In recent decades, multinational corporations (MNCs) have increased their foreign direct investment (FDI) activities worldwide and developing countries have been the main destination. FDI is the international flow of firm-specific asset such as production technologies, managerial and organizational practices, and trademarked brands. Usually, MNCs realize FDI through the establishment of new production facilities, or the merger and acquisition of an existing firm. The literature has traditionally distinguished two motives for FDI. Resource seeking FDI refers to firms willing to reduce production costs by relocating production to foreign countries abundant in necessary inputs, such as labor or natural resources. Market seeking FDI sees firms entering countries to produce goods and services for local sales. This strategy aims at reducing the burden of trade barriers and transportation costs. During the period between 1990 and 2013, the amount of FDI inflows had jumped from 35,000 to 77,000 million dollars and their share directed to the developing nations was 52% of global FDI inflows in 2013 (UNCTAD  ). To explain such massive evolution in international economic flows, a vast literature has focused on how governments compete to attract FDI. Indeed, the democratization process occurring in the 1990s, with a large number of countries adopting free market economic policies, paved way for interstates competition to attract foreign investment. Indeed, FDI would yield infrastructure upgrading, promote technology transfer and improve managerial knowledge; those factors are of most importance to compete in global markets (Dunning  ). FDI are also credited with boosting economic growth (Alfaro et al.  ) and reducing unemployment (Spiezia  ; Vacaflores  ). Such interstate competition has been witnessed mainly in developing countries as developed countries already exhibit high quality infrastructure, property rights protection and an educated work force, thus reducing the need for them to compete. Unsurprisingly, relaxing FDI regulation such as those on ownership, the creation of special economic zones, and lowering taxation on corporations have been widespread policies to attract FDI1. For instance, Kobrin  showed that during the period between 1992 and 2001, 95% of the 1086 individual policy changes either lessened restrictions on inflows of FDI or provided incentives to attract them.
However, interest groups’ influence has been credited to be of great importance in shaping FDI policy changes. Using an original dataset of FDI regulation covering the period between 1962-2000 over 150 countries, 57 industries and 12 specific regulations that represent the most common barriers to FDI2, Pandya  found that industries in which firms invest to gain market access such as service industries are more likely to be restricted, as powerful domestic producers lobby for protection. FDI policies and regulations may thus serve as a way to protect the national economy. So far, this theoretical debate has been investigated by political economists who focused on the interaction between FDI and trade policy. It started with the seminal argument that foreign firms try to circumvent protectionist barriers that impediment exports sales via setting up local production (Corden  ). Other formulation refers to the quid pro quo FDI, where the motive for foreign investment is the pre-emption barriers that might otherwise be implemented (Baghwati  )3. Dinopoulos and Wong  have shown that increased FDI by foreign entities will reduce protection threats against foreign imports in the host country. Quid pro quo investment is thus described as “indirect rent-seeking” via consensual policies (Hillman and Ursprung  ). The quid pro quo investor seeks a liberal trade policy, and will eventually be deterred if trade liberalization does not occur. For instance, Zhao  has demonstrated how such investment is pre-empted if labor unions lobby a host country’s government for protectionist policies. Similarly, Hillman and Ursprung  argue that such foreign investment could not take place if the foreign investor acquires a domestic monopoly firm. Confronting to import competition, the foreign investor supplying the domestic market may thus lobby for protection. The focus of this literature is whether the presence or threat of protection gives rise to FDI. On the other hand, other papers have studied how the presence of multinational firms affects the emergence of protection. In Hillman and Ursprung  ’s model, domestic and multinational firms exposed to foreign competition lobby for protection in the jurisdictions where they have plants4. Contrary to Hillman and Ursprung  , Ellingsen and Warneryd  developed a model where the domestic industry lobby for less protection, as a high level of protection could attract more FDI aimed at avoiding trade barriers and thereby harming domestic firms.
