectively, and the West attracted less than a billion USD.
$! # $ $ #% ! "# %"# #! "" ! "
If one looks at annual flows, the oblasts in the Center got, on average, about a billion USD in M&A capital per year, of which 70 percent were in non-financial sector (see Table 5.2.2). However, this investment accounted for about 20 percent of all foreign capital flows received by this region, a considerably smaller share than that for the East (about 29 percept) or the South (about 24 percent). In per capita terms, about 79 USD per person were on average invested in the Center annually. The Southern region was consistently the second destination for M&A capital with 730 million USD mean annual inflow, of which 72 percent flowed to real economic sectors (see Table 5.2.2). The Eastern region received less than the country’s average M&A flows, 437 million USD, and the Western regions lagged behind substantially with about 145 million USD (or 13 USD per person) in foreign M&A investment per year, a quarter of the Ukraine’s average.
66 &" % !'#"!## " ! #
$!" #$ $" !%$ #$ !$
5'1#,,.5'23-'.3-+,,+/.! ./.+.#.%+#, '%3/1-+,,+/.! '1#0+3#.5'23-'.3! '1%'.33/,,/1'+)..5'23-'.3 '#. (/13*'%/1'234&60'1+/& ! 43*/172%#,%4,#3+/.2$#2'&/.#(1/-."'.341'.5'23-'.31/40
Figure 5.2.5 shows the distribution of per capita foreign M&A investment in non-financial sectors across Ukrainian regions and oblasts between 2001 and 2013. Evidently, the Kyiv (in the Center), Dnipro and Zaporizhia (in the South) and Donetsk (in the East) oblasts were key target locations for foreign M&A capital, followed by the Kharkiv (in the East), Odesa and Mykolaiv (in the South), Poltava (in the Center) and Lviv (in the West) oblasts. Kyiv city and oblast attracted 36 percent of all foreign M&A investments in non-financial sectors during the period. The variety of target areas spanned across a whole spectrum of economic activity: from agriculture, food and beverages, to logistics, retail, real estate and IT. The investment originated mostly from the EU countries, the United States, and China, and about a quarter of deals were signed with Russians. Another most targeted oblast in this region was Poltava oblast where M&A investment concerned oil and gas drilling and agriculture and came mainly from the EU countries. In the Southern region, the most lucrative deals concerned the privatization of several big metallurgic SOEs—“Kryvorizhstal” in the Dnipro oblast (acquired by a German investor), and “Zaporizhstal” and “Zaporizhia Aluminium Factory” in the Zaporizhia oblast (both acquired by Russian investors). Besides these big transactions, most of the M&A capital in this region targeted agriculture, power generation and transportation infrastructure. Also, this foreign capital originated principally from the EU countries, Canada, the United States and China, and only a few deals in the region were done by Russian investors. Curiously, during the whole period, in Crimea there were just four deals—three originating from Russia and one from the United States. This observation is very different from the FDI data described above, where Crimea received a relatively large per capita FDI equity stock.
67 &# & & % '!# '$% % ! %!#$# " %#!$$ # ! $ $%$
In the Donetsk oblast in the Eastern region, almost all the M&A capital went to heavy industry. Moreover, out of the total 12 deals in this oblast, 9 were by Russian businesses. A similar dynamic was observed in the Luhansk oblast where 5 out of a total 6 deals in non- financial sectors were by Russian investors. In contrast, the Kharkiv oblast (another oblast in the Eastern region, bordering Russia) had a different experience—less than a half of its foreign M&A capital in non-financial sectors came from Russia (5 out of total 11 deals) and the rest were from the Western European countries and Turkey. Finally, the Lviv oblast in the West received the largest share of the M&A capital in this region. The investment concerned oil and gas drilling, retail and services (including IT). Other deals in the region were done in agriculture and power generation infrastructure. In terms of the countries of origin for this capital, the Western oblasts’ experience echoed the one of the Central and the Southern regions.
International public investment data in Ukrainian regions come from two institutions, the EBRD and NED, and are publicly available. The EBRD and the NED perform very different activities in Ukraine. Whereas the EBRD provides funding to private enterprises and local public institutions (in most cases in the form of loans that have to be repaid or minority equity
68 positions63), the NED essentially disburses the U.S. Department of State foreign aid (in the form of grants) by supporting local non-governmental organizations (or NGOs), civic initiatives, independent media, and human rights groups.64 The EBRD data include all the projects it financed in Ukraine since 1996. For each project they list its date, title and information about a recipient organization—its name, economic sector and ownership (public/private). The status of each project (completed, signed, disbursing, repaying) is also provided. Each project has its own web page where one can find its detailed description, the EBRD investment amount and other technical details. For the purposes of my analysis, I used the subset of these data spanning the period between 2001 and 2015. Based on the given information in each project description, I expanded the dataset to include the location of key operations of the target organizations. The NED data contain all the funded projects between 2001 and 2015. Additional information includes the fiscal year in which the project was conducted,65 the recipient organization, the purpose (e.g. media, human rights, elections etc.), a brief description, and the funding amount. As with the other datasets, I expanded the NED data to include the location of the recipient organizations. Also, the project activities/initiatives performed by the American or European organizations locally in Ukraine were excluded from the dataset because these funds most likely never reached Ukrainian citizens. The final dataset for the core period (2001-2013) includes 158 EBRD projects and 394 NED projects. Even though the number of the NED projects is significantly higher than that for the EBRD, their funding amount is much smaller. Between 2001 and 2013, the EBRD provided about 7.2 billion USD in investment to Ukrainian entities in the non-financial sector (124 projects or 78 percent) and about 1.6 billion USD in the banking sector (34 projects or 22 percent).66 The target economic sectors were agriculture, manufacturing, infrastructure and transport (including local public transportation projects), natural resources and power generation. During the same period, the NED disbursed about 18 million USD via grants to Ukrainian organizations.
63 Additionally, the EBRD can help hedge the financial risks associated with projects. It also works with local banks to provide loans for small and medium enterprises. For more information see http://www.ebrd.com/what-we- do/products-and-services.html. 64 Of all the U.S. foreign aid channels, in my research I consider only NED investment because their projects usually target community activities in different localities across the country, whereas other funding sources (such as USAID) focus primarily on the overall institutional development. Since Ukraine is a unitary country where policy- making is concentrated in its capital city, Kyiv, a potential effect of the USAID investment can’t be examined on the regional level. 65 For the purposes of this analysis I assigned the same calendar year to each project as the fiscal year, assuming that most of the actual project funding was disbursed in the corresponding calendar year. 66 Similar to FDI and M&A data, the EBRD investment in the banking sector were allocated to oblasts based on their financial market development measure, except for three project that targeted local banks in the Lviv and Kharkiv oblasts.
69 Figure 5.2.6 shows the distribution of the EBRD and NED investment across the regions. Yet again, the Central region received most of the investment of this type—more than 3 billion USD from the EBRD (or 34 percent of all funding) and 12 million in NED grants (or 67 percent of all funding). However, the investment pattern for other regions differs from that observed before. For example, the EBRD invested almost as much in the Southern region as in the Central region (almost 3 billion USD). Moreover, the Western region received almost twice as much EBRD funding as did the Eastern region—1.8 billion USD versus a billion USD respectively. The NED projects, on the contrary, focused more on the Eastern region than on either the Southern or Western region, and provided about 3 million in grants to this region. In contrast, the Western region received very few NED funds—a bit more than a million USD or 7 percent of the total investment during the period.
# # # "$ $!" " !! !
If one looks at the annual public capital flows across Ukrainian regions between 2001 and 2013 (Table 5.2.3), the regional pattern becomes somewhat different. The Central region received slightly less EBRD financing per capita than the Southern region—17.92 USD versus 18.08 USD respectively. Also, the share of this capital type in total foreign investment inflow was smaller for the Center (about 16 percent) than for the South (about 20 percent). Another curious fact is that, compared to other types of capital, the West received, on average, about 139 million USD in total EBRD capital, which was more than its mean annual M&A and FDI debt inflows. Additionally, in this region this investment accounted for almost a third of all its foreign capital. In contrast, the Eastern region received the least of the EBRD project financing, less than a half of the Ukraine’s average (and about a half in per capita terms).
70 &" % $" $ % ! $ '#" ## " #
$ " #$ $" %$ #$ $ 3%/!** ,3%01+%,1+)**)-, ,-,),!,#)!* %#1-/+)**)-, %/!.)1! ,3%01+%,1 %/#%,11-**-/%)',,3%01+%,1 3%/!**,3%01+%,1+)**)-, %/!.)1!,3%01+%,1 %/#%,11-**-/%)',,3%01+%,1 %!, &-/1(%#-/%012$4.%/)-$ 21(-/50#!*#2*!1)-,0"!0%$-,$!1!&/-+ !,$
As for the NED investment flows, their amounts were very low across all the regions, even though the Center stood out as a main recipient of these funds. In a given year between 2001 and 2013, a bit less than a million USD was invested on average in the Central region (or 7 cents per person), a bit less than a quarter million flowed to the Eastern oblasts (or 2 cents per person), and about 100 thousand USD was invested annually in the West and the South (1 cent per person). Interestingly, overall the NED funding flows have an observable trend during the study years. For instance, the total grant amount increased more than two-fold after the 2004 Orange revolution, was considerably small during 2008 financial crisis year, and rose significantly since 2010, when Ukraine’s former president Yanukovych was elected (especially in Crimea and the Donetsk and Luhansk oblasts, areas of Yanukovych’s main electorate). Similar to other investment data sources described above, Figures 5.2.7 and 5.2.8 illustrate per capita international public investment by the EBRD and NED across regions and oblasts during 2001-2013. The EBRD invested predominantly in the Kyiv and Zhytomyr oblasts in the Center, the Rivne and Lviv oblasts in the West, in all oblasts in the South (except Crimea), and to some extent in the Donetsk oblast in the East. In particular, the Western oblast received loans mainly in manufacturing, infrastructure, natural resources and power generation. For example, several projects concerned modernization of two nuclear power plants (and related power infrastructure) that are located in the Rivne and Khmelnytskyi oblasts in the Western region. Also, several projects targeted the construction of a new freeway that connected Kyiv city with Warsaw in Poland, a part of the infrastructure investment before the 2012 European Football Championship. Besides these infrastructure credits, the Zhytomyr and Kyiv oblasts received a large amount of capital for their agricultural businesses and various manufacturing companies.
71 (% (!( ' ) ") &'! "'& ""#" "" '#%& % $ '%#&& % " " #"&" &'&
(% (!( ' ) ") &'! "'& % $ '%#&& % " " #"&" &'&
The EBRD projects in the Southern oblasts focused almost universally on agriculture and infrastructure, such as upgrading ports on the Dnipro river (the Dnipro and Kherson oblasts) and
72 on the Black Sea (the Mykolaiv and Odesa oblasts). There was also considerable investment into local public transportation and infrastructure in the Dnipro city. In the Eastern region, the EBRD projects focused mostly on power and energy (especially, renewable energy) and natural resources, such as mining and heavy industry, in the Donetsk and Luhansk oblasts, while the Kharkiv oblast attracted capital mostly for its agricultural and manufacturing businesses. Figure 5.2.8 shows that the NED projects, in the contrary to those conducted by the EBRD, primarily focused on the Central and Eastern oblasts, especially those bordering Russia. For instance, in the Luhansk oblast, these projects focused predominantly on democratic participation and civil society, whereas in the Donetsk and Kharkiv oblasts, besides these same activities, the grants were also disbursed to the local Committees of Voters and human rights groups. In the Central region, the NED funded many organizations in the fields of research and education, especially in the Chernihiv oblast and the Kyiv city, and regional youth initiatives (e.g., in the Sumy oblast). In the Mykolaiv oblast, a major recipient of the funding in the South, the projects focused predominantly on local initiatives. Also, as mentioned above, Crimea became a major destination for NED investment only since 2010. Its projects have focused mainly on local media and human rights, and have often supported Crimean Tatars (Crimean native ethnicity). Another interesting observation is that these funding channels to Crimea did not stop after its annexation by Russia, though the recipients’ names became concealed. Finally, in the Western regions, the NED backed mostly educational institutions (in the Lviv oblast), the local Committees of Voters and human rights groups.
Socio-economic and investment data for Ukrainian regions between 2001 and 2013 show clear differences across the four regions. The socio-economic indicators included in my study are regional GDP (total, growth, and per capita), population size, labor force, measures of human capital (students enrolled in tertiary educational institutions and HDI), institutional development (crime), local financial market development (institutions in financial sector) and civic engagement (public libraries and voter turnout). The investment data consist of both domestic (local domestic investment in buildings and equipment, domestic M&A deals), and international capital flows in the form of FDI as equity and debt, international M&A capital, and public investment by EBRD (credits) and NED (grants). The data reveal a systematic pattern in the socio-economic development and attractiveness to foreign capital of the four regions. Specifically, between 2001 and 2013, the Central region was consistently the most prosperous and received most of the private and public foreign capital. The Southern region has steadily become the second in economic development, though it lagged considerably behind the Center with regard to its institutional development. In contrast to the Center and the South, the picture was mixed for the Eastern region. Even though its oblasts were rather wealthy relatively to other parts of the country, since 2011 the region was on a declining
73 path. It also systematically fell behind in terms of attracted foreign capital, except for NED public financing. Finally, the Western region was the poorest across all the Ukrainian regions, but had high level of human capital and civic engagement. The oblasts in the West consistently were not attractive to foreign capital flows with an exception of EBRD financing.
