Categories
Africa Economics Student Essays

The Impact of Elections on FDI in East Africa

Sub-Saharan Africa and other developing regions suffer from a lack of foreign direct investment. This paper analyzes how democracy and certainty factors in elections impact FDI inflows in Kenya, Tanzania, and Uganda from 1971-2015.

Abstract

Sub-Saharan Africa and other developing regions suffer from a lack of foreign direct investment (FDI), while developed regions take in the vast share of FDI.1 However, it is critical for regions like Africa to attract FDI due to its positive economic effects. Scholarly literature identifies many economic and political factors as influencing FDI inflows and investor decision-making. This paper analyzes how democracy and certainty factors in elections impact FDI inflows in Kenya, Tanzania, and Uganda from 1971-2015. Comparing each country’s average change in FDI inflows during election and non-election years evaluates how significant elections, in general, are in altering FDI. Further comparison of average election democracy and outcome certainty to changes in FDI inflows around elections tests the impact of these election factors on FDI inflows. My findings reveal no consistent relationship between changes in FDI inflows and electoral democracy, certainty, or elections in general. Showing elections as insignificant to changes in FDI inflows challenges the theories that elections, election democracy, or outcome certainty influence investment. If true, governments attempting to attract FDI would see more benefits from prioritizing improvements in their political institutions and enacting favorable economic policies, rather than addressing election factors.


Introduction

Just before you skydive, two things likely go through your mind: “holy cow, that looks incredible!” and, a few seconds later, “but I really hope they packed this parachute right.” Investors have similar, albeit less exciting, thoughts when they are considering whether or not to invest in a developing country. When deciding whether to build a factory in a rapidly growing African country, investors think about the numerous rewards like market access, tax breaks, and low wage costs. However, they also weigh the potential risks associated with that country, such as expropriation, tax hikes, or inflation. So far, the risks of investing have largely outweighed the benefits when it comes to foreign direct investment (FDI) in Africa. In other words, investors have not been jumping out of the plane.

Compared to other regions, Africa has received the lowest share of FDI inflows since 1990.2 Since “FDI occurs when a firm based in one country builds a new plant or factory, or purchases an existing one, in a second country,”3 FDI has been lauded as one of the best forms of investment for developing countries to receive. Firms being physically present, can lead to local job creation, skill and knowledge transfers, and increased tax revenue.4 These positive externalities make it crucial for countries like Kenya, Tanzania, and Uganda to attract more FDI. Scholars have long debated what factors influence FDI and how significant they are. My paper investigates elections in an attempt to help clarify this factor’s relationship to FDI inflows and how countries in Africa can better address investors’ concerns to attract FDI.

Investor Considerations

To begin, I will give a brief overview of the literature on the common factors thought to influence FDI. Firms consider economic factors when making investment decisions, however, since policy changes have a significant impact on the economy, firms also look to political factors when investing. Election events have the potential to change a country’s political environment and, resultantly, the economic factors thought to impact a firm’s investment decisions.

Economic Factors

All firms look for characteristics in foreign countries that allow them to increase their profits by both reducing certain costs and capitalizing on profits. Specifically, firms look for what are called locational advantages when they invest directly in a country. There are three general subsets of locational advantages: natural-resource investments, market-oriented investments, and efficiency-oriented investments.5 

Natural-resource investments occur when a foreign firm is looking to extract and utilize large amounts of natural resources found in another country.6 This type of investment has been prevalent in Africa since the colonial era and, even recently, there is still a positive association between FDI and natural resource wealth.7 Common natural resources that firms desire access to are fossil fuels, various minerals (e.g., gold, copper, lithium, etc.), and agricultural products. Dunning adds that resource investments can also be driven by the desire to gain access to cheap unskilled or semi-skilled labor8 or to acquire new technical, managerial, or marketing knowledge.9 Notably, this last point–gaining knowledge–matters less to firms investing in a developing country, since they are usually the ones bringing technological and managerial skills with them.10 

Market-oriented investments arise when a firm attempts to enter a large and rapidly growing local consumer market to sell its products.11 Entering these emerging markets is a priority to firms wishing to maximize profits and capture greater market share.12 Since market size and growth potential are key investor considerations, firms look at a country’s partnership in free trade areas13 (such as the recent African Continental Free Trade Area (AfCFTA)) and bilateral trade agreements. Firms also weigh how culturally similar the new market is to their home market, since adapting to more dissimilar markets will lead to higher entry costs for the firm.14 Similarly, prior experience in a foreign market is valuable to a firm because it reduces the uncertainty about market conditions and the costs spent adapting to the new market.15

Finally, efficiency-oriented investments stem from the availability of low-cost factors that are used in a firm’s production process.16 According to Dunning, firms look for one of two factors when making this type of investment. Firstly, firms may invest to take advantage of “differences in the availability and relative cost of traditional factor endowments”17,18 like labor, energy, and infrastructure. Second, in countries with similar markets and income levels, firms may invest to profit from “differences in economies of scale and scope, differences in consumer tastes, and supply capabilities.”19 In this case, factor endowments matter less than “created” differences in the market such as consumer behavior, local competition, the quality of institutions, and government economic policies.20

Stability and Policy Factors

Economic factors ultimately drive firms’ investment decisions. However, political and macroeconomic stability and government policies play a significant role in shaping many of the economic factors important to firms. Politics are especially prevalent in FDI, since the act of investing requires the company to own immobile assets in a country (i.e., a factory, farm, etc.). Once the firm builds the assets, the country’s government where it invested can change policies that the firm will have to accept.21 For this reason, firms pay close attention to policy, regulations, and stability.