In this paper, we seek to unify these two dimensions of FDI politics within the framework of the common agency model of lobbying (Grossman and Helpman  ). Contrary to conventional political economy models that assume citizens are rationally ignorant of the functioning of the economy and the impact of economic policies (Converse  ; Page and Shapiro  ; Deli et al.  ), we assume that the population exhibit heterogeneous political beliefs concerning international economic flows and in particular on the effects of FDI liberalization policies. In doing so, we fill a methodological gap highlighted by Rodrik  who insists on the necessity to account for ideas and beliefs in political economic modelling. The main contribution of the paper is to show that information provision related to the beneficial effects of FDI liberalization policies may paradoxically lead to less liberalization. This counter-intuitive result lies with the structure of the political economy setting and interdependence effects between lobbies’ activities and the evolution of citizens’ beliefs.
The paper is organized as follows: Section 2 presents the framework of our model in which we consider that society is not homogeneous with respect to citizens’ beliefs over the effects of FDI liberalization policies. In Section 3, we develop the common agency model of lobbying. We analyze the setting of the reform in FDI policies as a political compromise pressured by interest groups. The latter are composed of a domestic industry lobby and a foreign MNC lobby that have opposed interests. They both lobby the government by offering political campaign contributions. In Section 4, we investigate the policy distortion resulting from interest groups’ influence by comparing the equilibrium level of policy change with its socially optimal value. We show that interest groups’ influence may not be socially distortive under certain conditions. Section 5 explores the exogenous effect of information provision on the equilibrium level of FDI policy change. The last section presents some concluding remarks.
2. The Model
In this section we develop the assumptions of our basic economic model. We describe the utility functions and their arguments. In the spirit of Tagkalakis  and Jaeck and Kim  , we formulate the unemployment equation as follows:
where u is the unemployment rate; w is the nominal wage; and is the inflation rate. Note that represents the real wage. We assume that there is a level of FDI policies, , that corresponds to the current level of FDI rules and regulations which is normalized to 1 so that its log is zero. Therefore, is a composite index that represents the strength of liberalization of FDI policies with and. Such indices consider FDI policies that affect foreign investors’ market access6. The liberalization process is specified as a continuous variable; when v increases, the liberalization process becomes stronger. The parameter a measures the impact of the real wage, (a > 0) and measures the impact of FDI liberalization policies (). This latter assumption is supported by empirical evidence showing the positive impact of FDI in reducing the unemployment rate in Latin America (Vacaflores  ), Central Europe (Radosevic et al.  ), the Czech Republic (Dinga and Munich  ) and in Italy (Spieza  ). Therefore, the unemployment rate increases with the real wage () and decreases with the index v (i.e. deviation from the current state of rules and regulation of FDI; so if v = 0, no FDI liberalization policy is undertaken).
The population is composed of N citizens. Following Caplan  ’s empirical contribution we assume that there are two types of citizens: those who believe in the beneficial effects of liberalization of FDI policies on reducing unemployment (informed citizens) denoted by, and those who do not (misinformed citizens) denoted by 7. Alternatively, informed and misinformed citizens are assumed to hold correct and biased beliefs, respectively. Thus, the society is not homogenous with respect to citizens’ beliefs over liberalization policies8.
This representative informed citizen believes in the positive effects of liberalization on the level of unemployment. When the deregulation process increases, the unemployment rate decreases, and thus the utility increases. This is represented by. The term may be interpreted as the citizen’s belief over the effect of the liberalization process. We assume, indicating that informed citizens have correct beliefs,.
The representative utility function of a misinformed citizen is represented by:
The representative misinformed citizen does not believe in the beneficial effect of FDI liberalization policies leading to a misconception of the unemployment rate, with 10. When FDI liberalization process increases, the unemployment rate increases, and thus the utility decreases. This is represented by. However, it is assumed that external information related to the beneficial effects of liberalization policies, , compensates this bias. With, and. Thus, , everything else being equal11.