74
As noted in the introduction, the hypothesis of this dissertation is that international capital flows, depending on their types and amounts, disproportionately affect regional economic growth. Additionally, I hypothesize that this relationship is bidirectional, i.e., that regional economic growth also affects international capital inflows. In this chapter, I develop a model to test this hypothesis, based on the neoclassical theoretical growth model discussed in Chapter 3. I specify the model, based on the literature review on factors influencing economic growth and foreign investment (Chapter 3 and case studies in Chapter 4), and then estimate it empirically, using the data on Ukraine’s regions. This chapter describes the model and its results, including heterogeneity over time and across urban/rural areas.
A starting point for my model is the neoclassical growth model (per Solow, 1959, and Swan, 1959), which is based on the production function approach with constant returns to scale. Assuming a Cobb-Douglas production function with two inputs, labor and capital, GDP of oblast i in year t is: (1)
where Ait is technology level, Kit is total capital stock, Lit is labor, α and (1-α) are corresponding elasticities of oblast GDP with respect to inputs. Rewriting equation (1) in natural logarithmic monotonic transformation yields:
(2) for oblast i in year t. As noted in Chapter 3, in the Solow-Swan model capital input accumulation is the crucial factor in economic growth, while advances in technology are assumed to be exogenous. Hence,
Ait is Hicks neutral (not affecting capital and labor shares in the production function) and gt assumed to grow at a constant exogenous rate g, i.e. Ait=Ai0e . Similarly, labor input grows at an nt exogenous rate n, i.e. Lit=Li0e . Hence, in equations (1) - (2), growth of output over time, before reaching its long-run equilibrium, is crucially dependent on accumulation of the capital input.67 However, while converging to this equilibrium based on capital accumulation, every economy has its own path, conditional on its specific characteristics and factors (i.e., a conditional convergence). Barro and Sala-i-Martin (1995) show that such conditional convergence exists across U.S. states and Japan prefectures and is 2-3 percent. Their findings imply that the gap between states/provinces in these countries is reduced by half every 25-35
67 Including human capital, based on the Solow-Swan model extension.
75 years. However, the data from the case studies (Chapter 4) and Ukraine (Chapter 5) reveal that, whereas the GDP of provinces within regions converge over time, the gaps between regions have persisted and been quite stable over time. That is why I assume that each province/oblast converges to the regional, rather than overall nationwide equilibrium, i.e. that there exists a conditional regional convergence. But these long-run regional growth paths may be influenced by changes in exogenous factors. Consequently, regional growth over time may be expressed as
a function of the initial level of regional output, Yj0, and a long-term regional equilibrium path,
Yjt* (using Barro’s (1996) approach): (3)
where Yjt* is a function of a vector of choice and environing variables, Xit, affecting each * oblast i within region j. The regional growth rate is diminishing in Yj0, given Yjt (signifying * conditional convergence of oblast within regions), and increasing in Yjt , given Yj0, over transitional period to the equilibrium. The latter notion means that there could be a number of factors in policy and institutional domains that could positively affect regional economic growth in each period. Moreover, the vector of environing variables, Xit, can vary both by region and by oblast, i.e. there might be region-specific characteristics affecting regional growth and oblast- specific features influencing their rate of convergence to the regional equilibrium. But in the
Ukrainian context, since its regions do not have substantial policy autonomy, Xit only varies by oblast. Combining equations (2) and (3) and using the Taylor approximation theorem, each oblast’s
growth can be expressed as a percentage deviation from its regional equilibrium Yjt*, given its initial economic development, inputs and environmental variables:
(4)
where Rj is a regional indicator variable. Based on the literature review in Chapter 3 and case studies in Chapter 4, it is evident that the capital input in equation (4) should be very broadly defined, and cannot be seen as exogenous. The notion of capital includes different kinds of investment such as those by private entities and public institutions, domestic and international. As mentioned in Chapter 1, private investment can also take various forms, e.g., direct investment, mergers and acquisitions, via debt or equity. When one disentangles domestic and foreign capital, equation (4) becomes:
(5)
where α1 and α2 are corresponding oblast output elasticities of domestic capital, Dit, and foreign capital, Fit(k), of type k. Furthermore, all case studies examined reveal a systematic relationship where the most developed regions are also those that attract the most international capital. Consequently, the model in equation (5) should be augmented to explain both the regional differences in attracted capital and the endogeneity of regions’ capital accumulation. To do this, I introduce an interaction variable between the regional indicator variable, Rj, and foreign capital input
76 accumulation, Fit(k) (by type, k). This new variable takes into account the differential growth effect of attracted foreign investment for each region j:
&# , * # * &# * &#-!. * + # * # * * &#! (6) Second, to recognize the possible endogeneity between international capital and regional economic development, I expand the model by adding another equation that identifies the foreign capital accumulation by analyzing an investor’s decision-making. To understand an international stakeholder’s decision to invest in an oblast i in year t, I look at the case of foreign merger or acquisitions of a domestic firm where a competitive foreign- owned enterprise maximizes its local profit, i.e. difference between domestic output produced and inputs costs (domestic labor and own capital of type k) for oblast i in year t:
# , # # + # # + #-!. #-!. (7)
where Yit is a Cobb-Douglas production function with diminishing returns to scale, inputs Lit
and Fit(k) are nonnegative, and (pit, wit, rit) is a vector of exogenous prices. Solving the problem in (5), one gets an equation defining the capital-labor ratio as a function of exogenous variables: $ % ')-(. , ') (8) ') "')-(. % where β and (1-β) are corresponding capital and labor elasticities of output. Given exogenous and Hicks neutral technology, the ratio of elasticities in equation (8) is constant over time. As a result, the investment choice depends fundamentally on the marginal productivities of local
labor, wit, and the opportunity cost of capital, rit(k). Yet, there are also a number of local environing factors influencing these inputs’ marginal productivities, such as local preexisting economic conditions, education, labor laws, financial market development, liquidity and risk. Rewriting equation (8) in natural logarithmic monotonic transformation and rearranging yields: % , - .+ - .* * (9) #! # # #-!. # % #
where Hit and Zit are the vectors of factor variables influencing local labor and capital marginal productivities. In the Ukrainian context, labor factors include human, social and institutional development, while capital cost factors are local preexisting economic conditions (i.e. regional growth) and financial market development (i.e. those affecting liquidity and financial risks). Adding the regional differential effects in equation (9) and combining it with equation (6) yields the following system of equations:
&# , * # * &# * &#-!. * + # * # * * &#-!. % , * * * * * * #-!. % &# &# # # &# (10)
77 where Xit is a vector of common environing exogenous factors for regional economic growth and foreign capital accumulation and Zit are exogenous factors specific to foreign investors’ financial decision-making. Model (10), then, is the system of equations describing the relationship between growth and foreign capital flows in a region. In the next section, I estimate the model empirically using Ukrainian regional data.
My empirical estimation starts with examining basic properties of Ukrainian oblast output (per equation (2) in previous section). Since my hypothesis primarily concerns variations between Ukrainian oblasts and regions over time, I use the random-effects panel approach because it allows variations both within and between observations.68 In addition, I eliminate any spurious trend between the output and its inputs by controlling for time effects. The results of this estimation are in Table 6.2.1 column (1). Table 6.2.1 shows that, similar to the finding of past research, capital input69 contributes about one third (.37) to oblast GDP and labor accounts, and twice as much in the oblast output (.79). Both of these elasticities are highly statistically significant. Also, the oblast GDP has had a positive statistically significant time trend for the study period. Finally, the random-effect estimation approach has a high goodness-of-fit, both within and between oblasts. Expanding the notion of capital, I add a human capital variable, as measured by the number of students in tertiary educational institutions, to the model. The new estimation results are in Table 6.2.1 column (2). Consistent with the literature, human capital plays an important role in oblast GDP. In fact, its elasticity is almost as large as that of the physical capital input—.28 vs. .30—and is highly statistically significant. Additionally, the inclusion of human capital considerably reduces the labor elasticity—from .79 to .56—signifying that human capital, combined with the labor force, enhances its quality and input in GDP. Next, I disaggregate the oblast capital input to domestic and foreign capital and estimate the model again. The results in Table 6.2.1. in column (3) show that the elasticity of domestic capital, .24, is much larger than that of foreign capital, .05, and both are statistically significant. The elasticities of other inputs, labor and human capital, remain almost identical and statistically significant. Also, the time trend remains positive and statistically significant.
68 In contrast to the fixed-effect approach which focuses on specific predetermined characteristics within observations. For more information see Gujarati et al. (2009). 69 In this part of analysis, the capital input variable pools together oblast domestic and foreign capital.
78 !""! $! #" " !
/)!/3#,#0+3#, /)/-'23+%#0+3#, /)/1'+).#0+3#, /)#$/1 /) 34&'.32 !+-'!1'.& ')+/. '.3'1 /43* #23 † /.23#.3 "+3*+. '36''. 5'1#,, !'0'.&'.35#1+#$,'+2,/)/(/$,#23 "'23+2#1'('1'.%'1')+/. 3#.& '11/12/(3*'%/'((+%+'.32#1'+.0#1'.3*'2+2 0 0 0 † 0
Finally, I look at any predetermined regional effects on oblast GDP by adding regional categorical variables to the model where the West is a reference region. The results of this estimation are in column (4) of Table 6.2.1. All the regions70 have statistically significantly different from zero fixed characteristics that positively contribute to their regional GDP, as compared to the West. This means that the oblasts in the Western region may have unique features that differentiate them from other Ukrainian oblasts and regions, and that are not captured by observable economic data. The results also indicate that the three other regions have statistically equivalent contributions to their regional GDP (see Table C in Appendix C1 for
70 The Eastern region is marginally statistically different from the Western region (at 10% significance level).
79 pairwise comparisons). Thus, three regions, Center, South, and East, have similar effects on regional GDP and their effects differ from, and are larger than, the effect of the Western region on local GDP.
Once I examined basic properties of Ukraine’s oblast GDP, I estimate the regional economic growth model with domestic and foreign capital flows as in equation (6) to see whether the rate of change in oblast GDP depends on region effects. Similar to my previous analysis, I start by first looking at aggregate capital flows and regional growth differences associated with this cumulative investment. Additionally, I include a regional institutional development variable as measured by crime incidence. Since Ukraine is a unitary country with few policy-making authorities at the oblast and regional level, the work of local law enforcement (in countering local crime) is a good proxy for local institutional capacity. Previous evidence on the role of crime in economic growth is mixed with some studies finding a negative correlation (e.g., Detotto and Otranto, 2010) and some indicating that such correlation depends on crime activity location within a country (e.g. see the research on Mexican states in Pan et al., 2012). It is important to note that the Solow-Swan growth model focuses primarily on output and capital per unit of effective labor. Since there are no good measures of effective labor (especially in Ukraine’s statistical data), a typical empirical solution is to examine the output growth in per capita terms and the capital accumulation as a ratio of output (e.g., see Barro, 1996). However, there are a number of important caveats to this approach. Firstly, it assumes that each individual, including children and the elderly, are part of the effective labor force and have the same productivity. Second, using capital not per unit of effective labor but rather as a ratio to output ignores a fundamental capital-labor trade-off in output decision-making. Third, this approach does not require measurement of other contributing growth factors in units of effective labor, resulting in a probable systematic measurement error. In contrast, this dissertation accounts for the labor input in general by focusing on the labor force rather than population. Previous research used population growth as a proxy for labor force growth (e.g., see Hanushek and Kimko, 2010). But general population decline (as in case of many developed countries and Ukraine as well) may not necessarily signify a decline in the labor force as employees may decide to retire later and/or increase in immigration may offset the reduction in natives’ birth rates. Also, I account for the labor-capital trade-off in the firm’s investment choice and, consequently, include the labor force and its quality measures into the international capital equation (see below). Table 6.2.2. presents the results of the regional economic growth model random-effects estimation. I performed the Dickey-Fuller unit root test to determine whether the oblast GDP time series are stationary. The test showed that these series are difference stationary with a drift (which is why I also include the constant). Also, the GDP and capital flows time series have an observed discontinuity in 2009 when Ukraine’s economy was badly impacted by a global
80 financial crisis. I control for this fact by including an indicator variable for 2009 in my estimation.