Political instability consists of events such as anti-government demonstrations, assassinations, cabinet changes, constitutional changes, coups, government crises, purges, revolutions, and riots.22 According to two Global Investment Competitiveness (GIC) surveys of multinational executives in 2017 and 2019, political stability was the most important factor in their firm’s decision to invest.23,24 Earlier literature results compiled by Moreira in 2009 further supports this finding.25 The reason for this emphasis on political stability is “that the host government may change the rules of the game within the industry where the multinational company is active.”26 In other words, a political disruption can lead to a disruption in the market that proves harmful for a firm’s investment. In fact, the negative impacts of political instability are so strong that they can spill over into neighboring countries. Especially in land-locked countries like Uganda, who rely on their neighbors to export their goods, instability may cost firms when exporting their goods. Thus, investors care about the stability and trade relations of neighboring countries in addition to those they are investing in.27

More evidence that firms understand this connection between politics and the market stems from investors ranking of macroeconomic stability and the legal and regulatory environment as top concerns after political stability.28,29 Macroeconomic policies resulting in high inflation, tax rates, and cost of labor can create an unfavorable investment environment.30 Moreover, specific policy acts like expropriation, delays in issuing business permits, or breaching a government contract31 can be indicative of regulatory risk and poor property protections–both of which also deter FDI.32 On the other hand, governments can, and do, enact policies to attract investors. Ethiopia, for example, uses a combination of tax holidays (a policy of initially not taxing foreign firms), subsidized land, and no private sector minimum wage to attempt to entice foreign investors.33,34 

Elections 

As you can tell, there are many factors that investors use to measure whether or not to make an investment. Elections are significant since they allow investors to gauge political and policy stability within a country by analyzing an election’s democracy and outcome certainty. Furthermore, investors look at elections themselves as a cause of uncertainty. Although there is agreement that elections do influence investment, there is not academic consensus on how, or to what extent, election factors like democracy, outcome certainty, and policy uncertainty influence investment. 

Democracy

Touchton explains that investors “use elections as a proxy for political unrest, policy stability, respect for the law and property rights.”35 Rather than analyzing individual aspects of complex political institutions, investors consider how democratic an election was in order to judge the political institutions and the government holding the election. Investors reason that, if the government holding the elections respects election laws, then it can also be expected to respect property rights and legal contracts.36 Democratic institutions are viewed as checking autocratic leaders from enacting selfish, sweeping policy changes such as expropriation or raising taxes.37,38 Fairer, more democratic elections represent less political risk and, thus, a less risky investment environment. Multiple academic studies have indeed found that increases in democracy do correlate to higher levels of FDI,39,40,41 giving support to this argument. 

Uncertainty

Notably, however, there is no agreement among academics on whether more democratic institutions, which are measured via elections, actually encourage more foreign investment. Some academics suggest that investment is negatively affected by higher levels of democratization.42 The core facet of this argument is that democratic institutions create a more uncertain political environment than authoritarian regimes. Firstly, authoritarian regimes can provide firms with better deals since leaders are insulated from public pressure or domestic interests such as labor unions–both of whom would increase labor costs by advocating for higher wages and stricter labor laws.43 Moreover, firms may think that they can bribe authoritarian leaders to secure extra-legal protections or benefits, like seizing land or mineral rights.44 Second, in more democratic regimes, elections have an increased possibility of unseating leaders. New leaders may have different policy preferences, which could lead to increased uncertainty about how new policies will change the investment environment.45 Authoritarian leaders are more unlikely to be replaced by an election, thus indicating that policy will likely remain consistent. Hence, in a situation where election outcomes are more certain or less democratic, investors may respond positively since they believe the incumbent regime will better secure their investment. 

While investors look to electoral democracy to measure governments and political institutions, they also view election events themselves as potentially harmful to the investment environment. As I just mentioned, elections, even ones with certain outcomes, can lead to a leadership change that can result in policy changes. In addition, incumbent leaders may delay making regulatory interventions until after elections.46 Leaders running for reelection either do not want to enact regulations that will lose them votes or are simply too busy campaigning to focus on regulatory action.47 However, after an election, leaders can implement new policies that may harm investors with more impunity. For investors, this means that the pre-election period contains significant uncertainty about what policies will be enacted after the election concludes. Indeed, findings by Julio and Yook confirm that FDI does drop before an election and resume afterwards due to policy uncertainty, especially in a competitive election.48

Beyond causing policy uncertainty, election events can provoke political instability in the form of riots, coups, and post-election violence. While Gossel finds that post-election violence does not appear to be a significant consideration for investors,49 it is still worth noting that such events might make elections a generally less auspicious time to invest. 

Hypotheses

Based on the literature I have just discussed, it is reasonable to conclude that elections should have an impact on FDI inflows. Those with political power shape national economic policies that can either harm or help investors. Elections provide investors with a visible method for assessing the political, and resulting economic, environment within a country. Specifically, investors use election democracy and outcome certainty to measure risk, however, there is disagreement about how investors interpret these factors and how investment is affected. Finally, elections themselves may generate policy uncertainty and political instability, thereby deterring foreign investment.

To test the existing literature, I propose three hypotheses. First, I hypothesize that FDI inflows will decrease during election years in general due to increased policy uncertainty and political instability. Second, I propose that FDI inflows will decrease as elections become less democratic due to the more recent findings that democracy correlates to greater investment. Third, I propose that FDI inflows will decrease as election outcomes become less certain due to firms desire to avoid policy uncertainty. 

For clarity, democracy is defined by the presence of both political and civil liberties. Political liberties involve “the right of all adults to vote and compete for public office, and for elected representatives to have a decisive vote on public policies,”50 while civil liberties are “rights to free expression, to organize or demonstrate, as well as rights to a degree of autonomy.”51 Election result certainty pertains to how likely a candidate, usually the incumbent, is to remain in power after an election.

Additionally, I should emphasize that I am using these hypotheses to determine the relationship between election factors and FDI inflows. As I have shown, the literature does not agree about how election variables like democracy or outcome certainty impact elections. Thus, my hypotheses should be viewed as testable statements, rather than pre-confirmed statements of fact.

Methodology

Description and Overview

To ascertain the relationship between elections and FDI inflows, I combine data on both into a single dataset to make a side-by-side comparison of executive election’s levels of democracy and outcome certainty in Kenya, Tanzania, and Uganda. I then compare the total averaged sums of each election factor to the total average change in FDI inflows for each country. This allows me to test both my second hypothesis, that FDI inflows will decrease as elections become less democratic, and my third hypothesis that FDI inflows will decrease as election outcomes become less certain. To test my first hypothesis, that FDI inflows will decrease during election years, in general, I compare changes in FDI inflows in election years to non-election years (see Graphs 1-3). 