3. The Political Process
In this section we develop the political model which is based on the common agency framework (Grossman and Helpman  ). We characterize how interest groups’ influence affect the political equilibrium. We consider two competing firms that operate in the service industry and produce the same good. The formation of the FDI policies easing market access in the home country is subject to the influence of interest groups. For methodological purpose, we first assumed that no external information on the beneficial effects of FDI liberalization policies is undertaken, therefore is given. A domestic and a foreign MNC that belong to the home and foreign country, respectively, from distinct lobbies j, with j=d; f representing the domestic and foreign lobby respectively. Each lobby offers political contributions to the government. This money is for ‘direct political influence’ as it supports the electoral campaign for the reelection of an incumbent government12. As a reward, lobbies expect to obtain a better access to the legislator (Grossman and Helpman  ). In our case, both lobbies are concerned with the degree of liberalization of FDI policies, , but with opposed interests, and therefore, compete for influence. The domestic lobby supports a low level of liberalization as it seeks protection from competing foreign investors13. On the other hand, the MNC lobby supports more liberalization as it seeks market access. Following Olson  , we also assume that citizens (informed and misinformed) are too numerous to overcome the free-rider problem and as such they do not organize themselves into a lobby group. To examine how political contributions provided by lobbies affect the equilibrium level of FDI policies, we use a common agency model of lobbying (Grossman and Helpman  ) which is a two-stage game. The timing of the game can be expressed as following: In the first stage, each lobby presents to the government a contribution schedule, s(v), which is contingent upon the level of FDI policies chosen by the government. In the next stage, the government sets the liberalization policy and collects the political contributions. Analyzing the determination of the political equilibrium requires first to define the aggregate welfare of each lobby j.
The domestic lobby’s gross welfare function is represented by the domestic firm profit:
where is the total revenue with, and the total cost function.
The MNC lobby’s gross welfare function is represented by the MNC profit:
where is the total revenue with, and. We further assume that 14. The relationship between and shows that an increase in the FDI liberalization process raises the profit of the MNC as it could expand its production, and reduces the one of the domestic firm as it loses market shares15. Each lobby sets up its contribution schedule so as to maximize its net welfare, which is the difference between the gross welfare and its contributions. Then, the government chooses the policy that maximizes a weighed sum of the aggregate social welfare and the received contributions. Following Grossman and Helpman  we assume that contribution schedules are globally truthful as they reflect everywhere the lobby’s true welfare (Bernheim and Winston  ). With this global truthfulness assumption, the political equilibrium is the solution of a program in which the government seeks to maximize an objective function, which is the sum of the aggregate social welfare W and the interest of the active lobbies. It is such that16:
The parameter represents the interest group’s ability to influence the policymaker. It depends on exogenously determined factors such as political skills. A higher value of means that the lobby j can better influence the government that puts a higher weight on the contribution offered by lobby j. We assume that interest groups can have different lobbying efficiencies and can be either greater or lower than. The social welfare function, W, exhibits three components: the informed citizen’s welfare, the misinformed citizen’s welfare, and the contribution received by the government respectively:
where is the informed citizen’s aggregate welfare equal to:
And is the misinformed citizen’s aggregate welfare equal to:
Differentiating Equation (6) with respect to v, we obtain the first-order condition of government optimization:
Here, measures the government’s marginal benefits from promoting FDI liberalization policies. The second-order condition requires that this variation is decreasing in v such that. The relation in (10) contains a key characteristic of the model. Around the truthful equilibrium, a policy change induces a variation in each lobby’s contribution that is exactly equal to the variation in this lobby’s gross welfare.
In particular, a marginal increase in v induces variations in the domestic lobby and MNC’s contributions by the amount of and, respectively. In what follows, we define respectively and as the domestic lobby and the MNC lobby’s marginal willingness to contribute (MWTC) for a marginal change in the FDI liberalization process, v. In equilibrium, is negative, meaning that the domestic lobby is willing to offer less political contributions in exchange for a strengthening of the FDI liberalization process, as such policies reduce its total revenue. Also, is positive. This means that the MNC lobby is willing to increase its contribution when the FDI liberalization process strengthens as it could enhance the MNC to expand its production. Equation (10) shows that the political equilibrium level of FDI policy change is chosen as a compromise between political contributions from interest groups and social welfare.