81
4 "04$-$1,4$--07 0.(34,&$1,4$--07 02(,*/$1,4$--07 "04$-$1,4$--07(/4(2 "04$-$1,4$--07!054+ "04$-$1,4$--07$34 02(,*/$1,4$--07(/4(2 02(,*/$1,4$--07!054+ 02(,*/$1,4$--07$34 $%02 !45'(/43 2,.( (*,0/ (/4(2 !054+ $34 2,3,35..9 0/34$/4 #,4+,/ (47((/ 6(2$-- "(1(/'(/46$2,$%-(,3*2074+0)-0*0%-$34 --&0/4,/50536$2,$%-(3$2(),234',))(2(/&(3 (8&(14)02 4 $/',/-0*$2,4+.3#(34,3$2()(2(/&(2(*,0/!4$/'$2'(220230)4+(&0()),&,(/43 $2(,/1$2(/4+(3,31 1 1 † 1
82 Column (1) in Table 6.2.2 shows that, across Ukrainian oblasts, a one percent increase in total pooled investment (both domestic and foreign) is associated, on average, with about .09 percent economic growth in the Western oblasts. This effect is statistically significant, slightly higher for oblasts in the Southern and Central regions and smaller for the Eastern oblasts, compared to the Western region (though the regional differences are not statistically significant, see Table C2a in Appendix C). Increase in crime and labor force growth is positively correlated with oblast economic development, but not statistically significantly so. Growth in human capital has the highest effect on oblast economic growth—one percent growth in the number of students is associated with .17 percent oblast economic growth. Table 6.2.2. column (2) lists the results of a similar random-effects estimation approach, except that the capital flows are disaggregated to domestic and foreign. The latter includes FDI (as equity and debt) and public investment by EBRD and NED.71 The results indicate that domestic capital is strongly statistically significant and is positively correlated with oblast economic development. Moreover, a one percent increase in these local investments is, on average, associated with about than .09 percent economic growth. In contrast, the effect of foreign capital flows, even though positively associated with oblast economic growth, is much smaller. Compared to the Western region, the effect in the Southern oblasts is the largest—a one percent increase in foreign capital flow is associated with almost .02 percent economic growth in those oblasts (see Table C2b in Appendix C for more details). For other growth factors, none of them are statistically significant in this specification, including human capital. The results in Table 6.2.2. illustrate that there are several robust factors contributing to local economic growth such as local preexisting conditions and exogenous shocks to the economic system (e.g., crises).72 However, discrepancies between capital flow estimates in column (1) and (2) indicate that the analysis of the effect of investment on regional economic growth depends on the types of capital considered, including types of foreign investment. I examine the notion that international capital flows are likely endogenous, i.e. dependent on regional economic growth, in the following estimations.
As previously discussed in Chapter 5, I consider a number of international capital flows received by each oblast—FDI, M&A, EBRD and NED investment. To estimate the impact of these various foreign investment on Ukraine’s regional economic growth, I use a simultaneous equation model (SEM), a standard method for analyzing economic variables that are jointly
71 I do not include M&A capital in the total foreign investment because there is considerable overlap between FDI and M&A data (see discussion in Chapter 5). Consequently, the foreign capital estimates of this model may be viewed as lower bounds of true effects. 72 Note that in contrast to level models in Table 6.2.1, both models in Table 6.2.2 show a poorer ‘between’ fit than ‘within’ fit, indicating that fixed effects approach might be more appropriate than random-effects when analyzing the rate of change in regional GDP.
83 dependent.73 SEM is a generalization of seemingly unrelated regression (SUR) and two-staged least squares regression (2SLS) methods. Whereas SUR assumes there are no endogenous regressors and 2SLS does not account for correlations of the error terms in different equations, SEM allows for both endogenous regressors and correlations of the error terms in different equations.74 The specified SEM follows the system of two equations in model (10), a regional growth equation and an international capital attraction equation (see below). The second equation in the SEM resolves the endogeneity of international capital variable in the economic growth model and uses instruments (Zit) identified from the case studies—social capital (following Putnam, 1993) and local financial market development (following Hermes and Lensik, 2003).
The selection of these instruments was done based on two factors: (a) the availability of relevant data in the Ukrainian context, and (b) the possibility of both time and regional differences. Data availability made it infeasible to include wages in the model. There are several reasons why the wage statistic, as reported by the SSSU, can’t be used for the purposes of this research. Firstly, it does not include employees at small businesses. Second, due to a high share of informal pay in Ukraine (including both in the shadow untracked economy and official business activities), the official wage statistic is not a good measure of labor productivity. Third, the reported wage is a yearly and monthly average which can’t be disaggregated to the hours effectively worked. Fourth, the SSSU substantially changed its data collection methodology in 2010 that made the wage statistics incomparable for years before and after 2010. As a result of these considerations, rather than using the SSSU reported wage, I use the labor force and student population as general measures of labor input and its quality. The model’s requirement for data that vary across time and oblasts/regions precludes me from using variables that are available only on the national level, such as inflation. The only regional inflation indicator reported by the SSSU is a consumer price index (or CPI), i.e. a price level of a narrow bundle of consumer goods. For the purposes of my analysis, a good measure of prices in the capital equation would be prices as faced by producers. The latter statistic is only available on Ukraine’s national level and, consequently, does not vary by region or oblast. Overall the SEM estimation results are consistent with previously discussed general conclusions (see Tables C3-C9 in Appendix C). Several robust factors such as local preexisting conditions, exogenous shocks to the economic system (both negatively related to oblast
73 For more information see Gujarati et al. (2009). 74 I report and compare the results of all three estimation approaches: SEM, SUR and 2SLS. See Appendix C for details.
84 economic growth), and regional predetermined characteristics (positively related to oblast economic growth) are statistically significant and consistent across the considered specifications. Also, increase in crime incidence is negatively related to local growth (except for models with M&A and FDI debt) and human capital is positively associated with growth, though its statistical significance varies. For domestic investment, their importance in oblast economic growth differs based on a type and amount of foreign capital – the larger the foreign capital flow is relative to domestic capital, the lesser the role of domestic capital is. For the factors behind regional attractiveness to international capital flows, local financial market development is a strong positive predictor of foreign investment. In contrast, social capital, as measured by public libraries and voter participation, is a weaker instrument, even though when it applies (in election years), voter turnout is consistently negatively associated with attracted foreign investment. Other important variables include human capital and crime. On the one hand, the effect of human capital depends on a measure under consideration, with an increase in student population being negatively associated with capital inflows, but improvement in HDI being positively related. On the other hand, change in crime incidence is repeatedly positively related to international capital flows, except for the M&A equation. Since crime, as reported by the SSSU, covers all types of crimes including tax evasion, this observation possibly illustrates a previously mentioned limitation of FDI data: that it includes Ukrainian oligarchic capital reinvested via offshore zones (likely for the purposes of tax evasion). Yet, the positive correlation between the change in crime prevalence and foreign investment holds even for the public capital flows. While this relationship in the EBRD case is ambiguous, the NED foreign aid often targets localities with weak institutional development, such as feeble law enforcement. Tables 6.2.3 and 6.2.4 summarize the core results of the SEM estimation for the relationship between regional economic growth and international capital flows. Table 6.2.3 shows regional variations (rows) in the effects of different international capital flows (columns) on economic
growth. It also lists p-values for these regional effects’ joint significance, where H0 is that all the corresponding coefficients equal to zero. Similarly, Table 6.2.4 shows regional variations (rows) in the effects of economic growth on different international capital flows (columns) and related p-values for these regional effects’ joint significance. Among private capital flows, increases in FDI inflows have the largest effect on economic growth across all regions. Moreover, all four regions are statistically different from each other in their corresponding growth effects (based on pairwise test of corresponding regression coefficients75). For instance, a one percent increase in FDI equity flows is associated, on average, with .37 increase in the Western oblasts’ growth. In contrast, this result is much smaller for the other regions: .13 percent for the Center, .15 percent for the South and .11 percent for the East. The latter notion signifies how much the Western region lacks FDI, i.e. a small absolute increase in such investment would greatly help this region’s development.
75 For full pairwise regional comparisons see Tables C3a-C3f in Appendix C.
85 In the second equation, regional economic growth is positively correlated with FDI inflows. Furthermore, this connection is strongly jointly statistically significant, i.e. as hypothesized, there is indeed a simultaneous relationship between economic growth and FDI. One percent growth is associated, on average, with about 7-9 percent increase in FDI inflow across all regions. However, this effect is not statistically different across regions. Hence, conceptually, the Western oblasts should be able to attract as much FDI as other oblasts as long as they grow comparably. But as these oblasts economically fall behind other regions, their attracted FDI inflows are consistently smaller.
" *% $* ($* %$" & *"$ "%,)%$!( $ .) %$"%$%# (%,* - -&
'+ *- '+ *- * $ * !*+ !&+!) ',+$ *+ '%&+%#&%"% & !)' $% ""! + $ &#!+ & ) !))')* )!%&( )!&+$!*%*
" *% !( $ .) %$"%$%# (%,* %$$* ($* %$" & *"$ "%,)- -&
'+ *- '+ *- * $ * !*+ !&+!) ',+$ *+ '%&+%#&%"% & !)' $% ""! + $ &#!+ & ) !))')* )!%&( )!&+$!*%*
In contrast to equity investment, FDI in the form of debt (i.e., credit between FDI-affiliated firms) has no statistically significant effect on regional economic growth, but these inflows are
86 affected by regional development. Moreover, the relationship between FDI debt and growth is negative and statistically significant. For example, a one percent economic growth in the Eastern region is associated with a 13 percent decline in an oblast’s borrowing. At the same time, oblasts in the Center are less likely to reduce their debts with economic growth as compared to other regions. When FDI equity and debt are combined, the joint effect is similar to that of equity investment. Specifically, the total FDI is positively correlated with regional economic growth and this relationship is statistically significant across all regions. The Western oblasts experience a mean .33 increase in growth as a result of one percent increase in cumulative FDI inflows—the largest impact as compared to other regions. It is noteworthy that the effect of total FDI for the Eastern oblasts is three times smaller than that in the West, even though these oblasts attracted barely twice as much capital as the West (see discussion in pervious Chapter 5). This means that a relatively small absolute increase in FDI would not generate as much growth in those oblasts as in the West, i.e. given existing technological level (or total factor productivity, TFP) there are negative returns to scale. As for the impact of economic growth on total attracted FDI, likewise FDI equity, it is positive, statistically significant and similar across all regions—one percent oblast growth, on average, results in about 4-5 percent more FDI inflows. M&A capital, which is a subset of FDI but better defined (see detailed discussion in Chapter 5), affect regional economic growth similarly to FDI equity.76 It is positively associated with regional growth and statistically significant. The Western oblasts are again different from the rest of the country—one percent increase in M&A is, on average, associated with .22 percent growth. For other regions this effect is .09-.10, even though their attracted M&A amounts vary substantially. Also, economic growth is not correlated with a regional propensity to attract M&A deals. The latter observation indicates that local economic growth might not be a crucial reason for acquisition of foreign assets. Finally, for public investment, both the EBRD and NED investment are positively correlated with regional economic growth. Their corresponding effects are statistically significant and the largest for the oblasts in the West. In fact, one percent increase in the NED projects funding is, on average, associated with .44 percent growth in the Western oblasts. The EBRD financing is only half as effective on the Western growth. Hence, in the West, even very small relative increase in foreign aid can have a significant positive effect on development. For other regions, the EBRD and NED funds impact growth in similar ways. One percent increase in EBRD and NED investment in these regions result in corresponding mean .06 and .08 growth. There is a simultaneous relationship between regional growth and the receipt of public funds for both EBRD and NED investment flows, but the effects have the opposite direction. While high economic growth is strongly positively associated with the receipt of EBRD funds, it is negatively linked to the receipt of NED grants (and only marginally statistically significant). A
76 In model specification with M&A, domestic M&A through offshores are added to domestic capital investment.
87 one-percent increase in growth in the Western oblasts is associated, on average, with 20 percent more EBRD capital flows, whereas in the other regions, a one percent increase in growth is associated with only 15-16 percent more EBRD capital. In the case of NED funding, this relationship is reversed; the more developed a region becomes the less foreign aid it attracts. For the Western oblasts, this effect is the strongest. A one-percent increase in growth there is associated with a 10-percent decline in NED project funding, whereas other regions would experience only a 4-to-5-percent reduction. Moreover, the Eastern oblasts are least likely to face a decline in NED foreign aid, perhaps because there are strategic or political reasons for foreign aid flows into this region. Thus, the findings suggest that my hypothesis of the differential effect of international capital flows on regional economic growth can’t be rejected. Moreover, there is indeed a simultaneous nexus between foreign investment and economic development and it depends on the type of investment under consideration. My results also suggest that the cyclical patterns of foreign investment described by Stiglitz (2010), especially for private debt, likely hold in the case of Ukraine. Finally, the nexus between foreign investment and economic growth is consistently distinct for the Western oblasts, where any small increase in international capital flows, either public or private, would create the largest effect on local economic development. At the same time, among the other three regions, the Eastern oblasts, on average, experience the smallest economic boost from a marginal increase in attracted private capital flows. The latter conclusion is consistent with observed stagnation of this region since 2011 (see Chapter 5 for detailed discussion).