Using these methods, I am able to demonstrate a correlation, or lack thereof, between election democracy and certainty factors, and changes in FDI flows over time and between countries. Notably, my analysis does include a number of limitations stemming from difficulty isolating election variables’ impacts, limited sample size, equally weighted election variables, use of binary data, and incomplete FDI data for Uganda (see Methodological Appendix).

Kenya, Tanzania, and Uganda

Of the 54 African countries, I chose just three–Kenya, Tanzania, and Uganda–to analyze for the following reasons. First, all three countries are presidential republics with a majority-elected president and parliaments elected through a combination of majority vote and some proportional representation.52 One small difference is that Kenya has a bicameral parliament, consisting of a Senate and National Assembly.53 Choosing countries with similar government structures is necessary since the prevalence of majoritarian and proportional representation, coupled with the number of veto players (political actors who need to agree to enact a policy54), in a political system are both found to impact what government policies are implemented.55,56 To better equalize policy factors, I picked three countries with relatively similar political institutions to ensure that one of the countries I chose was not more or less able to implement policy changes. This would negate the possibility that one country would endemically have higher policy uncertainty due to the potential for more rapid policy change.

Second, I selected my three countries because they have presidential and parliamentary elections at the same time (as opposed to staggering elections for the two branches). This meant that the impact of an election would be more distinct in election years and easier to measure. The impact (and election data) of elections was similar, which allowed me to study executive election results as a proxy for elections in general. 

Finally, my countries are in close geographic proximity to each other. As Jared Diamond discusses in Guns, Germs, and Steel, location differences alter disease prevalence, agricultural capacity, and, ultimately, development outcomes.57 By investigating three countries in close proximity, I control for climate, disease outbreaks (such as AIDS), wars, and other situational phenomena that impact FDI. For instance, during the AIDS outbreak we might expect to see a universal drop in FDI since the countries would likely be similarly affected. But, because of this universal effect, an election’s impact on FDI change will remain distinct. Hence, doing this will allow me to better see past universal changes in FDI caused by external phenomena, and instead pinpoint changes caused by elections. 

NELDA and World Bank FDI Inflow Data

The National Elections Across Democracy and Autocracy (NELDA) dataset asks 58 questions related to elections from 1945-2015 and the events surrounding them.58 All 58 questions are asked for every executive and legislative election (separately), year, and country. Answers are a binary variable (i.e., “yes” or “no”), however, when data is not available or when the answer is inapplicable the answer is coded “N/A.” Broadly, I chose the dataset since it provided data on individual aspects of elections over a significant time span.59 Survey data, on the other hand, did cover a longer time span, but was inconsistent and did not allow me to test election year factors.

I used the World Bank FDI Net Inflows in U.S. dollars (USD) dataset for FDI inflows.60 This dataset detailed the dollar amount of FDI that entered each country every year from 1970 to 2019. Where data is unavailable it is labeled ‘0.’ I picked this dataset because it allowed me to measure the change in FDI inflows during election and non-election years over a forty-year span.

Testing the Impact of Elections on FDI

To determine what relationship FDI inflows have to elections in general, I plotted the yearly dollar amount change in FDI inflows from 1970 to four years after each country’s last recorded election year before 2015 (see Appendix 8).61 After making the plot, I highlighted election years and the years following the election and looked for any consistent increases or decreases in FDI inflows over time in each country. Finally, I compared my findings for each country to identify trends across countries (see Graphs 1-3). 

If my first hypothesis is supported, there should be a decrease in FDI inflows in election or post-election years in comparison to other yearly changes in FDI inflows (i.e., a positive relationship). These decreases should occur consistently in each country over time. However, if my hypothesis is not true, I should see either a consistent increase in FDI inflows during election or post-election years (i.e., an inverse relationship), or no discernible trend in FDI inflows across countries over time (i.e., no relationship).

Testing the Impact of Election Factors on FDI

Overview

To analyze the relationship between FDI inflows, election democracy, and outcome certainty, I selected questions from the NELDA dataset to represent election democracy and (separately) outcome certainty, assigned the answers a numeric value from 1 to -1, then added up all the values for each election year. To compare the democracy and outcome certainty scores between countries, I averaged the sum total of each category’s scores. My FDI inflow values measured time spans before, during, and after each executive election cycle to isolate election year changes. Specifically, I took the average change in FDI inflows from the two years before the election year, the election year (or the average of the election year and the year before), and the two years following the election year. To summarize the total impact of the election, I averaged the sum of two years before and after each election. And finally, I added all those sums up and recorded the average percent change in FDI inflows for all election cycles in each country (see Appendix 8). 

Pairing the total average of election democracy and (separately) outcome certainty scores with the average percent change in FDI inflows allowed me to see how these two election factors impacted the change in FDI inflows (see tables 1-3 and Appendices 1-3). Comparing each country’s results allowed me to identify trends across countries and make conclusions about the relationship between the two (see Graphs 4-6).

Supporting my second hypothesis should mean that there is a decrease in average FDI inflow changes alongside lower democracy scores across countries (i.e., a positive relationship). Likewise, supporting my third hypothesis would also show a decrease in average FDI inflow changes with lower outcome certainty scores. Should my hypotheses be untrue, there should be either an increase in average FDI inflow changes with lower democracy or outcome certainty scores (i.e., a negative relationship), or no consistent trend between FDI inflow changes and election factor scores (i.e., no relationship).

NELDA Data Selection and Manipulation 

Before continuing, I would like to explain why I selected and changed the data as I did. To begin, I only used NELDA questions from executive elections. Executive and legislative answers are almost all identical since both elections occurred in the same years, making analyzing both results redundant. Additionally, political power in African politics is often dominated by a single leader who usually heads the executive branch, making executive elections more important than legislative.62 In the cases of Kenya, Tanzania, and Uganda, all three have presidential systems making them more likely to have power centralized in the executive branch. 