4. Political Equilibrium Policy
In this section, our primary focus is to compare the two equilibria () and ().is the socially optimal level of FDI policy change, which corresponds to an equilibrium when there are no direct lobbying. is the equilibrium level of FDI policy change that results from the direct influence of lobbies. Analyzing the deviation between and will provide us with a better understanding of the distortion resulting from direct political influence. Using Equation (7) and rearranging, we first set out the socially optimal level of FDI policy change, which is the solution of the following first-order condition17:
The second order condition requires that. Equation (11) shows that relaxing FDI regulations increases the informed citizen’s welfare, decreases the misinformed citizen’s welfare and increases (decreases) the MNC lobby’s welfare and (the domestic lobby’s welfare), respectively. The optimal level of FDI policy reform should balance the four factors. From (11) we get:
To compare and, we evaluate Equation (10) at and given the first order condition we obtain:
In (13), with the second order condition, if is positive (negative), the equilibrium level of FDI policy change is higher (lower) than the social optimum. It is equal, and when. If, then . When the MNC is the only one lobbying actively, the equilibrium level is higher than the optimal level. Similarly, if, then. When the domestic lobby is the only one lobbying actively, the equilibrium level is lower than the optimal level. These results are expected. We consider the particular case where the two lobbies have the same political efficiency, such that. Combining (12) and (13), we get:
From Equation (14), the following proposition shows the relationship between and when:
Proposition 1: If the two lobbies have the same political efficiency (), the equilibrium level of FDI policy change is equal to the optimal level if the ratio of informed over misinformed citizens equals a critical value A. These results hold everything else being equal.
From Equation (14), when, the sign of depends on the sign of. Therefore, we have for , and thus equals. The intuition behind proposi-
tion 1 is that in the case where, when the level of policy change is optimally set, the marginal welfare effects of changing is equal to zero, so the value of depends on the two lobbies’ MWTCs. From Equation (13), implies that, the domestic lobby’s MWTC for a marginal change in FDI liberalization policies equals the MNC one, resulting in the equilibrium level equals to its socially optimal level.
Corollary: The equilibrium level is higher than its socially optimal value if this ratio falls below the critical value A. It is lower than the optimal level if the ratio increases above A. These results hold everything else being equal.
When and thus. Similarly, under the government has to balance the impacts of strengthening the level of liberalization on informed citizens’ welfare (), on misinformed citizens’ welfare () and on the contributions from the domestic lobby () and the MNC lobby (). Since, , and, the condition implies that when increases under the influence of the growing number of misinformed citizens, either must increase or must decrease. The intuition is that because misinformed citizens and the domestic lobby have common interests as they both support the status quo towards strong FDI regulations, interdependence effects occur between these two groups. Given that strengthening the liberalization process would decrease the welfare of misinformed citizens, the government has an incentive to weaken the liberalization process. This would incentivize the domestic lobby to reduce its contribution for a marginal tightening of FDI regulation as it indirectly benefits from the change in the welfare of the misinformed. Comparatively, the domestic lobby contributes less than the MNC lobby for a marginal change in the FDI liberalization policies, resulting in a higher equilibrium level of liberalization. Similarly, when and thus. The condition implies that when increases under the influence of the growing number of informed citizens, either must decrease or must increase. Given that strengthening the liberalization process would increase the welfare of informed citizens who become a large number in the society, the government has an incentive to do so. Because informed citizens and the MNC lobby have common interests as they both support more FDI liberalization, interdependence effects might also occur between these two groups. This would incentivize the MNC lobby to decrease its contribution for a marginal softening of FDI regulation as it indirectly benefits from the change in the welfare of the informed. Comparatively, the domestic lobby contributes more than the MNC lobby for a marginal change in the FDI liberalization policies, resulting in a lower equilibrium level of liberalization.
5. The Effect of Information Provision
In this section, we investigate the effect of an exogenous provision of information regarding the beneficial effects of FDI liberalization, , on the equilibrium level of FDI liberalization policy. We assume that such external information may come from different sources of the civil society such as think tanks for instance. Since , the sign of depends on the sign of since. By differentiating with respect to, we obtain:
With, we reach the following proposition on the effect of information provision on the level of FDI policy change. This proposition holds for any.