In this section, I examine the heterogeneity of my main findings over time and across urban/rural areas. Specifically, I study whether and how my core results change if one looks at the effects of international capital flows on regional economic growth over medium term, or if one adds urbanization to the system as in (10). The goal is to provide a more nuanced picture of the nexus between foreign investment and regional growth. The presented supplemental explorations are separate from one another.
My core model examines the simultaneous relationship between various types of international capital flows and regional economic growth on average during 2001-2013. In this section, I describe the heterogeneity of the growth and capital effects over time by adding lagged capital and growth variables (and their region interactions) to my SEM. Since my core study period is relatively short, I include two lags (T-1 and T-2) in addition to the baseline (T-0). These auxiliary variables provide insights about the concurrent relationship between different foreign investment and regional development over the medium-term (1-3 years).
88 Tables 6.3.1 and 6.3.2 summarize the main results of this medium-term model (see full results in Appendix C Table C10). Table 6.3.1 shows regional variations (rows) in the effects of different international capital flows (columns) on economic growth for each time period up to 3 years (T0, T-1, and T-2). It also lists p-values for these regional effects’ joint significance, where
H0 is that all the corresponding coefficients equal to zero for the corresponding type of capital and period. Similarly, Table 6.3.2 shows regional variations (rows) in the effects of economic growth on different international capital flows (columns) and related p-values for these regional effects’ joint significance for each time period up to 3 years (T0, T-1, and T-2). As in my main model, cumulative FDI (equity and debt) has a strong statistically significant medium-term effect on regional economic growth.77 However, now this effect varies both across regions and over time. In the baseline year, total FDI has a negative correlation with oblast growth for all regions with the Western oblasts experiencing the strongest effect. For the next periods, this effect flips its sign and becomes positive for all regions, except for the South. In the Southern oblasts, FDI starts to enhance growth only 2 years after its initial inflow. In contrast, in the Western oblasts, the effect remains positive. In the Central and Eastern regions, growth rates decline again two years after the initial boost from FDI. In the second equation, regional economic growth is still correlated with cumulative FDI. Furthermore, this correlation again depends on both region and period. In baseline and subsequent year, it is not statistically significant. But after two years (T-2), higher economic growth does lead to higher FDI in three out of four regions, especially in the West. For M&A capital, the results are analogous to those for cumulative FDI flows; however, its effect on regional growth is distinct in several ways. First, it is strongly statistically significant for the whole medium term. Second, its dynamic changes from negative in baseline, to positive in a subsequent year, and to negative again in T-2. Third, the direction of the M&A effect is the same for all regions, with the largest impact in the West. In the opposite direction, growth leads to increased M&A investment, and the increase is most likely to occur a year after the spike in growth, especially in the Western and Central oblasts. As for the effects of public capital, EBRD investment enhances regional economic growth only in the short term. Moreover, a region’s propensity to attract EBRD financing is also highest in the short run and it declines substantially over time. In contrast, NED investments do not have a significant effect on growth in the medium term, nor does growth have a significant effect on future NED investments in the medium term. This may be because it can take considerable time to realize the potential of foreign aid, or because the disbursement of foreign aid can involve a long decision-making process.
77 It is noteworthy that for the separate effects of FDI equity and FDI debt, equity is often positively correlated with regional growth (except for the East) and debt is almost always negatively correlated. Moreover, equity effect is only marginally statistically significant, whereas debt effect is not statistically significant at all.
89 ! )$ #)'#) $#! % )!# !$,($# ' #.( $#!$#$" '$,) - -%$+' *"'"
&* )- &* )- ) # ) !*+ !&+!) ',+$ *+ '%&+%#&%"% & !)' $% !*+ !&+!) ',+$ *+ '%&+%#&%"% & !)' $% !*+ !&+!) ',+$ *+ '%&+%#&%"% & !)' $% ""! + $ &#!+ & ) !))')* )!%&( )!&+$!*%*
Overall, the observations about the simultaneous effect of regional economic growth on capital attraction imply that private investors likely take past oblast development into account, though this decision is rather dynamic and complex. The decisions of public investors, in contrast, differ considerably based on the particular institution involved. A better understanding of this process for both private and public stakeholder would require a much longer panel time series.
90 ! )$ ' #.( $#!$#$" '$,) $##)'#) $#! % )!# !$,(- -%$+' *"'"
&* )- &* )- ) # ) !*+ !&+!) ',+$ *+ '%&+%#&%"% & !)' $% !*+ !&+!) ',+$ *+ '%&+%#&%"% & !)' $% !*+ !&+!) ',+$ *+ '%&+%#&%"% & !)' $% ""! + $ &#!+ & ) !))')* )!%&( )!&+$!*%*
Finally, I analyzed the effect of urbanization on growth and international capital flows by adding a measure of the urbanization rate to my model. I hypothesized that the variable might capture enough of the variation across regions to make region effects disappear. In other words,
91 the region effects might then be due to urban vs. rural economic differences, rather than pre- existing characteristics. Oblasts with large urban populations might be expected to grow more quickly and attract more foreign capital (e.g., Saha et al., 2014; Bobonis and Shatz, 2007). In practice, however, adding the urbanization rate did not considerably influence the model parameter estimates (see detailed results in Table C11 in Appendix C). Previously important factors responsible for influencing an oblast’s economic development—such as pre-existing economic development, external shocks, law enforcement and human capital—remain robust, no matter which type of capital is considered. Region effects also remain highly statistically significant. Similar to the main results, the effect of foreign capital flows on regional growth is statistically significant for all types of capital, except for FDI debt, though its magnitude is slightly reduced. The urbanization rate itself is either not statistically significant or negatively related to oblast growth (in the case of FDI and NED financing). In the second equation, the results remain almost identical to those from earlier model specification. As in my main model, regional growth does not affect M&A capital attraction and marginally influences NED financing. Finally, the urbanization rate is statistically significantly positively related to foreign capital attraction (except for EBRD and marginally for M&A), with the strongest correlation being in the NED case. One possible explanation of these results is that a curious trend that emerges when one looks at the regional average urbanization rates. In the Western region, slightly less than 50% of the population lives in urban areas. In both Central and Southern regions, about 60% live in cities. In contrast, in the Eastern region this number is more than 80%. Since the Central and Southern regions were consistently prosperous and the primary destination for foreign investors between 2001 and 2013, whereas the Eastern oblasts stagnated since early 2010, the effect of urbanization appears to be less important in explaining Ukraine’s regional economic differences than it is in other countries.
My model of regional economic growth and international capital flows, which accounts for the simultaneous relationship between the two phenomena, fits Ukraine’s regional data well and produces several important insights. The results show that the effect of international capital flows on regional economic growth indeed varies by both investment type and destination region. There is a simultaneous connection between foreign investment and regional economic growth, especially for FDI and public capital. Moreover, several factors such as pre-existing economic development, external shocks, law enforcement and human capital are rather robust and strongly important in regional economic development, no matter what type of capital is considered. Additionally, region effects—the
92 unobserved variation across regions due to their specific geographic, cultural, and historical characteristics—are also important. The nexus between foreign investment and economic growth has a distinct pattern in the Western oblasts, where any small increase in international capital flows, either public or private, would create the largest effect on local growth and development. At the same time, among the other three regions, the Eastern oblasts, on average, experience the smallest economic boost from a marginal increase in attracted private capital flows. In most of Ukraine’s oblasts, an increase in foreign private capital, and particularly in equity, has a larger positive effect on regional economic growth than that of public capital. But, given the low level of initial public investment, their increase has substantial and positive effect on regional economic growth. Supplemental examinations of effect heterogeneity across time and urbanization dimensions reveal that the relationship between growth and international capital flows is dynamic and complex. The positive influence of private capital on regional growth often occurs with some time lag, with the effect becoming evident after at least a year. In the case of public capital investments, EBRD financing is found to enhance regional growth in the short term, whereas NED investment requires a longer time to translate into enhanced development. Also, in the medium term, the simultaneous effect of regional development on capital attraction suggests that, while decisions of public investors differ considerably based on a particular institution, private investors consider a region’s past development trajectory when making dynamic investment decisions. Finally, adding a measure of urbanization does not considerably influence the model estimates. No matter what type of capital is considered, the model findings are robust.
93
Intra-national regional differences are frequently noticeable. Policy discussions about regional differences regularly refer to historical, sociological, and identity-related factors, and less often to economic differences. But it is important to understand why persistent regional socio-economic disparities exist within countries, what factors drive them, and what can be done to reconcile these differences. It is also essential to identify the role of capital flows in regional divergent paths, because prosperous localities amass thriving businesses. The goal of this dissertation was two-fold. First, I summarized existing research on the topic of regional disparities within countries, drew parallels between factors affecting regional economic differences in different countries, and highlighted those aspects specifically related to the movement of international capital. Second, I answered a broad research question about how international capital flows are connected to differential regional economic development, exploring why some regions attract more foreign investment than other regions, and how regional differences in FDI receipts influence regional economic development (and vice versa). The hypothesis of this research was that international capital flows affect regional economic growth differently, depending on their type and size. Additionally, I hypothesized that the relationship between growth and capital flows is bidirectional, i.e., that capital flows drive growth, while growth influences the extent and nature of capital flows. I tested my hypothesis empirically using a robust modeling approach and a comprehensive economic data set from Ukraine, compiled from a range of data sources.
In the literature, the most important factors influencing a country’s economic growth are those related to labor, capital, institutions, and policies. Labor-related factors include population growth, life expectancy, fertility, mobility, and urbanization. Capital-related factors are typically divided into three categories—private, public, and human—which are seen as having distinct effects. Private capital can be domestic capital stock and foreign investment (e.g., FDI or private foreign debt). Public capital is usually viewed as either development aid (e.g., provided by international organizations such as the World Bank and USAID), financial aid (e.g., provided by international financial organizations such as IMF), or public infrastructure (roads, bridges, utilities, transportation, cultural/sports), which can also be
94 financed through development or financial aid. Human capital is usually measured in terms of schooling, R&D, universities, managerial skills, and health. Institutional development, another crucial factor, is usually measured in terms of the strength of the rule of law; social capital, political rights, and democracy (sometime measured using data on voter participation, protests, and internet penetration), market development, economic activity, the nature of interactions between people and firms, and ease of doing business (e.g., the World Bank ranking). Monetary and fiscal policies thought to contribute to economic growth include targeted inflation, cost of capital/interest rate, exchange rate regimes, government subsidies (either debt- financed or tax financed), and taxes collected/government consumption. In addition, existing studies point to a host of other factors that play a role in economic growth. These factors include openness to trade, the extent of exports and imports, tourism, volatility or output instability, geography, natural resources, living standards, and the location of industry.
China, Italy, and the United States are three examples of countries with pronounced regional differences where similar combinations of factors have influenced the movement of international capital and development. Despite having very different histories, cultures, and socio-economic systems, all three countries have substantial intra-national variation. Their regional economic disparities are commonly attributed to several key factors, including local geography, human and social capital, institutional quality, infrastructure, public policies, and clustering of economic activities. However, the relative importance of these factors considerably varies across the countries. In China, local geography and specific government policies, such as creation of SEZs and industrial clusters, played an important role in divergent paths of its Eastern developed region and the rest of the country. The same is true for the inflow of foreign capital that was motivated by the preferential regulatory regimes of SEZs and industrial clusters and presence of the large sea ports in proximity. In the Italian case, the quality of social and human capital, local institutions, and the level of corruption are seen as instrumental in driving the substantial regional disparities between the Northern and Southern provinces, both in economic development and attractiveness to foreign investors. The United States has also highly unequal regional development. Its pre-existing regional disparities are reinforced by high labor mobility, clustering of economic activities and agglomeration: rich states along the coasts draw more human capital and get richer, and poor inland states become poorer. Similarly, foreign capital is also attracted by big markets that are
95 concentrated in the country’s Mideast and Great Lakes, as well as in California in the Far West and Texas in the Southwest.
Across all case studies, the most prosperous regions are also those with the largest international capital inflows. For example, three Chinese Eastern provinces alone – Shanghai, Jiangsu, and Guangdong—received half of all country’s FDI. In Italy, the wealthy Northern region and Lazio province in its Central part (where the capital city, Rome, is) are the most attractive to international capital. In the U.S., FDI is concentrated in just two states, California and Texas, and three regions—Great Lakes, Mideast, and New England. These U.S. localities also have the highest per capital income.
The socio-economic and investment data for Ukrainian regions between 2001 and 2013, revealed a systematic pattern in both regional socio-economic development and attractiveness to foreign capital. Specifically, across all regions, the Central region is consistently the most prosperous, and it also received the bulk of the private and public foreign capital flowing into the country. The Southern region has consistently come in second place with regard to economic development and foreign investment receipts, but lagged considerably behind the Center with regard to institutional development. Although the oblasts in the Eastern region are wealthy relative to other parts of the country, the region has seen economic decline since 2011. It also systematically fell behind in terms of foreign capital receipts, except for NED public financing. Finally, the Western region is the poorest across all the Ukrainian regions, but has high level of human capital and civic engagement. Aside from EBRD financing, the oblasts in the West do not attract much foreign capital.