When choosing which questions to include, I selected groupings and questions to measure an election’s democracy and the uncertainty surrounding an election’s outcome. These categorizations were selected to indicate aspects of democracy and outcome certainty. To reiterate the definitions I gave in my hypothesis section, democracy within elections is characterized by the right of all adults to vote, compete for public office, rights to free expression, and to organize or demonstrate.63 Election outcome certainty refers to how likely it is for a candidate, usually the incumbent, to remain in power after the election.

Under democracy, I created four groupings of questions: First Multiparty Elections, Perceived Behavioral Fairness, Behavioral Fairness, and Election Violence (see Appendix 7). These four factors were chosen to isolate which parts of an election democracy (if any) have the most significant connection to investor decision-making. First, Multi-party Elections refers to “when a country holds the first multi-party elections after a significant period of non-democratic rule, or when a country transitions from single-party elections to multi-party elections.”64 The term “multi-party” means that several parties that are nominally independent of the ruling regime are allowed to run.65 I chose this grouping to measure how free individuals are to compete for public office and express themselves through making choices between candidates. Perceived Behavioral Fairness denotes “whether the conduct of the election can be deemed to be free and fair,”66 while Behavioral Fairness “refers to overt election manipulation.”67 In this grouping, actions that might restrict voting or election competition are included to measure election democracy. Finally, Election Violence refers to violence that in part occurred in relation to an election. Violence, especially when used by the incumbent regime against demonstrators, can indicate an undemocratic election and failure to respect election outcomes or allow free expression.

To measure election outcome certainty, I used questions on: whether the incumbent regime was confident of victory, was the incumbent leader contested in the election, and did the incumbent or a chosen successor (someone chosen by the incumbent leader or regime to replace the incumbent) run. All three categories help quantify how certain it was that an election would be won by the incumbent regime or a non-incumbent beforehand. 

For the questions themselves, I assigned the answers to each question a numeric value of either 1, 0, or -1. For questions under election democracy, answers that indicate a more democratic election are coded as “1,” while answers that indicate less democracy are coded as “-1.” Similarly, for the election certainty factor, answers that indicate an election’s outcome is more certain are coded as “1,” while answers indicating less certainty are coded as “-1.” All answers without data, those labeled “N/A,” are coded as “0” for all factors. Importantly, for the First Multi-party Elections factor, all “No” answers are also coded as “0” since a “Yes” answer can only occur once, and “No” would not indicate a trend towards less democracy. Also, NELDA questions 21 and 22 are combined so that if the incumbent runs a successor automatically cannot be chosen, making question 22 “0.” (For above, see Appendix 7.)

While using individual questions was an option, the binary nature of the answers in the NELDA dataset means that trends are difficult to observe since the data is either very erratic (e.g., yes, no, yes, no) or flat (e.g., yes, yes, yes, yes). Combining multiple answers solves this problem by allowing multiple binary answers to build on each other and show trends. Picking a single question also increases the chance of having a false positive. But, by combining several related questions to a common election factor, the chance of a flawed response is decreased.

Aligning numerically valued answers with the years when elections took place allows me to show changes in election factors over time. Of note, since Uganda has had fewer elections than either Kenya or Tanzania, I found it necessary to take the average of all the numeric answer values for each individual factor. Doing this allows me to compare factors across countries. Without using averages, the raw amount of data available would skew the results. Thus, averaging the sum of all available data per factor solves this issue, and makes comparison between countries possible (see Appendices 1-3).

FDI Inflow Data Selection and Manipulation

Measuring the total change in FDI from two years before an election to two years after, allows me to see the impact that an election had on FDI inflows during that time (see Appendix 8). Taking the average of these two-year cycles negated anomalous yearly changes that might not be caused by an election. 

Another important point I should qualify is what “election year(s)” actually are. For elections that occur within the last six months of the year, the election year is only that year. However, if an election occurs in the first six months of a year, then the election year is the averaged sum of the election year and the previous year. In this instance, the two years before the election become the two years before the year prior to the election. So, for a February election in 1996, the average of 1996 and 1995 is the “election year” and the two years before the election is the average of 1994 and 1993. The reason I did this is that I wanted to equally incorporate the lead-up to the election in the FDI inflow data. Kenya and Tanzania generally hold their election around October and December, while Uganda holds theirs around February and March. 

Were I only to include the FDI inflows for the actual election year in Uganda, I would not be assessing the complete follow-up to the election, but rather the results of it. If my hypothesis is that FDI inflows change due to election events, it would be misleading not to include the full election event (lead-up included) into my “election year” FDI value. Hence, averaging the year of and year before is my attempt to include the entire election event in my FDI inflow numbers. 

Lastly, my decision to average the sum of the change in FDI inflows two years before an election to two years after was done to compare election-specific FDI inflow data between countries. This choice was made primarily since I needed total average FDI data to compare with the NELDA data. 
 

Findings

Impact of Elections on FDI

Graphs 1-3 show each country’s election years in red and the year after in purple. Bars represent the change in FDI inflows from the year before to that year. For all three countries, there is no consistent change in FDI inflows during election or post-election years themselves or when compared to non-election years. This pattern remains true over time and across countries.

Case Study: Kenya

Election History Overview

In 1963, Kenya held its first pre-independence elections. The two main parties participated, but the Kenya African National Union (KANU) party won a majority in both houses of parliament and the party’s leader, Jomo Kenyatta, became prime minister.68 Kenyatta quickly consolidated political power through several constitutional amendments, including the one making him president.69 Before the 1969 election, KANU banned any opposition party candidate from running, making the country a one-party state.70 

On Kenyatta’s death in 1978, his vice president, Daniel arap Moi, took over and was later elected president in the 1979 election. The power transfer occurred smoothly due to a rise in coffee prices and the leadership of now vice president, Mwai Kibaki.71 An uprising in 1990 by anti-Moi forces organizing a pro-democracy rally prompted KANU to amend the constitution to allow opposition parties to run.72 However, in 1992 Moi won reelection and KANU maintained a strong majority despite the elections being marred by thousands of deaths and mass violence.73

In 2002, Moi selected Uhuru Kenyatta–Jomo Kenyatta’s son–to run as KANU’s presidential candidate. The National Rainbow Coalition (NARC) party, with Mwai Kibaki leading the ticket, won the presidency and a majority in parliament, ending KANU’s 40-year reign.74 The 2007 election was marked by mass violence and the displacement of over 600,000 people after Kibaki was declared the winner likely due to election manipulation.75 The situation was so bad that the UN and African Union stepped in to broker a power-sharing agreement between Kibaki and his opponent, Raila Odinga. In 2013, Uhuru Kenyatta, running under The National Alliance (TNA) party, successfully defeated Odinga with just over 50% of the vote in a pre-runoff election that remained largely peaceful.76
 

Democracy Findings 

Democracy scores strongly correlate to major events in Kenya’s election history (see Table 1). Higher scores occur in 1979 (when Moi is elected after Kenyatta’s death), 1992 (Kenya holds its first multi-party elections in decades), 2002 (KANU and Moi are defeated by opposition), and in 2013 (in which Uhuru Kenyatta won a highly competitive election). 