Proposition 2: Information provision on the beneficial effects of FDI leads to weaken the FDI liberalization process.
The political effects of information provision on the stringency of FDI policies depend on the impact of information provision on, the marginal benefit to the government arising from tightening the liberalization process. Here, the negative sign of shows that the effect of information provision leads paradoxically to less liberalization of FDI policies. This is because the marginal benefits to the government increase when the liberalization process is weakened. The intuition lies with interdependence effects between lobby groups and citizen’s welfare. Indeed, the increase in information provision compensates for the loss of misinformed citizens’ welfare resulting from more liberalization, By analogy with proposition 1, this would incentivize the MNC lobby to decrease its contribution for a marginal softening of FDI regulation as it indirectly benefits from the change in the welfare of misinformed citizens. Comparatively, the domestic lobby contributes more than the MNC lobby for a marginal change in the FDI liberalization policies, leading to less liberalization.
Recognizing that the political economy of FDI demand has so far received less attention as opposed to the study of FDI supply, this paper has proposed a challenging analysis seeking to unify these two dimensions of FDI politics. Using a common agency model of politics, we have proposed a modelling of a reform aimed at relaxing FDI policies that account for the impact of direct lobbying of a domestic lobby and a foreign MNC lobby which compete for influencing the FDI policy process affecting firms’ market access. The domestic lobby supports a low level of liberalization as it seeks protection from competing foreign investors. On the other hand, the MNC lobby supports more liberalization at it seeks market access. Contrary to conventional political economy models, we have assumed that the population exhibits heterogeneous political beliefs over the effects of FDI liberalization policies.
Our general model has suggested interesting results under specific conditions. First, proposition 1 has shown that when interest groups have the same lobbying efficiency, the lobbies’ influence is not necessarily distortive. Indeed, taking into account the heterogeneity of citizen’s beliefs over the effects of FDI policies, the equilibrium level of policy change is equal to its socially optimal value for a specific distribution of informed over misinformed citizens. Although conventional results of common agency models of lobbying have demonstrated that interest groups’ activities bias policy choices against unorganized stakeholders, our model shows that it is not always the case. This result lies with the structure of the political game where the two lobbies holding opposed interests may provide the same lobbying efforts. Second, we have shown that a change in the distribution of informed over misinformed citizens creates political distortions as the equilibrium outcome deviates from its socially optimal level. For instance, when the number of misinformed increases, we paradoxically observe a strengthening of liberalization of FDI policies. This counter-intuitive result has to do with interdependence effects that occur between the two groups who share a common interest, namely the domestic lobby and the misinformed. Second, our model has also accounted for the effect of exogenous provision of information on the beneficial effects of the liberalization of FDI policies. In extending proposition 1, proposition 2 has shown that the liberalization process is weakened when such provision occurs. This paradoxical outcome also results from interdependence effects between lobby groups and the evolution of citizens’ beliefs. This theoretical result has important policy implications especially for education policies aiming at creating a support for FDI and international economic flows in general. Indeed, findings of the literature showing that educated individuals are persuaded by the economic benefits of FDI and support them, may not be transferred into more FDI liberalization depending on the structure of the political game of well-organized interests.
For further research, our model has, however, the following limitations and could be extended in many directions. First, we have considered a general index of FDI policies that affect foreign investors’ market access. Further research would consider specific policies and integrate them into the industry’s profit functions. In addition, developing an economic model that accounts for material sources of utility such as wages would enrich the realism of the model and bring interesting insights given the labor market effects of FDI. Then, testing our result would bring interesting insights into the empirical literature of the political economy of FDI. Finally, the high level of generality of our model allows us to apply it to other important political economy issues such as those related to labor market regulations.
We thank participants of Research Seminar at the Department of Economics and Finance, United Arab Emirates University, February 2016. We are also thankful for the UAEU support, Grants number G00001588. Finally we thank Mr. Shahabuddin Abdulrouf, Mr. Enayatulah Abdal Ghafoor and Mr. Hamid Tajjudin for their excellent research assistance.