In Chapter 6 of this monograph, I developed a model of regional economic growth and international capital flows, taking into account a simultaneous relationship between these phenomena. I then tested and estimated the model using Ukraine’s regional data. The results show that the effect of international capital flows on regional economic growth is positive, but the magnitude and statistical significance of the effect varies both by the type of investment and the destination region. The positive effect is also bidirectional: higher local economic growth attracts more FDI and public financing, and less foreign aid. Moreover, several
96 factors, such as pre-existing economic development, external shocks, law enforcement, and human capital are robust and have important effects on in regional economic development. Additionally, unobserved region-specific differences are important, but they cannot fully explain regional differences. Finally, adding a measure of urbanization does not considerably influence the model parameter estimates – no matter what type of capital is considered, the model results remain robust.
The nexus between foreign investment and economic growth has a distinct pattern in the Western oblasts, where any small increase in international capital flows, either public or private, would create the largest effect on local growth and development. At the same time, among the other three regions, the Eastern oblasts, on average, experience the smallest economic boost from a marginal increase in attracted private capital flows. In most of Ukraine’s oblasts, an increase in foreign private capital, and particularly in equity, has a larger positive effect on regional economic growth than that of public capital. Supplemental examinations of effect heterogeneity across time and urbanization dimensions reveal that the relationship between growth and international capital flows is dynamic and complex. The positive influence of private capital on regional growth often occurs with some time lag, with the effect becoming evident after at least a year. In the case of public capital investments, EBRD financing is found to enhance regional growth in the short term, whereas NED investment requires a longer time to translate into enhanced development.
Economic underdevelopment and the scarcity of foreign investment have been chronic problems in Ukraine since the fall of the Soviet Union. Without improvement in Ukraine’s ability to attract foreign capital, Ukraine will have difficulty raising its standard of living and competitiveness in the world market. My study highlights the significant regional differences in Ukraine’s economic development. Specifically, between 2001 and 2013, the Central region was consistently the most prosperous region whereas the Western region was the poorest region. The regional distribution of foreign capital receipts was also uneven and followed a similar pattern. In addition, the Eastern region has stagnated since 2011 and experienced the lowest return on its foreign investment in terms of growth.
97 ! Given the substantial existing regional differences in both development and foreign investment, the goal of Ukraine’s regional economic policy should be to achieve balanced and sustainable growth across all regions. Failure to fulfil this goal could lead to increasing popular discontent and potential social unrest. My research suggests several ways that Ukraine could address growing regional disparities and help less developed regions catch up.
!! # " ! ! The Western oblasts face consistent problems in their economic development, attract little investment and lack rich natural resources. But they have high human capital and civic engagement. Hence, people are the major contributors to this region’s growth. Service industries that rely on the labor force (rather than capital), such as IT and tourism, could become drivers for both foreign investment in this region and its growth.
! ! ! ! " ! # " # ! " The Eastern oblasts and Crimea chronically have fallen behind the Central and Southern oblasts in terms of international capital inflows. Furthermore, when foreign investment did go to Crimea, Donetsk and Luhansk oblasts, it came predominately from two sources: (1) Ukrainian domestic capital reinvested through offshore zones (such as Cyprus) and (2) Russian capital (either invested directly or via offshore areas). In these two Eastern oblasts and Crimea, Russian businesses play a substantial role, whereas in other regions, Russian capital plays a far less significant role. For example, between 2001 and 2013 in the Kharkiv and Odesa oblasts, private investment came not only from Russia, but also from a mixture of other countries, including the European Union, Canada, the United States, Turkey and China. Consequently, the local population and firms in Crimea, Donetsk and Luhansk oblasts are particularly vulnerable to Russia’s influence. Second, the Eastern oblasts have lower human development than Ukraine’s national average. Consistent with the Ukrainian government’s official State Strategy for Regional Development until 2020, priority should be given to improving human capital and the competitiveness of the labor force in the Eastern region. Doing so would boost the positive effect of foreign capital on local growth.
! Foreign investment not only enhances regional growth, but it itself depends on prior regional development. As regions become more developed, they generally receive more foreign
98 investment. Regional development, in turn, depends on institutions and public policies, including fiscal policy. Taking into account the fact that (a) more than a third of FDI in Ukraine’s regions come from offshore zones and (b) that this investment positively correlates with local crime incidence (which includes tax evasion crimes), improvement in the tax system could boost foreign investment. Reducing the overall tax burden and streamlining related regulations could not only help generate additional tax revenues by retaining some of Ukraine’s domestic capital at home (rather than causing it to be reinvested via offshore areas), but also attract more foreign investment. This would in turn enhance regional economic growth.
! Executives and managers employed mainly in major international companies consistently emphasize that fighting corruption is the number one priority for improving the business climate in Ukraine.78 In these surveys, anti-corruption measures are also closely linked to overall judiciary reform and improvement in law enforcement. Because the judicial system is part of the rule of law and institutional development in general (factors that influence both economic growth and foreign capital attraction), prioritizing these reforms is crucial for Ukraine’s business climate and future economic growth.
Foreign investment not only enhances regional economic growth, but it also depends on prior regional development. As regions become more developed, they receive more foreign private (FDI) and more international public capital, but less foreign aid. In Ukraine, this nexus between foreign investment and economic growth holds across all regions, but is particularly distinct for the poorest provinces. In these regions, a small increase in international capital flows would create the largest effect on local growth. At the same time, the likelihood of receiving these capital flows is directly related to economic growth. Thus, policy measures targeted at attracting more foreign investment should be implemented together with an economic strategy of enhancing local economic growth.
78 For example, see American Chamber of Commerce in Ukraine Corruption Perception Survey 2015; European Business Association in Ukraine, Dragon Capital, and Center for Economic Strategy Survey of Foreign Investors 2017.
99 When a poor region consistently lacks capital, foreign aid has a potential to stimulate local economic growth. However, as seen in Ukraine’s case, because foreign aid usually focuses on strengthening civil society and community organizations, sometimes more foreign aid goes to areas of donor’s strategic political interests (e.g., the Eastern oblasts along the Russian border) rather than the least developed regions (e.g., the Western oblasts along the EU border). Even if foreign aid potentially changes local political and social preferences, it should not do it at a cost of deeper regional economic disparities.
Ignoring persistent regional disparities may create vulnerabilities that could potentially be exploited. Recent examples of cases where regional differences are particularly emphasized (besides Ukraine) include current political divisions within the United States and the Great Britain’s domestic divergent views about Brexit. In the Great Britain, the dividing lines go starkly between Northern Ireland and Scotland, on one hand, and Wales and England, on the other.79 In the United States, cultural and political differences between the economically distinct coastal and inland states (in Great Lakes, Plains, Mountains, Southeast, and Southwest regions) have become increasingly stark.80 In many other countries, existing regional differences may be unnoticeable today but could become evident in future. While addressing regional disparities, policy-makers should decide on the approaches that would work best depending on local conditions and preferences. These approaches could include either redistribution or making more efficient use of existing resources, or a combination of both. Policymakers could target key factors known to influence growth, and seek to balance them out across regions. While geographic characteristics can’t be equalized, human and social capital could be. Cross-regional gaps in the quality of local institutions and available infrastructure could also be addressed through the prioritization of investments in areas where they are weakest. An alternative policy approach may be to let regional differences persist, but to pay closer attention to related issues, such as labor outflow from poor to wealthy regions (and abroad), income inequality, and potentially divergent political preferences.
79 “EU referendum: The result in Maps and Charts”, BBC News, June 24, 2016. 80 2016 Presidential Election Results. For more details see https://www.270towin.com/maps/2016-election-state- winners
100 Even though regional disparities are common across countries, studies about specific aspects of these regional differences are complicated by the lack of good regional data. Nonetheless, future research should aim at investigating further the contribution and mechanism of influence of each regional factor as mentioned in this monograph. Moreover, as the share of portfolio investment tends to increase and the overall foreign liabilities of many countries expand, gathering and analyzing related data is essential. Such further explorations could shed light on the complex relationship and feedback mechanisms between foreign investment and regional economic growth. Knowing the range of each factor’s effect would also help policymakers to develop targeted policy instruments. Additionally, investigating the effect of regional policies on local growth, especially in countries where regions have a substantial fiscal and/or policy autonomy, is essential in order to understand what policies could help reduce economic gap between regions. Finally, exploring cross-disciplinary issues related to regional socio-economic differences, such as those mentioned in the previous section of this chapter and that are beyond the scope of this study, is important when developing relevant policy alternatives.
101 "
# !!! #
102
103
104 99.7-2@"+5*;<(*:2*<287;27%8,28,87862, 7-2,*<8:;
! '" "#!# " # !" %! &
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
105 $! !!
"#+&"*05*1&.,/"0*1.-1.'"0(*-"+*-&"00&%*$2*.-1'.0&(*.-"+ "0*"#+&1*-.%&+ *- "#+& +"12*$*2*&1.'&(*.-"+-/32"$2.01
3,,6 8 8 ! ! # ! "!# ! !# ! "!# ! !# ! !# "!
"#+& ""*05*1&.,/"0*1.-1.'"0(*-"+*-&"00&%*$2*.-1'.0&(*.-"+ "0*"#+&1*-.%&+ *- "#+& &(*.-"+$.-.,*$0.52)"$2.01
.2"+"/*2"++.5&(*.- .-20"12 2%00 7 878 $.-'-2&04"+!
! # ! "!# ! !# ! "!# ! !# ! !# "!
"#+& #"*05*1&.,/"0*1.-1.'"0(*-"+*-&"00&%*$2*.-1'.0&(*.-"+ "0*"#+&1*-.%&+ *- "#+& &(*.-"+$.-.,*$0.52)"$2.01
.0&*(-"/*2"++.5&(*.- .-20"12 2%00 7 878 $.-'-2&04"+!
! # ! "!# !
!# ! "!# ! !# ! !# "!
106 ! ! "" #
!$ #& # !$ #& # !$# #" " %#
,
0 0 0 (*
,- !',+
*$&!
(&!+,$ )$, %
(*!$"' )$, %
(*!$"' )$, %!',!*
(*!$"' )$, %(-,#
(*!$"' )$, % +,
!"$('-&&/
*$+$+-&&/
0 ('+, ',
*( #$ !$# #" " #
(*
,- !',+
*$&!
0 *(.,#
107 < 4296-*16*4
4296-"276-
< 4296-&56
.1&1(.&/ 156.676.215
!*,.21700; %" %" %" %" %" %"
< < 4.5.5700;
2156&16
! " 42'(-.
#//8&4.&'/*5&4*+.456).++*4*1(*5*:(*36+24 6 &1).1/2,&4.6-05*:(*36+24+.1&1(.&/.156.676.215 $*56.5&4*+*4*1(*4*,.21"6&1)&4)*442452+6-*(2*++.(.*165&4*.13&4*16-*5.5 3 3 3 † 3
108 $ #*/#+ '%(*#+'&+' *!#&$#& ** # ,#'&+ '* !#'&$*# $ +3 )-#,0' $
)-#,0 !#'& '&,*+, , ** 1 212 '& &, *.$
*'/," !#'& '&,*+, , ** 1 212 '& &, *.$
$ #*/#+ '%(*#+'&+' *!#&$#& ** # ,#'&+ '* !#'&$*# $ +3 ,' $
, !#'& '&,*+, , ** 1 212 '& &, *.$
*'/," !#'& '&,*+, , ** 1 212 '& &, *.$
109 &! %,1%-!)'* ,%-)(-)" ,#%( &%(! ,,! % .%)(-"),!#%)( & ,% &!-5 +/%.2! .) !&
+/%.2! .!#%)( )(., -. . ,, 3 434 )("(.!,0 &
,)1.$!#%)( )(., -. . ,, 3 434 )("(.!,0 &
&! %,1%-!)'* ,%-)(-)" ,#%( &%(! ,,! % .%)(-"),!#%)( & ,% &!-5 *%. &) !&
*%. &!#%)( )(., -. . ,, 3 434 )("(.!,0 &
,)1.$!#%)( )(., -. . ,, 3 434 )("(.!,0 &
110 # "(,") &$'(")&%)& ( "%#"% (( " *"&%) &( "&%#("# )/ '"*#& #
'"*# "&% &%*()* * (( - .-. &% %* (+#
(&,*! "&% &%*()* * (( - .-. &% %* (+#
# "(,") &$'(")&%)& ( "%#"% (( " *"&%) &( "&%#("# )/ '"*#& #
'"*# "&% &%*()* * (( - .-. &% %* (+#
(&,*! "&% &%*()* * (( - .-. &% %* (+#
111
"! ! #! / *- /0"#)/. -&(# *(#./&! +&/ ' ,0&/2 ,0&/2#)/#- ,0&/2 *0/% ,0&/2 ./ #$&*)0((2 -&.&.0((2 *)./ )/ -* !%& "! ! "!$ *- /0"#)/. -&(# -*1/%
3 -*1/%#)/#-
-*1/% *0/%
112 "96>;1*:; 25*5,2*3 5:;2;<;265: 2+9*92.: '6;.9%<956<; #.0265<44@ )$ )$ )$ )$ )$ 92:2:<44@ 65:;*5; # $ "96+,12
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† 7
113
"! ! #!