The first multi-party elections in 1992 did not correlate to a marked increase in FDI inflows (there is a larger change in FDI inflows in 1974 and 2007). Perceived behavioral fairness is consistently much lower than actual behavioral fairness. From 1992 on, behavioral fairness hovers around a high of 4-6, while perceptual fairness remains low and even drops in 1992 and 1997. Finally, the election violence score does approximate the occurrence of violent elections in Kenya, like the notable 2007 election.

Result Certainty Findings

Negative certainty scores correlated strongly to power transitions and highly competitive elections (see Table 1). Kenya’s last three elections, two of which resulted in an incumbent losing power, have the lowest certainty scores. Notably, certainty does not appear to have a relationship to changes in FDI inflows. 

Case Study: Tanzania

Election History Overview

In 1962, Tanzania held its first post-independence election, which brought to power Julius Nyerere and the Tanganyika African National Union (TANU) party. Soon after, the United Republic of Tanganyika and Zanzibar became one country in 1964, with Nyerere as president and Abeid Amani Karume as vice president. A constitutional amendment made the country a one-party state until 1992. Of note, TANU merged with another party, making Chama Cha Mapinduzi (CCM) the sole official party.77

Nyerere ran unopposed until 1985, when he resigned and appointed Ali Hassan Mwinyi as his successor. Mwinyi won, unopposed, in 1985. In 1992, the constitution was amended to allow multi-party elections, CCM candidate Benjamin Mkapa was elected president and the party maintained a huge majority in parliament. Mkapa went on to win reelection in 2000, but the election was marred by voter fraud allegations and violence.78

In 2005, Jakaya Mrisho Kikwete, the CCM candidate, was elected president, with the party continuing to hold a strong majority in parliament. CCM lost some seats to the opposition in 2010, but Kikwete still easily won reelection with 61% of the vote. Kikwete was replaced by CCM candidate John Magufuli in 2015.79 The election was closely contested, but Magufuli still managed to win.80

Democracy Findings

Democracy scores for Tanzania correlate to notable election events and increasing democracy in the country (see Table 2). Tanzania’s first multi-party election in 1995, is marked by a large democracy score increase, from 2 to 12. The following year the country’s score decreased to a 6 in alignment with the violence and fraud surrounding the 2000 election. Both 2010 and 2015, show increases in democracy as the CCM is losing seats in parliament and contending with tighter races.

First multi-party elections are the only factor that correlates to a marked increase of 240% in FDI inflows. From 1975-1990, perceived fairness is relatively equal to actual fairness but, from 1995 on, perceived fairness is 2-3 points lower than actual fairness. Both metrics increase in 1995 following the country’s first multi-party elections. The only exception to this is in 2000, where perceived fairness falls back to pre-1995 levels at a dismal -1. Elections became slightly more violent in 2000 alone, which is consistent with Tanzania’s history of political stability.81

Result Certainty Findings

Election result certainty fell dramatically in 2005, and less so in 2010 and 2015 (see Table 2). This tracks well with new candidates taking power in 2005 and 2015, and the CCM facing increasing opposition in elections. Result certainty did not show a correlation to the change in FDI inflows.

Case Study: Uganda

Election History Overview

Milton Obote was elected prime minister in 1969, months before the country gained its independence. Obote enacted an unpopular new constitution that appointed him president and redistricted the country. In January of 1971, General Idi Amin, who had been helping Obote maintain power, staged a coup and assumed power.82

In 1979, after causing chaos in Uganda and attempting to invade Tanzania, Amin was deposed by Yusufu Lule, who was then replaced by Godfrey Binaisa two months later. Obote and his party were then reelected in 1980 in a controversial election. However, Lule and Yoweri Museveni refused to accept Obote’s victory and formed the National Resistance Movement (NRM) to fight the government. Obote was deposed by a military coup in 1985.83

Museveni and the NRM took power and drafted a new constitution in 1986 without an election. A new constitution in 1995 paved the way for the 1996 election, in which Museveni easily won. Technically, the election was a “no-party” election. For the 2006 election, Museveni extended his term limits and also allowed parties to form for the first time since 1980. Despite this change, Museveni and the NRM won reelection in 2006 and 2011 by large margins. Notably, these large margins may be due, in part, to election manipulation. In 2006, Museveni’s main rival was imprisoned for months leading up to the election, and elections are thought to be influenced by military intimidation and bribery.84
 

Democracy Findings

Overall democracy scores in Uganda correlate to significant election markers (see Table 3). In 1996, the country experienced its first election since 1980 and received a corresponding high democracy value of 4. In 2006, the country received -1 despite this being the country’s first official multi-party election. However, this value is easily explained when we note that while opposition parties were not officially allowed in prior elections, unofficial opposition was still allowed to run.85 In addition, Museveni’s extension of his term limit and imprisoning his main opposition both point to the undemocratic nature of the election. 

Like Tanzania, Uganda’s first official multi-party elections in 2006 correlated to the highest increase in FDI inflows (206%), while the 1996 elections saw the second highest change at 170%. No other groupings showed a correlation to changes in FDI inflows. Similar to both Kenya and Tanzania, perceptual fairness was lower than actual behavioral fairness–meaning elections were fairer than they were perceived. Finally, election violence remained flat at 0, with a slight improvement in 2011. 