1The literature has focused on spatial econometrics to explore the extent of competition in tax, environmental standards, economic policy reforms, bilateral investment treaties and labor standards. The reader may refer to Cooray et al.  for an excellent survey of this literature.
2The United Nations identifies sixty distinct policy measures that countries use to restrict FDI (UNCTAD  ). Pandya restricts his analysis to the following: bans on foreign ownership, majority local ownership requirements, government monopoly, mandatory joint ventures, compulsory investment pre-screening, local content requirements, minimum exports quotas, discriminatory tax policy, caps on capital and profit repatriation, limits on access to foreign exchange, local employment minimums and mandatory local representation on boards of directors. Note that the first five policy instruments listed above restrict FDI by limiting foreign investors’ market access. The remaining seven are regulations that apply to foreign and not to nationally-owned firms.
3See Baghwati et al.  for a survey.
4They show, however, that the presence of multinational firms in an industry may encourage trade liberalization under specific conditions.
5Surveys from 1995, 1998, and 2001 included the question: “Do you consider that foreign investment, in general, is beneficial or is it harmful for the economic development of the country?” The 1998 and 2001 surveys ask a different but related question: “Do you strongly agree, agree, disagree, or strongly disagree with the phrase: foreign investment should be encouraged?” The author also provides a comprehensive test of FDI’s predicted effects on individual income. In support of Pandya  ’s argument, Leeson et al.  have highlighted that ideas matter in shaping citizens’ attitudes towards free market policies in the U.S. Relying on questions from the General Social Surveys that ask respondents their opinions about the government’s relationship to the economy, they found that where think tank spending is higher, citizens have more “pro-mar- ket” attitudes toward economic policy.
6As an example, and building on Pandya  , we assume that FDI regulations subjected to the softening process are related to bans on foreign ownership, majority local ownership requirements, mandatory joint ventures, or compulsory investment pre-screening.
7Based on the Survey of Americans and Economists on the Economy from Harvard University, Caplan  found that citizens would have an “anti-foreign” bias leading them to underestimate the benefits from free trade. We assume that such biased perception may also hold for the benefits of FDI liberalization policies.
8This assumption contrasts with the seminal standard assumption of citizens’ rational ignorance that prevails in the public choice literature. The question of the origin of biased beliefs is left out of the paper. Biased beliefs may result from exposure to activist groups (Saint-Paul  ; Jaeck  ) or ideology (MacDonald and Rabinowitz  ).
9The assumption is in line with the seminal results of MacCulloch et al.  who showed that individual happiness increases when the unemployment rate decreases.
10This assumption is motivated by Caplan  ’s empirical evidenced demonstrated an “anti-foreign bias” as well as Scheve and Slaughter  showing that individuals are less supportive of FDI if they perceive their jobs to be insecure.
11This assumption is motivated by Pandya’s  study that relies on the annual public opinion survey conducted in 18 Latin American countries. Results clearly indicate that attitudes in favor of FDI increase with the level of education.
12IDEA  lists 41 countries with bans or other forms of restrictions on foreign donations to political parties. This includes the United Kingdom, France, Brazil, Canada, and Argentina. However, there are some examples of countries which do not impose bans such as Australia, Colombia, Denmark and Germany (Aidt and Hwang  ). Gawande et al.  ) have shown that foreign lobbies significantly impact the setting of tariffs and non-tariff trade barriers in the U.S. and that foreign activity may be welfare-enhancing.
13This assumption is in line with Pandya  ’s results showing that industries in which firms invest to gain market access such as service industries are more likely to be restricted, as powerful domestic producers lobby for protection.
14Aitken, Harrison, and Lipsey  evidenced declining productivity for local firms competing with MNC as they face increased production costs due to greater labor demand.
15These assumptions are in line with FDI’s negative effects for local firms such as loss of market share due to increased foreign competition (Chari and Gupta  ).
16The objective function of the government without considering the global-truthfulness assumption would be as follow,. If one considers the global-truthfulness property, the original objective function is re-written as Equation (6).
17When deriving the following condition, we use the relationship and. This relationship occurs as a result of the global-truthfulness property of the contribution schedules.