.
1 *,
./"#).-
,&(#
*(#-.&! +&. '
# .
# .#).#,
# . */.%
# . -.
#$&*)/((0
,&-&-/((0
1 1 *)-. ).
,* !%& "! ! !
*,
./"#).-
114 924.
B B B "96>;1
B B "96>;1.5;.9
"96>;1$6<;1
"96>;1*:;
25*5,2*3 5:;2;<;265:
2+9*92.:
'6;.9%<956<;
#.0265<44@ )$ )$ )$ )$ )$
B B 92:2:<44@
65:;*5;
# $ "96+,12
!%33,65;25<6<:=*92*+3.:*9./29:;-2//.9.5,.:.?,.7;/69 "; *5-=6;.9;<956<;*5-25360*92;14: .?,.7;/69/25*5,2*325:;2;<;265:32+9*92.:*5-=6;.9;<956<;(.:;2:*9./.9.5,.9.0265 92:2:-<44@2:642;;.-25;1.9.09.::265>2;1=6;.9;<956<;+.,*<:.;1.9.>.9.56.3.,;265:25 $ 2:*0.5.9*32A*;2656/$&#*5- $ $*:2;*336>:/69+6;1.5-60.56<:9.09.::69:*5-,699.3*;265:6/;1..9969 ;.94:25-2//.9.5;.8<*;265:$.;16-*::<4.:;1.9.*9.56.5-60.56<:9.09.::69: $ $-6.:56;*,,6<5; /69,699.3*;265:6/;1..9969;.94:25-2//.9.5;.8<*;265: $;*5-*9-.9969:6/;1.,6.//2,2.5;:*9.257*9.5;1.:2: 7 7 7 † 7
115
"! ! #!
/
*-
/0"#)/.
-&(#
2 *(#./&! +&/ '
,0&/1 )"# /
,0&/1 )"# /#)/#-
2 ,0&/1 )"# / *0/%
2 ,0&/1 )"# / ./
#$&*)0((1
-&.&.0((1
*)./ )/
-* !%& "! ! "!$ !
*-
/0"#)/.
116 813-
A A !85=:0
!85=:0-4:-8
!85=:0#5;:0
!85=:0)9:
14)4+1)2 49:1:;:1549
1*8)81-9
&5:-8$;845;:
"-/154;33? (# (# (# (# (#
A A A 81919;33?
549:)4:
" # !85*+01
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† 6
117
#" "! ! $"
2 /
2 !+-
/0#$*/.
-')$
2 +)$./'" ,'/ (
,'/ (
,'/ ($*/$-
,'/ ( +0/&
,'/ ( ./
$%'+*0))1
-'.'.0))1
2 +*./ */
-+!"&' #" "! "
2 !+-
/0#$*/.
118 :35/
C # :7?<2
# :7?<2/6
# :7?<2%7=<2
C # :7?<2+;<
36+6-3+4 6;<3<=<376;
3,:+:3/;
(7
$/1376=55A *% *% *% *% *%
:3;3;=55A
76;<+6<
$ % #:7,-23 !
!"&44-76<36=7=;>+:3+,4/;+:/03:;<.300/:/6-/;/@-/8<07: #< +6.>777
119
"! ! #!
.
1 *,
./"#).-
,&(#
1 *(#-.&! +&. '
+&. '
1 +&. '#).#,
+&. ' */.%
1 +&. ' -.
#$&*)/((0
,&-&-/((0
*)-. ).
,* !%& "! ! !
*,
./"#).-
120 924.
B "96>;1
"96>;1.5;.9
B "96>;1$6<;1
B "96>;1*:;
B 25*5,2*3 5:;2;<;265:
2+9*92.:
'6;.9%<956<;
#.0265<44@ )$ )$ )$ )$ )$
92:2:<44@
65:;*5;
# $ "96+,12
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† 7
121
"! ! #!
.
1 *,
./"#).-
,&(#
*(#-.&! +&. '
+&. '
+&. '#).#,
+&. ' */.%
+&. ' -.
#$&*)/((0
,&-&-/((0
1 1 *)-. ).
,* !%& "! ! !
*,
./"#).-
122 B 924.
B "96>;1
B B B "96>;1.5;.9
"96>;1$6<;1
"96>;1*:;
25*5,2*3 5:;2;<;265:
B 2+9*92.:
B '6;.9%<956<;
#.0265<44@ )$ )$ )$ )$ )$
92:2:<44@
65:;*5;
# $ "96+,12
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† 7
123 ! ! "" #
! # ! # ! "
*
- &(
*+ %*)
- ("$
- &$ )*" '"*#
&( " %'"*# *
&( " %'"*# %* ( *
&( " %'"*# &+*! *
&( " %'"*# )* *
&( " %'"*# *
&( " %'"*# %* ( *
&( " %'"*# &+*! *
&( " %'"*# )* *
&( " %'"*# *
&( " %'"*# %* ( *
&( " %'"*# &+*! *
- &( " %'"*# )* *
"&%+$$,
124 (#)#)+%%-
. . '&)* &*
(' "#
'(
*+ !&*)
(#%!
. (',*" *
(',*" !&*!( *
(',*" '+*" *
(',*" )* *
(',*" *
(',*" !&*!( *
(',*" '+*" *
(',*" )* *
. (',*" *
(',*" !&*!( *
(',*" '+*" *
(',*" )* *
. #& & # $&)*#*+*#'&)
125 !*,.21700: %" %" %" %" %" %"
4.5.5700:
2156&16
! " 42'(-.
#//8&4.&'/*5&4*+.456).++*4*1(*5*9(*36+24 6 &1).1/2,&4.6-05*9(*36+24+.1&1(.&/.156.676.215 $*56.5&4*+*4*1(*4*,.21 1 (&3.6&/(&5*)20*56.((&3.6&/.1(/7)*5)20*56.( //53*(.+.(&6.215&4**56.0&6*)75.1," "6&1)&4)*442452+6-*(2*++.(.*165&4*.13&4*16-*5.5 3 3 3 ;3
126 (# '.3'/#+), .'/+*/+$ .%'* ('*# ..#"'!0'+*/$+.#%'+* ( .' (#/7 -1'04+"#(
-1'04#%'+* +*0. /0 0".. 5 656 !+*$*0#.2 (
.+30&#%'+* +*0. /0 0".. 656 !+*$*0#.2 (
127
%! $*.$+!(&) *$+('+(" *#$' %$'! **! $ ,$('+"(*!#$(' % *$ %!+1 ! ,( !%
! ,!#$(' (',* +, , ** / 0/0 ('"',!*- %
128 /,41'$&(,+ ,+1/ 01 1# // 767 ",+% +1$/3 )
!)$ " (/4(0$,*- /(0,+0,% /&(+ )(+$ //$#("1(,+0%,/$&(,+ ) /( !)$08 .2(15$!1,#$)
.2(15$!1$&(,+ ,+1/ 01 1# // 6 767 ",+% +1$/3 )
129
$#(& #" #"&$%& &$$ *)* #" "& $'!
130 '" ! &,0&-"*(+ ,&-*)-*# ,$&) '&)" ,,"!& .&*)-#*,"$&*) ' ,& '"-3 +&. '*!"'
+&. '"$&*) *)., -. .!,, 1 212 *)#).",/ '
,*0.%"$&*) *)., -. .!,, 212 *)#).",/ '
131
# "(,") &$'(")&%)& (!"%#"% (( " *"&%) &( !"&%#(" # )/ '"*#& #
'"*# !"&% &%*()* * (( - .-. &% %* (+#
132 +)/-$!#%)( )(-+ ,- - ++ 101 )("(-!+. &
&! " %+/%,! )'* +%,)(,)" +#%( &%(! ++! % -%)(,")+!#%)( & +% &!,2 *%- &) !&
*%- &!#%)( )(-+ ,- - ++ 0 101 )("(-!+. &
133
$#(& #" #"&$%& &$$ *)* #" "& $'!
134 " !
!$ #& !$ #& # # !$# #" " %#
-
1 )+
1 -.!"(-,
1 +%'"
)'",-% *%- &
)+"%#( *%- &
)+"%#( *%- &"(-"+
)+"%#( *%- &).-$
)+"%#( *%- & ,-
+ (%0 -%)(
"#%)(.''/
+%,%,.''/
1 )(,- (-
+) $% !$# #" " #
1 )+
-.!"(-,
+%'"
135 ? ! 64;8/
? ! 64;8/,38,6
! 64;8/#498/
? ! 64;8/(78
03(3*0(1 37808980437
? %6)(30>(8043
",.043922= '# '# '# '# '# '#
? ? 60707922=
4378(38
" # !64)*/0
$11:(60()1,7(6,-0678+0--,6,3*,7,<*,58-46 !8 (3+0314.(608/27,<*,58-46-03(3*0(1037808980437(3+ 96)(30>(80436(8, 3 *(508(1*(7,+42,780**(508(103*19+,7+42,780* &,7807(6,-,6,3*,6,.043 #8(3+(6+,664674-8/,*4,--0*0,387(6,035(6,38/,707 5 5 5 † 5
136
Abdioglu, N., Khurshed, A., & Stathopoulos, K. (2013). Foreign institutional investment: Is governance quality at home important?. Journal of International Money and Finance, 32, 916-940. Aiello, F., & Scoppa, V. (2000). Uneven regional development in Italy: explaining differences in productivity levels. Giornale degli Economisti e Annali di Economia, 270-298. Aiello, F., & Scoppa, V. (2008). Convergence and regional productivity divide in Italy: evidence from panel data. Akai, N., & Sakata, M. (2002). Fiscal decentralization contributes to economic growth: evidence from state-level cross-section data for the United States. Journal of Urban Economics, 52(1), 93-108. Akulava, M., & Vakhitova, G. (2010). The impact of FDI on firm’s performance across sectors: evidence from Ukraine. Kyiv School of Economics, (26). Ali, H., & Ahmad, S. (2010). Foreign direct investment, economic growth and regional disparities: the Malaysian experience. International Journal of Interdisciplinary Social Sciences, 5(8), 405-416. Ascari, G., & Di Cosmo, V. (2005). Determinants of total factor productivity in the Italian regions. Scienze Regionali. Aschauer, D. A. (1989a). Is public expenditure productive?. Journal of Monetary Economics, 23(2), 177-200. Aschauer, D. A. (1989b). Public investment and productivity growth in the Group of Seven. Economic Perspectives, 13(5), 17-25. Aschauer, D. A. (1995). Infrastructure and macroeconomic performance: Direct and indirect effects. The OECDJob Study1995: Investment, Productivity and Employment, 85-101. Atesoglu, H. S. (2004). Defense spending and investment in the United States. Journal of Post Keynesian Economics, 27(1), 163-170. Balasubramanyam, V. N., Salisu, M., & Sapsford, D. (1996). Foreign direct investment and growth in EP and IS countries. The Economic Journal, 92-105. Ballarino, G., & Schadee, H. (2005). Civicness and economic performance. A longitudinal analysis of Italian provinces, 1980–2000. European Sociological Review, 21(3), 243-257.
137 Bao, S., Chang, G. H., Sachs, J. D., & Woo, W. T. (2002). Geographic factors and China's regional development under market reforms, 1978–1998. China Economic Review, 13(1), 89- 111. Barro, R. J. (1990). Government spending in a simple model of endogenous growth. Journal of Political Economy, 98(5, Part 2), S103-S125. Barro, R. J. (1991). Economic growth in a cross section of countries. The Quarterly Journal of Economics, 106(2), 407-443. Barro, R. J. (1996a). Determinants of economic growth: a cross-country empirical study (No. w5698). National Bureau of Economic Research. Barro, R.J. (1996b). Democracy and growth. Journal of Economic Growth, 1(1), 1-27. Barro, R. J., Mankiw, N. G., & Sala-i-Martin, X. (1995). Capital mobility in neoclassical models of growth. American Economic Review, 85(1), 103–115. Barro, R. J., Sala-i-Martin, X., Blanchard, O. J., & Hall, R. E. (1991). Convergence across states and regions. Brookings papers on economic activity, 107-182. Bartik, T. J. (1985). Business location decisions in the United States: Estimates of the effects of unionization, taxes, and other characteristics of states. Journal of Business & Economic Statistics, 3(1), 14-22. Basile, R., Benfratello, L., & Castellani, D. (2005). Attracting foreign direct investments in Europe: are Italian regions doomed?. Centro Studi Luca d'Agliano Development Studies Working Paper No. 200 Basile, R., Castellani, D., & Zanfei, A. (2003). Location choices of multinational firms in Europe: the role of national boundaries and EU policy. Quaderni di Economia, Matematica e Statistica, Urbino Basu, P., Chakraborty, C., & Reagle, D. (2003). Liberalization, FDI, and growth in developing countries: A panel cointegration approach. Economic Inquiry, 41(3), 510-516. Bengoa, M., & Sanchez-Robles, B. (2003). Foreign direct investment, economic freedom and growth: new evidence from Latin America. European Journal of Political Economy, 19(3), 529-545. Benoit, E. (1978). Growth and defense in developing countries. Economic Development and Cultural Change, 26(2), 271-280. Bobonis, G. J., & Shatz, H. J. (2007). Agglomeration, adjustment, and state policies in the location of foreign direct investment in the United States. The Review of Economics and Statistics, 89(1), 30-43.