Result Certainty Findings

Election results were very certain and remained at 1 for every election (see Table 3). There was no connection between result certainty and FDI inflows.

Impact of Election Factors on FDI

Comparing the average change in FDI inflows before and after an election to total average democracy scores for each country reveals that Kenya and Tanzania’s greater democracy scores equate to greater changes in FDI inflows. Uganda, having the lowest average democracy score and highest change in FDI inflows, is an exception to this trend. While this large average change in FDI inflows may be an inflated result (see Methodological Appendix), without a larger data set containing several more countries, I can find no consistent relationship between democracy factors and changes in FDI inflows (see Graphs 4 and 5).

A comparison of the average change in FDI inflows before and after an election to total average certainty scores for each country shows that Kenya and Tanzania have the lowest certainty scores and the lowest average change in FDI inflows. Uganda had the highest score and the highest change in FDI inflows. Notably, Kenya and Tanzania had equivalent certainty scores, making the increase or decrease not proportionate to the change in FDI. Due to the small number of questions used to characterize election certainty, this result is not surprising. Since the results are not proportional to each other, I cannot find a consistent relationship between election result certainty and changes in FDI inflows (see Graphs 4 and 6).

Discussion

My findings present a challenge to previous literature. The lack of a consistent relationship between changes in FDI inflows and electoral democracy, certainty, and elections in general undermines two key notions. First, investors may not use elections to gauge the political environment within a country or make election decisions. Contrary to past literature, my results show no indication that FDI inflows changed based on an election, and little evidence that inflows varied in connection to democracy or outcome certainty. If true, investors may not accept the premise that election democracy is an effective proxy for a country’s political structure or that elections cause enough policy uncertainty to delay investment. Instead, investors may look to the other economic and political factors mentioned in my investor considerations section. For countries looking to attract FDI, this may mean holding elections or boosting election integrity will not change the levels of FDI entering a country. Rather, governments may want to consider adopting more favorable economic policies or structural reforms to improve investment flows.

Given my small selection of countries, my results are far from definitive. However, my election democracy and certainty scores were predictive of major election events in all three case studies. This affirms the accuracy of my study’s methodology. Future studies may want to replicate my approach and sample more countries to further confirm the absence of a relationship between FDI inflows, election factors, and elections in general. In addition, further studies should compare policy actions and economic factors (such as inflation, deficits, and taxes) against changes in FDI inflows to determine their relationship to investment. 

Conclusion

As Africa strives to develop economically, the continent will need to improve its low FDI numbers, given this type of investment’s benefits to job creation, skill transference, and increasing tax revenues among other factors. To accomplish this, African countries will need to address investors’ economic concerns to attract more investment. Literature identifies many economic factors that may alter the investment environment. Political factors such as stability and policy are purported to greatly influence economic factors and also contribute to investment decisions. Election democracy is thought to be used by investors as a proxy of political institutions, while election outcomes are believed to create policy uncertainty. These factors directly affect the economic factors that investors care about, and thus, may impact FDI inflows.

To test these findings, I compared changes in FDI to election democracy, election outcome certainty, and elections in general for Kenya, Tanzania, and Uganda from 1971-2015. My findings indicate no consistent relationship between changes in FDI and elections across countries or over time. Due to my limited sample size, future studies should analyze more countries to confirm this paper’s findings. That said, my results challenge the contention that elections play a significant role in investor decision-making. If true, the role of other economic and political factors in influencing FDI should be studied further by academics. Moreover, governments wishing to attract FDI should prioritize improving their political institutions and enacting favorable economic policies, rather than election factors.





_________

Appendix 1: Kenya Election Factors and FDI Inflow Changes Summary

Appendix 2: Tanzania Election Factors and FDI Inflow Changes Summary

Appendix 3: Uganda Election Factors and FDI Inflow Changes Summary

Appendix 4: Kenya Election Factors Data

Appendix 5: Tanzania Election Factors Data  

Appendix 6: Uganda Election Factors Data

Appendix 7: Election Factors Questions and Numeric Score Assignments 

Appendix 8: Foreign Direct Investment (FDI) Data

Methodological Appendix

Limitations

Absence of Uganda Data

Uganda made equivalent analysis difficult because the country has held far fewer elections than Kenya or Tanzania, and because the World Bank had spotty FDI inflow data on the country in the 1980s. Uganda has only held four executive elections starting in 1996, while Kenya and Tanzania have each held nine elections between 1971 and 2015. The lack of election data means that the average change in FDI inflows for Uganda in my second and third hypothesis tests are inflated. In the mid-1990s to early 2000s, Africa saw a massive up-tic in FDI. Since I only take FDI inflow data for Uganda starting in 1996, my changes in FDI inflows show very large increases, but not the slow or even negative numbers associated with the 1970-80s. By taking the average of these very positive numbers I see a much higher number than in both Kenya and Tanzania since their FDI data does include the 1970s to 1980s. Hence, the Ugandan data in my second and third hypothesis tests is questionable.

In an attempt to obtain more accurate FDI data for Uganda, I created “phantom election dates” at five-year intervals before 1996. By simulating elections going back to 1971 and measuring the change in FDI inflows, I hoped to get a more accurate total average FDI inflow value to compare with Kenya and Tanzania. Doing this did produce a significantly lower average change in FDI, however, the number was inaccurate because the World Bank did not have FDI data for Uganda in 1981, 1984, 1986, or 1987.

Isolating Election Impacts

Next, isolating whether changes in FDI inflows are caused by elections or election factors is challenging. I have tried to account for extraneous variables through my choice of countries, two-year FDI inflow averages, and measurement of change in FDI rather than raw FDI inflow numbers. Despite this, it remains difficult to tell whether FDI numbers are fluctuating in response to elections or an outside factor. At best, my study can show a relationship between FDI inflows and elections.

Equally Weighted Election Factors

Since I did not give questions or election factors weighted values, all election factors seem to be equally influential to investor decision making. Investors may see certain aspects of an election as more indicative of a country’s situation than others. Moreover, they may care more about certain factors than others. Considering this, not weighing my questions or factors could lead to inaccurate findings.

Binary NELDA Data

As a result of NELDA answers being either “yes” or “no,” the answers do not leave room for complexity. Questions like “Is the country said to be in good relations with the US before the elections?” are labeled with a simple answer that lacks depth. This calls into question the validity of NELDA answers. Though my approach of giving the answers values and adding them does somewhat negate the possibility of one answer being wrongly labeled, more accurate results would be better achieved through a ranked response. 

Limited Number of Countries

Using only three countries also poses a risk to the accuracy of my results. Especially considering that Uganda has such a lack of data, any of my findings should be considered suspect due to my lack of data points. Were I to include more countries in my study, say all Sub-Saharan African countries, then trends in election factors and FDI inflows would be more reliable since I would have more data points.  



ENDNOTES:

1. Thomas Oatley, International Political Economy : International Student Edition, 6th ed. (New York, NY: Routledge, 2018), 164, https://doi.org/10.4324/9781351034661.
2. Oatley, 166.
3. Oatley, 163.
4. Oatley, 176–77.
5. Oatley, 168.
6. Oatley, 168.
7. Berhanemeskel Solomon, “Determinants of Foreign Direct Investment Inflow to Ethiopia: Time Series Analysis” (Thesis, Addis Ababa University, 2018), 12, http://etd.aau.edu.et/handle/123456789/18625.
8. John H Dunning and Sarianna M Lundan, Multinational Enterprises and the Global Economy (Cheltenham, UK ; Northampton, MA: Edward Elgar, 2008), 68, https://catalog.hathitrust.org/Record/007272172.
9. Dunning and Lundan, 69.
10. Geoffrey C. Sigei, “Factors Influencing Investment in Kenya by Multinational Corporations (MNCS) in the Transport Sector: The Case of Rift Valley Railways (K) Limited” (Thesis, University of Nairobi., 2009), 27, http://erepository.uonbi.ac.ke/handle/11295/23430.
11. Oatley, International Political Economy, 168.
12. Sigei, “Factors Influencing Investment in Kenya by Multinational Corporations (MNCS) in the Transport Sector,” 24.
13. Sigei, 24.
14. Sigei, 24.
15. Sigei, 24.
16. Oatley, International Political Economy, 169.
17. Dunning and Lundan, Multinational Enterprises and the Global Economy, 72.18. Solomon, “Determinants of Foreign Direct Investment Inflow to Ethiopia,” 6. 
19.  Dunning and Lundan, Multinational Enterprises and the Global Economy, 72. Dunning and Lundan, 72.
20. Raymond Vernon, Storm Over the Multinationals: The Real Issues (Harvard University Press, 1977).
21. Sandrina Berthault Moreira, “The Determinants of Foreign Direct Investment: What Is the Evidence for Africa?,” Poiésis – Revista Do Programa de Pós-Graduação Em Educação 2, no. 3 (June 30, 2009): 94, https://doi.org/10.19177/prppge.v2e3200983-104.
22. Peter Kusek and Andrea Antonia Lim Silva, What Investors Want: Perceptions and Experiences of Multinational Corporations in Developing Countries (Rochester, NY: The World Bank, 2018), 8, http://hdl.handle.net/10986/29566.
23. Abhishek Saurav and Ryan Kuo, The Voice of Foreign Direct Investment : Foreign Investor Policy Preferences and Experiences in Developing Countries, Policy Research Working Papers (The World Bank, 2020), 12, https://doi.org/10.1596/1813-9450-9425.
24. Moreira, “The Determinants of Foreign Direct Investment,” 94.
35. Sigei, “Factors Influencing Investment in Kenya by Multinational Corporations (MNCS) in the Transport Sector,” 25.
36. Sigei, 26.
37. Kusek and Silva, What Investors Want, 8.
38. Saurav and Kuo, The Voice of Foreign Direct Investment, 12.
29. Moreira, “The Determinants of Foreign Direct Investment.”
30. Saurav and Kuo, The Voice of Foreign Direct Investment, 16.
31. Saurav and Kuo, 3.
32. Ashwin Parulkar, “African Land, Up for Grabs,” World Policy Journal 28, no. 1 (March 1, 2011): 103–10, https://doi.org/10.1177/0740277511402796.
33. Marie Durane, “‘MADE IN ETHIOPIA’ — THE NEW NORM IN THE GARMENT INDUSTRY: EBSCOhost,” Sustainable Development Law & Policy 15, no. 2 (2015): 24.
34. Michael Touchton, “Campaigning for Capital: Fair Elections and Foreign Investment in Comparative Perspective,” International Interactions 42, no. 2 (March 14, 2016): 275, https://doi.org/10.1080/03050629.2016.1093475.
35. Touchton, 275.
36. Mancur Olson, “Dictatorship, Democracy, and Development,” American Political Science Review 87, no. 3 (September 1993): 571, https://doi.org/10.2307/2938736.
37. Nathan Jensen, “Political Risk, Democratic Institutions, and Foreign Direct Investment,” The Journal of Politics 70, no. 4 (October 1, 2008): 1049, https://doi.org/10.1017/S0022381608081048.
38. Jo Jakobsen and Indra De Soysa, “Do Foreign Investors Punish Democracy? Theory and Empirics, 1984–2001,” Kyklos 59, no. 3 (2006): 383–410, https://doi.org/10.1111/j.1467-6435.2006.00338.x.
39. Philipp Harms and Heinrich W. Ursprung, “Do Civil and Political Repression Really Boost Foreign Direct Investments?,” Economic Inquiry 40, no. 4 (2002): 651–63, https://doi.org/10.1093/ei/40.4.651.
40. Matthias Busse, “Transnational Corporations and Repression of Political Rights and Civil Liberties: An Empirical Analysis,” Kyklos 57, no. 1 (2004): 45–65, https://doi.org/10.1111/j.0023-5962.2004.00242.x.
41. Quan Li and Adam Resnick, “Reversal of Fortunes: Democratic Institutions and Foreign Direct Investment Inflows to Developing Countries,” International Organization 57, no. 1 (February 13, 2003): 202, https://doi.org/10.1017/S0020818303571077.