138 Bologna, J., Young, A. T., & Lacombe, D. J. (2016). A spatial analysis of incomes and institutional quality: evidence from US metropolitan areas. Journal of Institutional Economics, 12(01), 191-216. Boone, P. (1996). Politics and the effectiveness of foreign aid. European Economic Review, 40(2), 289-329. Borensztein, E., De Gregorio, J., & Lee, J. W. (1998). How does foreign direct investment affect economic growth?. Journal of international Economics, 45(1), 115-135. Boschma, R., & Iammarino, S. (2009). Related variety, trade linkages, and regional growth in Italy. Economic Geography, 85(3), 289-311. Bosworth, B., Collins, S. M., Chodorow-Reich, G., & Tille, C. (2007, January). Returns on Foreign Direct Investment: Does the United States Really Do Better? [with Comment and Discussion]. In Brookings Trade Forum (pp. 177-210). Brookings Institution Press. Braun, J. (1993). Essays on economic growth and migration, Ph.D. dissertation, Harvard University. Brehm, S. (2013). Fiscal incentives, public spending, and productivity–county-level evidence from a Chinese province. World Development, 46, 92-103. Bresnahan, T., & Gambardella, A. (Eds.). (2004). Building high-tech clusters: Silicon Valley and beyond. Cambridge University Press. Bronzini, R. (2004). Foreign direct investment and agglomeration: evidence from Italy. Banca d’Italia Tema di Discussione 526. Buss, T. F. (2001). The effect of state tax incentives on economic growth and firm location decisions: An overview of the literature. Economic Development Quarterly, 15(1), 90-105. Cabinet of Ministers of Ukraine. (2014). State Strategy for Regional Development until 2020. Cai, F., Wang, D., & Du, Y. (2002). Regional disparity and economic growth in China: The impact of labor market distortions. China Economic Review, 13(2), 197-212. Campbell, J. Y., & Mankiw, N. G. (1989). Consumption, income, and interest rates: Reinterpreting the time series evidence. NBER Macroeconomics Annual, 4, 185-216. Cass, D. (1965). Optimum growth in an aggregative model of capital accumulation. The Review of Economic Studies, 32(3), 233-240. Castells, A., & Solé-Ollé, A. (2005). The regional allocation of infrastructure investment: The role of equity, efficiency and political factors. European Economic Review, 49(5), 1165- 1205.
139 Chen, A., & Groenewold, N. (2013). Does investment allocation affect the inter-regional output gap in China? A time-series investigation. China Economic Review, 26, 197-206. Chen, B., & Feng, Y. (2000). Determinants of economic growth in China: Private enterprise, education, and openness. China Economic Review, 11(1), 1-15. Cheong, T. S., & Wu, Y. (2014). The impacts of structural transformation and industrial upgrading on regional inequality in China. China Economic Review, 31, 339-350. Choong, C. K., Baharumshah, A. Z., Yusop, Z., & Habibullah, M. S. (2010). Private capital flows, stock market and economic growth in developed and developing countries: A comparative analysis. Japan and the World Economy, 22(2), 107-117. Coccorese, P. (2009). Market power in local banking monopolies. Journal of Banking & Finance, 33(7), 1196-1210. Coe, D. T., & Helpman, E. (1995). International R&D spillovers. European Economic Review, 39(5), 859-887. Coppola, G., De Blasio, G., & Gallo, M. (1998). Development of Italian regions: the role of human capital. Rivista Internazionale di Scienze Economiche e Commerciali, 45, 543-562. Costa-i-Font, J., & Rodriguez-Oreggia, E. (2005). Is the impact of public investment neutral across the regional income distribution? Evidence from Mexico. Economic Geography, 81(3), 305-322. Daniele, V. (2015). Two Italies? Genes, intelligence and the Italian north–south economic divide. Intelligence, 49, 44-56. De Benedictis, L. (2007). 8. Italian comparative advantages, persistence and change in overall specialization. Industrial agglomeration and new technologies, 205. De Mooij, R. A., & Ederveen, S. (2003). Taxation and foreign direct investment: a synthesis of empirical research. International Tax and Public Finance, 10(6), 673-693. Deger, S., & Sen, S. (1983). Military expenditure, spin-off and economic development. Journal of Development Economics, 13(1), 67-83. Deger, S., & Smith, R. (1983). Military expenditure and growth in less developed countries. Journal of Conflict Resolution, 27(2), 335-353. Demurger, S. (2001). Infrastructure development and economic growth: an explanation for regional disparities in China?. Journal of Comparative Economics, 29(1), 95-117. Démurger, S., D Sachs, J., Woo, W. T., Shuming, B. A. O., & Chang, G. (2002). The relative contributions of location and preferential policies in China's regional development: being in the right place and having the right incentives. China Economic Review, 13(4), 444-465.
140 Destefanis, S., Barra, C., & Lubrano Lavadera, G. (2014). Financial development and local growth: evidence from highly disaggregated Italian data. Applied Financial Economics, 24(24), 1605-1615. Detotto, C., & Otranto, E. (2010). Does crime affect economic growth?. Kyklos, 63(3), 330-345. Di Liberto, A., Pigliaru, F., & Mura, R. (2008). How to measure the unobservable: a panel technique for the analysis of TFP convergence. Oxford Economic Papers, 60(2), 343-368. Diamond, P. A. (1965). National debt in a neoclassical growth model. The American Economic Review, 55(5), 1126-1150. Dumais, G., Ellison, G., & Glaeser, E. L. (2002). Geographic concentration as a dynamic process. Review of Economics and Statistics, 84(2), 193-204. Easterly, W., & Levine, R. (2001). What have we learned from a decade of empirical research on growth? It's Not Factor Accumulation: Stylized Facts and Growth Models. World Bank Economic Review, 15(2), 177-219. Ellison, G., & Glaeser, E. L. (1997). Geographic concentration in US manufacturing industries: a dartboard approach. Journal of Political Economy, 105(5), 889-927. Faini, R., Annez, P., & Taylor, L. (1984). Defense spending, economic structure, and growth: Evidence among countries and over time. Economic Development and Cultural Change, 32(3), 487-498. Fan, S., & Zhang, X. (2002). Infrastructure and regional economic development in rural China. China Economic Review, 15(2), 203–214. Fiorino, N., & Ricciuti, R. (2007). Legislature size and government spending in Italian regions: forecasting the effects of a reform. Public Choice, 131(1-2), 117-125. Fiorino, N., Galli, E., & Petrarca, I. (2012). Corruption and growth: Evidence from the Italian regions. European Journal of Government and Economics, 1(2), 126-144. Fischer, S. (1991). Growth, Macroeconomics, and Development. NBER Macroeconomics Annual, 6, 329-379. Forbes, K. J. (2010). Why do foreigners invest in the United States?. Journal of International Economics, 80(1), 3-21. Frattini, F., & Prodi, G. (2013). Industrial clusters in China: Policy tools for further and more balanced development. European Review of Industrial Economics and Policy, 5. Friedman, J., Gerlowski, D. A., & Silberman, J. (1996). Foreign direct investment: The factors affecting the location of foreign branch plants in the United States. Global Finance Journal, 7(2), 209-222.
141 Fujita, M., & Mori, T. (1996). The role of ports in the making of major cities: self-agglomeration and hub-effect. Journal of Development Economics, 49(1), 93-120. Fujita, M., Krugman, P. R., & Venables, A. J. (1999). The Spatial Economy: Cities, Regions, and International Trade. MIT Press. Gallup, J. L. & Sachs, J. D. (1998). Geography and economic development. Working paper, Harvard Institute for International Development, Harvard University, Cambridge, MA 02138. Ganong, P., & Shoag, D. (2012). Why has regional income convergence in the U.S. stopped?. Cambridge: Harvard University Mimeo. Garofalo, G. A., & Yamarik, S. (2002). Regional convergence: Evidence from a new state-by- state capital stock series. Review of Economics and Statistics, 84(2), 316-323. Gennaioli, N., La Porta, R., De Silanes, F. L., & Shleifer, A. (2014). Growth in regions. Journal of Economic Growth, 19(3), 259-309. Gerking, S., & Morgan, W. (1998). State fiscal structure and economic development policy. Growth and Change, 29(2), 131-144. Glaeser, E. L. (2001). The Economics of Location-Based Tax Incentives (No. 1932). Harvard- Institute of Economic Research. Glaeser, E. L., & Gottlieb, J. D. (2008). The economics of place-making policies. Brookings Papers on Economic Activity, 155-239. Glaeser, E. L., & Redlick, C. (2009). Social capital and urban growth. International Regional Science Review, 32(3), 264-299. Goh, C. C., Xubei, L., & Nong, Z. H. U. (2009). Income growth, inequality and poverty reduction: a case study of eight provinces in China. China Economic Review, 20(3), 485-496. Goldstein, G. S., & Gronberg, T. J. (1984). Economies of scope and economies of agglomeration. Journal of Urban Economics, 16(1), 91-104. Grossman, G. M., & Helpman, E. (1993). Innovation and growth in the global economy. MIT press. Guerrieri, P., & Iammarino, S. (2007). Dynamics of export specialization in the regions of the Italian Mezzogiorno: persistence and change. Regional Studies, 41(7), 933-948. Guiso, L., Sapienza, P., & Zingales, L. (2004). Does local financial development matter?. The Quarterly Journal of Economics, 119(3), 929-969. Gujarati, D. N., & Porter, D. C. (2009). Basic Econometrics. McGraw-Hill/Irwin.
142 Hansen, H., & Tarp, F. (2000). Aid effectiveness disputed. Journal of International Development, 12(3), 375-398. Hansen, T. J., & Kalambokidis, L. (2010). How are businesses responding to Minnesota’s tax- free zone program?. Economic Development Quarterly, 24(2), 180-192. Hanushek, E., & Kimko, D. (2000). Schooling, Labor-Force Quality, and the Growth of Nations. The American Economic Review, 90(5), 1184-1208. Head, C. K., Ries, J. C., & Swenson, D. L. (1999). Attracting foreign manufacturing: Investment promotion and agglomeration. Regional Science and Urban Economics, 29(2), 197-218. Head, K., Ries, J., & Swenson, D. (1995). Agglomeration benefits and location choice: Evidence from Japanese manufacturing investments in the United States. Journal of International Economics, 38(3), 223-247. Helpman, E. (1998). Institutions and Economic Performance. Harvard University Press. Helsley, R. W., & Strange, W. C. (1990). Matching and agglomeration economies in a system of cities. Regional Science and Urban Economics, 20(2), 189-212. Henderson, D. J., Tochkov, K., & Badunenko, O. (2007). A drive up the capital coast? Contributions to post-reform growth across Chinese provinces. Journal of Macroeconomics, 29(3), 569-594. Henderson, J. V. (1988). Urban Development: Theory, Fact and Illusion. Oxford Univ. Press. Henning, M., Enflo, K., & Andersson, F. G. (2011). Trends and cycles in regional economic growth. How spatial differences shaped the Swedish growth experience from 1860-2009. Explorations in Economic History, 48(4), 538-555. Hermes, N., & Lensink, R. (2003). Foreign direct investment, financial development and economic growth. The Journal of Development Studies, 40(1), 142-163. Higgins, M. J., Levy, D., & Young, A. T. (2006). Growth and convergence across the United States: evidence from county-level data. The Review of Economics and Statistics, 88(4), 671- 681. Higgins, M. J., Young, A. T., & Levy, D. (2009). Federal, state, and local governments: evaluating their separate roles in US growth. Public Choice, 139(3-4), 493-507. Jarreau, J., & Poncet, S. (2009). Export sophistication and economic performance: evidence from Chinese provinces. CEPII Working Document, (2009-34). Jayaratne, J., & Strahan, P. E. (1996). The finance-growth nexus: Evidence from bank branch deregulation. The Quarterly Journal of Economics, 111(3), 639-670.