42. Nathan M. Jensen, “Democratic Governance and Multinational Corporations: Political Regimes and Inflows of Foreign Direct Investment,” International Organization 57, no. 3 (ed 2003): 593, https://doi.org/10.1017/S0020818303573040.
43. Touchton, “Campaigning for Capital,” 278.
44. Kexin Chen, He Nie, and Zhenyu Ge, “Policy Uncertainty and FDI: Evidence from National Elections,” The Journal of International Trade & Economic Development 28, no. 4 (November 21, 2018): 421, https://doi.org/10.1080/09638199.2018.1545860.
45. J. Tyler Leverty and Martin F. Grace, “Do Elections Delay Regulatory Action?,” Journal of Financial Economics 130, no. 2 (November 1, 2018): 426, https://doi.org/10.1016/j.jfineco.2018.06.010.
46. Chen, Nie, and Ge, “Policy Uncertainty and FDI,” 420.
47. Brandon Julio and Youngsuk Yook, “Policy Uncertainty, Irreversibility, and Cross-Border Flows of Capital,” Journal of International Economics 103 (November 1, 2016): 23, https://doi.org/10.1016/j.jinteco.2016.08.004.
48. Sean Joss Gossel, “FDI and Elections in Sub-Saharan Africa,” The European Journal of Development Research 32, no. 4 (September 1, 2020): 1163, https://doi.org/10.1057/s41287-020-00260-5.
49. Raymond D Gastil and Richard W Cottam, Freedom in the World: Political Rights and Civil Liberties, 1978 (New York: Freedom House, 1978), 7.
50. Gastil and Cottam, 7.
51. “Countries – The World Factbook,” accessed March 17, 2021, https://www.cia.gov/the-world-factbook/countries/.
52. “Countries – The World Factbook.”
53. George Tsebelis, Veto Players: How Political Institutions Work (Princeton University Press, 2002), https://www.jstor.org/stable/j.ctt7rvv7.
54. Carolyn L. Evans, “A Protectionist Bias in Majoritarian Politics: An Empirical Investigation,” Economics & Politics 21, no. 2 (2009): 278–307, https://doi.org/10.1111/j.1468-0343.2009.00346.x.
55. Witold Henisz and Edward Mansfield, “Votes and Vetoes: The Political Determinants of Commercial Openness,” International Studies Quarterly, March 1, 2006, 189–212, https://doi.org/10.1111/j.1468-2478.2006.00398.x.
56. Jared M. Diamond, Guns, Germs, and Steel: The Fates of Human Societies (W. W. Norton & Company, 1997).
57.  Susan D. Hyde and Nikolay Marinov, “Which Elections Can Be Lost?,” Political Analysis 20, no. 2 (2012): 191–210, https://www.jstor.org/stable/23260172.
58. Susan D. Hyde and Nikolay Marinov, “Which Elections Can Be Lost?,” Political Analysis 20, no. 2 (2012): 191–210, https://www.jstor.org/stable/23260172.
59. Other datasets, like the Corruption Perceptions Index (CPI), did not address election factors or cover years before Africa’s influx of FDI since they only went back to the mid 1990s.
60. “Foreign Direct Investment, Net Inflows (BoP, Current US$) | Data,” accessed February 22, 2021, https://data.worldbank.org/indicator/BX.KLT.DINV.CD.WD.
61. By yearly change in FDI inflows, I mean the dollar change from the year before to the current year. For example, the change in FDI inflows for 2000 would be the dollar amount increase or decrease in FDI from 1999 to 2000.
62. Todd J. Moss and Danielle Resnick, African Development: Making Sense of the Issues and Actors (Boulder: Lynne Rienner Publisher, Inc., 2018), 42.
63. Gastil and Cottam, Freedom in the World, 7.
64. Hyde and Marinov, “Which Elections Can Be Lost?,” Code Book pg. 6.
65. Hyde and Marinov, Code Book pg. 6.
66. Gossel, “FDI and Elections in Sub-Saharan Africa,” 1153.
67. Gossel, 1153.
68. “Kenya,” Encyclopedia Britannica, accessed March 17, 2021, https://www.britannica.com/place/Kenya.
69. Musambayi Katumanga and Institute for Education in Democracy (Kenya), National Elections Data Book: Kenya, 1963-1997. (Nairobi: Institute for Education in Democracy, 1997), 2, https://aceproject.org/ero-en/regions/africa/KE/kenya-kenya-election-history-1963-2013-the-kenyan.
70. Katumanga and Institute for Education in Democracy (Kenya), 3.
71. “Kenya.”
72. Katumanga and Institute for Education in Democracy (Kenya), National Elections Data Book, 7.
73. Katumanga and Institute for Education in Democracy (Kenya), 7.
74. “Kenya.”
75. Katumanga and Institute for Education in Democracy (Kenya), National Elections Data Book, 17.
76. “Kenya.”
77. “Tanzania,” Encyclopedia Britannica, accessed March 17, 2021, https://www.britannica.com/place/Tanzania.
78. “Tanzania.”
79. “Tanzania.”
80. “Tanzania Overview,” National Democratic Institute, April 8, 2008, https://www.ndi.org/sub-saharan-africa/tanzania.
81. “Tanzania Overview.”
82. “Uganda,” Encyclopedia Britannica, accessed March 17, 2021, https://www.britannica.com/place/Uganda.
83. “Uganda.”
84. “Uganda.”
85. “Uganda,” Encyclopedia Britannica, accessed March 17, 2021, https://www.britannica.com/place/Uganda.





Processing…
Success! You're on the list.

___
Latest






___
Editor’s Picks

By Daniel Grasso

Daniel Grasso is a transfer student and UCLA graduate with a B.A. in political science. His studies have focused on international political economy, especially in Sub-Shanahan Africa. At UCLA, he serves on the Bruin Political Union board as Director of Policy Innovation. He currently interns for the Corporate Counsel on Africa (CCA) and has worked for several successful U.S. Senate and
California Assembly campaigns from 2016 to 2020. Daniel aspires to pursue a career in international affairs, encouraging partnerships and development in growing regions like Africa.

LinkedIn: https://www.linkedin.com/in/dan-grasso/
Email: daniel.c.grasso@gmail.com

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s