143 Jiang, Y. (2011). Understanding openness and productivity growth in China: An empirical study of the Chinese provinces. China Economic Review, 22(3), 290-298. Jin, J., & Zou, H. F. (2005). Fiscal decentralization, revenue and expenditure assignments, and growth in China. Journal of Asian Economics, 16(6), 1047-1064. Jolley, G. J., Lancaster, M. F., & Gao, J. (2015). Tax incentives and business climate: executive perceptions from incented and nonincented firms. Economic Development Quarterly, 29(2), 180-186. Kenney, M. (2000). Understanding Silicon Valley: the anatomy of an entrepreneurial region. Stanford University Press. King, R. G., & Rebelo, S. (1990). Public policy and economic growth: developing neoclassical implications. Journal of Political Economy, 98(5, Part 2), S126-S150. Koopmans, T. C. (1965). On the concept of optimal economic growth. Kremer, M. (1993). Population growth and technological change: One million BC to 1990. The Quarterly Journal of Economics, 108(3), 681-716. Krugman, P. (1991). Increasing returns and economic geography. Journal of Political Economy, 99(3), 483-499. Kuzio, T. (1998). Ukraine: Coming to terms with the soviet legacy. Journal of Communist Studies and Transition Politics, 14(4), 1-27. La Costituzione della Repubblica Italiana (1947). Senato della Repubblica. Lambrinidis, M., Psycharis, Y., & Rovolis, A. (2005). Regional allocation of public infrastructure investment: The case of Greece. Regional Studies, 39(9), 1231–1244. Larrabee, F. S., Johnson, S. E., Gordon, J. I., Wilson, P. A., & Baxter, C. (2012). NATO and the Challenges of Austerity. RAND Corporation. Latzko, D. A. (2013). The geographic concentration of economic activity across the eastern United States, 1820–2010. Journal of Historical Geography, 41, 68-81. Lebovic, J. H., & Ishaq, A. (1987). Military burden, security needs, and economic growth in the Middle East. Journal of Conflict Resolution, 31(1), 106-138. Lessmann, C. (2013). Foreign direct investment and regional inequality: A panel data analysis. China Economic Review, 24, 129-149. Levine, R., & Renelt, D. (1992). A Sensitivity Analysis of Cross-Country Growth Regressions. The American Economic Review, 82(4), 942-963. Lim, D. (1983). Another look at growth and defense in less developed countries. Economic Development and Cultural Change, 31(2), 377-384.
144 Lin, J. Y., & Liu, Z. (2000). Fiscal decentralization and economic growth in China. Economic Development and Cultural Change, 49(1), 1-21. Little, J. S. (1978). Locational decisions of foreign direct investors in the United States. New England Economic Review, 4, 43-63. Liu, T., & Li, K. W. (2001). Impact of liberalization of financial resources in China’s economic growth: evidence from provinces. Journal of Asian Economics, 12(2), 245-262. Liu, T., & Li, K. W. (2006). Disparity in factor contributions between coastal and inner provinces in post-reform China. China Economic Review, 17(4), 449-470. Liu, X., & Wang, C. (2003). Does foreign direct investment facilitate technological progress?: Evidence from Chinese industries. Research Policy, 32(6), 945-953. Long, C., Yang, J., & Zhang, J. (2015). Institutional impact of foreign direct investment in China. World Development, 66, 31-48. Lucas, R. E. (1988). On the mechanics of economic development. Journal of Monetary Economics, 22(1), 3-42. Lucchetti, R., Papi, L., & Zazzaro, A. (2001). Banks’ inefficiency and economic growth: A micromacro approach. Scottish Journal of Political Economy, 48(4), 400-424. Lynn, R. (2006). Race differences in intelligence: an evolutionary analysis. Washington Summit Publishers. Lynn, R. (2010). In Italy, north–south differences in IQ predict differences in income, education, infant mortality, stature, and literacy. Intelligence, 38(1), 93-100. Lynn, R. (2012). IQs in Italy are higher in the north: A reply to Felice and Giugliano. Intelligence, 40(3), 255-259. Mankiw, N. G., Phelps, E. S., & Romer, P. M. (1995). The growth of nations. Brookings Papers on Economic Activity, 1995(1), 275-326. Mankiw, N. G., Romer, D., & Weil, D. N. (1992). A contribution to the empirics of economic growth. The quarterly journal of economics, 107(2), 407-437. McConnell, J. E. (1980). Foreign direct investment in the United States. Annals of the Association of American Geographers, 70(2), 259-270. Mishra, D., Mody, A., & Murshid, A. P. (2001). Private capital flows and growth. Finance & Development, 38(2), 2-2. Mullen, J. K., & Williams, M. (2005). Foreign direct investment and regional economic performance. Kyklos, 58(2), 265-282.
145 Mulligan, C. B., & Sala-i-Martin, X. (1993). Transitional dynamics in two-sector models of endogenous growth. The Quarterly Journal of Economics, 108(3), 739-773. Murphy, K. M., & Topel, R. H. (2013). Some basic economics of national security. The American Economic Review, 103(3), 508-511. National Development and Reform Commission. (2015). Vision and Actions on Jointly Building Silk Road Economic Belt and 21st-Century Maritime Silk Road: China. Neumark, D., & Muz, J. (2016). The “business climate” and economic inequality. Review of Income and Wealth, 1(62), 161-180. Newman, R. J. (1983). Industry migration and growth in the South. The Review of Economics and Statistics, 76-86. Nicoletti, G. (2002). Institutions, economic structure and performance: Is Italy doomed?. ISAE Annual Report “Monitoring Italy, 129-197. Pan, M., Widner, B., & Enomoto, C. E. (2012). Growth and crime in contiguous states of Mexico. Review of Urban & Regional Development Studies, 24(1-2), 51-64. Pedroni, P., & Yao, J. Y. (2006). Regional income divergence in China. Journal of Asian Economics, 17(2), 294-315. Perucca, G. (2014). The role of territorial capital in local economic growth: evidence from Italy. European Planning Studies, 22(3), 537-562. Peters, A., & Fisher, P. (2004). The failures of economic development incentives. Journal of the American Planning Association, 70(1), 27-37. Petraglia, C. (2002). Total factor productivity growth and public capital: the case of Italy. Studi Economici. Phelps, E. (1961). The golden rule of accumulation: a fable for growthmen. The American Economic Review, 51(4), 638-643. Phillips, K. L., & Chen, B. (2011). Regional growth in China: An empirical investigation using multiple imputation and province-level panel data. Research in Economics, 65(3), 243-253. Picci, L. (1999). Productivity and infrastructure in the Italian regions. Giornale degli Economisti e Annali di Economia, 329-353. Piffer, D., & Lynn, R. (2014). New evidence for differences in fluid intelligence between north and south Italy and against school resources as an explanation for the north–south IQ differential. Intelligence, 46, 246-249. Plaut, T. R., & Pluta, J. E. (1983). Business climate, taxes and expenditures, and state industrial growth in the United States. Southern Economic Journal, 99-119.
146 Porter, M. E. (1998). Clusters and competition: new agendas for companies, governments, and institutions. Harvard Business School Press. Porter, M. E. (1998). Clusters and the new economics of competition (Vol. 76, No. 6, pp. 77-90). Boston: Harvard Business Review. Putnam, R. D., Leonardi, R., & Nanetti, R. Y. (1994). Making democracy work: Civic traditions in modern Italy. Princeton university press. Ramsey, F. P. (1928). A mathematical theory of saving. The Economic Journal, 38(152), 543- 559. Ran, J., Voon, J. P., & Li, G. (2007). How does FDI affect China? Evidence from industries and provinces. Journal of Comparative Economics, 35(4), 774-799. Rappaport, J., & Sachs, J. D. (2003). The United States as a coastal nation. Journal of Economic growth, 8(1), 5-46. Resende, G. M. (2011). Multiple dimensions of regional economic growth: The Brazilian case, 1991− 2000. Papers in Regional Science, 90(3), 629-662. Romer, P. M. (1986). Increasing returns and long-run growth. Journal of Political Economy, 94(5), 1002-1037. Romer, P. M. (1987). Crazy explanations for the productivity slowdown. NBER Macroeconomics Annual, 2, 163-202. Romer, P. M. (1990). Endogenous technological change. Journal of Political Economy, 98(5, Part 2), S71-S102. Romp, W., & De Haan, J. (2007). Public capital and economic growth: A critical survey. Perspektiven der Wirtschaftspolitik, 8(S1), 6-52. Rosenthal, S. S., & Strange, W. C. (2001). The determinants of agglomeration. Journal of Urban Economics, 50(2), 191-229. Sachs, J. (1997). Geography and economic transition. Cambridge: Harvard Institute for International Development. Saha, D., Fikri, K., & Marchio, N. (2014). FDI in US metro areas: the geography of jobs in foreign-owned establishments. Brookings Institution, June, 20. Samuelson, P. A. (1958). An exact consumption-loan model of interest with or without the social contrivance of money. Journal of Political Economy, 66(6), 467-482. Saxenian, A. (1996). Regional advantage. Harvard University Press. Scully, G. W. (2003). Optimal taxation, economic growth and income inequality. Public Choice, 115(3), 299-312.
147 Shaver, J. M., & Flyer, F. (2000). Agglomeration economies, firm heterogeneity, and foreign direct investment in the United States. Strategic Management Journal, 1175-1193. Shenggen, F. A. N., & Zhang, X. (2004). Infrastructure and regional economic development in rural China. China Economic Review, 15(2), 203-214. Shepotylo, O. (2012). Spatial complementarity of FDI: the example of transition countries. Post- Communist Economies, 24(3), 327-349. Shiu, A., & Heshmati, A. (2006). Technical change and total factor productivity growth for Chinese provinces: a panel data analysis. Bonn, Germany: Institute for the Study of Labor (IZA) Solow, R. M. (1956). A contribution to the theory of economic growth. The Quarterly Journal of Economics, 70(1), 65-94. Stansel, D. (2005). Local decentralization and local economic growth: A cross-sectional examination of US metropolitan areas. Journal of Urban Economics, 57(1), 55-72. Stiglitz, J. E. (2000). Capital market liberalization, economic growth, and instability. World Development, 28(6), 1075-1086. Sussell, J., & Thomson, J.A. (2015). Are Changing Constituencies Driving Rising Polarization in the U.S. House of Representatives?. RAND Corporation. Swann, G., Prevezer, M., & Stout, D. (1998). The dynamics of industrial clustering: International comparisons in computing and biotechnology. Oxford University Press. Tuan, C., Ng, L. F., & Zhao, B. (2009). China's post-economic reform growth: The role of FDI and productivity progress. Journal of Asian Economics, 20(3), 280-293. Turner, C., Tamura, R., & Mulholland, S. E. (2013). How important are human capital, physical capital and total factor productivity for determining state economic growth in the United States, 1840–2000?. Journal of Economic Growth, 18(4), 319-371. U.S. Department of State. (2015). Italy Investment Climate Statement. Uzawa, H. (1965). Optimum technical change in an aggregative model of economic growth. International Economic Review, 6(1), 18-31. Vaona, A. (2008). Regional evidence on financial development, finance term structure and growth. Empirical Economics, 1(34), 185-201. Wang, Y. (2013). Fiscal decentralization, endogenous policies, and foreign direct investment: Theory and evidence from China and India. Journal of Development Economics, 103, 107- 123.
148 Wasylenko, M. (1997). Taxation and economic development: the state of the economic literature. New England Economic Review, (Mar), 37-52. Whalley, J., & Xian, X. (2010). China's FDI and non-FDI economies and the sustainability of future high Chinese growth. China Economic Review, 21(1), 123-135. Whynes, D. K. (1979). The economics of Third World military expenditure. Palgrave Macmillan UK. Wilson, A. (1995). The Donbas between Ukraine and Russia: The use of history in political disputes. Journal of Contemporary History, 30(2), 265-289. Xie, D., Zou, H. F., & Davoodi, H. (1999). Fiscal decentralization and economic growth in the United States. Journal of Urban Economics, 45(2), 228-239. Yao, S. (2006). On economic growth, FDI and exports in China. Applied Economics, 38(3), 339- 351. Yekelchyk, S. (2007). Ukraine: Birth of a modern nation. Oxford University Press. Youtie, J., & Shapira, P. (2008). Building an innovation hub: A case study of the transformation of university roles in regional technological and economic development. Research Policy, 37(8), 1188-1204. Yu, K., Xin, X., Guo, P., & Liu, X. (2011). Foreign direct investment and China's regional income inequality. Economic Modelling, 28(3), 1348-1353. Zeng, D. Z., & Zeng, D. Z. (2011). How do special economic zones and industrial clusters drive China's rapid development?. Washington, DC: World Bank. Zhang, H., Zhang, H., & Zhang, J. (2015). Demographic age structure and economic development: Evidence from Chinese provinces. Journal of Comparative Economics, 43(1), 170-185. Zhang, T., & Zou, H. F. (1998). Fiscal decentralization, public spending, and economic growth in China. Journal of Public Economics, 67(2), 221-240. Zheng, X., Li, F., Song, S., & Yu, Y. (2013). Central government's infrastructure investment across Chinese regions: A dynamic spatial panel data approach. China Economic Review, 27, 264-276.
149