• No results found

THE INFLUENCE OF CORRUPTION ON EXPORT PERFORMANCE

N/A
N/A
Protected

Academic year: 2021

Share "THE INFLUENCE OF CORRUPTION ON EXPORT PERFORMANCE"

Copied!
42
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

THE INFLUENCE OF CORRUPTION

ON EXPORT PERFORMANCE

AN EMPIRICAL ANALYSIS OF LATIN AMERICA

by

RIANNE VAN RAAIJ

University of Groningen Faculty of Economics and Business Master International Economics and Business

Supervisor: Dr. M.S.S. Krammer Co-assessor: Dr. A.A. Erumban

June 2015

(+31) 06 30420963 riannevanraaij@gmail.com

(2)

Abstract  

This paper examines the influence of corruption on the export performance in Latin America. Corruption is defined the misuse of power by a public official through accepting, soliciting or extorting a bribe as well as the offering of a bribe by a private agent in order to circumvent public policies and processes. Academic literature shows that corruption has a negative impact on firms because it leads to increased operational costs, uncertainty and deterioration of public finances, infrastructure and public services. Empirical evidence is provided on a firm level, based on data from the Enterprise Survey of the World Bank. No clear evidence is found on how corruption influences export performance.

(3)

Table  of  Contents  

Introduction ... 3 Literature Review ... 6 Corruption ... 6 Exports ... 8 Conceptual framework ... 12 Methodology ... 13

Background of export performance and corruption in Latin America ... 13

Data ... 16 Variables ... 17 Dependent variables ... 17 Independent variables ... 17 Control variables ... 18 Interaction variables ... 19 Model ... 19 Analysis ... 21 Results ... 22 Summary statistics ... 22 Model estimation ... 23 Robustness ... 26 Conclusion ... 29 Discussion ... 29 Implications ... 30

Limitations and future research ... 31

Bibliography ... 33

Appendices ... 38

Appendix A. Country Overview Latin America and the Caribbean ... 38

Appendix B. Country Sample ... 38

Appendix C. Results Berja-Berque Normality Test ... 39

Appendix D. Robustness Checks - Probit Estimation ... 39

Appendix E. Robustness Check - Tobit Estimation Export Intensity ... 40

(4)

Introduction  

Corruption has received renewed interest of managers due to the rise of emerging markets in the last decades. Firms have identified corruption as a challenge in their business practices (Enterprise Survey, 2010). World Bank (1997) defines corruption as ‘the misuse of public power for private gain’. A public official misuses his power when he accepts, solicits or extorts a bribe. Similarly, when a private agent offers a bribe to circumvent public policies and processes to gain competitive advantages he is guilty of corruption (World Bank, 1997; Transparency International, 2015). In general, low-income countries show higher levels of corruption because corruption is linked to poor quality of institutions and public sector functioning (Kimuyu, 2006; Lalountas et al., 2011). Academics argue that corruption affects the social and economic development of a country negatively (Mauro, 1995; Kimuyu, 2006; Charoensukmongkol and Sexton, 2011). More specifically, literature on the subject argues that corruption has a negative effect on firms because it raises operational costs, causes uncertainty, and diminishes public finances, infrastructure, and public services (Mauro, 1995; Tanzi, 1998; Gaviria, 2002; Kimuyu, 2006; Charoensukmongkol and Sexton, 2011; Rehman and Perry, 2013). In other words, corruption, in whatever form, influences the performance of firms (Tanzi, 1998; Gaviria, 2002).

(5)

monitoring and settlement and carry sufficient cash in case they need to bribe someone (De Jong and Bogmans, 2011).

About half of the interviewed Latin American firms, identified corruption as a major constraint to operations (Enterprise Survey, 2010). Analysis by Transparency International (TI) shows that Latin American countries still range among the lowest performing countries in the world in the Corruption Perceptions Index (CPI), with few exceptions such as Chile, Uruguay and Costa Rica (Transparency International, 2015). For over two decades, Latin America has failed to combat corruption (Transparency International, 2015). Since the 1990s, regional integration processes have been initiated, where Mexico and Brazil are seen as regional leaders that should carry the load in the fight against corruption (Gardini and Lambert, 2011). However, these countries are unable to take the lead in combating corruption. Last year, one of the biggest companies in Latin America, Petrobras was accused of corruption. Petrobras is a Brazilian state-controlled petroleum company producing more than 90 percent of petroleum in Brazil and owning all national refineries (Watts, 2015; Costas, 2015; The Economist, 2015). These allegations of corruption shed light on the widespread corruption activities throughout the private and public sector in Brazil, indicating that Latin American countries perform poorly in addressing corruption.

This paper aims to identify the impact of corruption on export. More specifically, the research question that will be answered in this paper is “How does corruption affect

export performance in Latin America?”

(6)

This paper does so by using detailed firm level data about corruption, export, and other firm characteristics, such as firm age, size and ownership, on 16 Latin American countries from the Enterprise Survey of the World Bank. The effects of corruption on export performance will be explained based on this data.

This study focuses on corruption and export performance in Latin America. The goal of this paper is to find empirical evidence that indicates that export performance is affected by corruption and in what way. In order to find grounds to refocus current public policies towards combating corruption. It is necessary to refocus public policies on combating corruption because in the last two decades anti-corruption initiatives by governments failed to fight corruption. Anti-corruption efforts are initiated to improve social and economic development. Thus, the impact corruption has on export performance is important to consider because firms are the engine of social and economic development.

(7)

Literature  Review  

Corruption    

Generally speaking, corruption is defined as ‘the misuse of public power for private

gain’ (World Bank, 1997). Corruption may consist of different actions, although each

of these has the same causes and outcomes (Tanzi, 1998; Gaviria, 2002). For example, a police officer who accepts money from drug traffickers, a custom agent who extorts businesses, or a politician who appropriates royalties; each of these people is guilty of corruption.

It has been widely accepted in the academic literature on this topic, that corruption restrains economic growth and lowers development outcomes. Zelekha and Sharabi (2012) explain that the theoretical literature about corruption describes transmission mechanisms with which corruption can negatively affect growth or other economic activities. These mechanisms are distortions in the allocation of resources within the economy, increased uncertainty in economic decision making, degradation of the legal mechanisms for settling business disputes, loss of leadership, reduced marginal productivity of capital, increased inequality in distribution of income and effect on small business sector (Zelekha and Sharabi, 2012). Besides, corruption affects the level and composition of FDI, increases the level of informality and lowers social spending as well as the ability to raise revenues (Mauro, 1995; World Bank, 1997; Gaviria, 2002). More importantly, international trade patterns are negatively affected by corruption. Although trade barriers are decreasing, bureaucratic requirements, such as permits, licenses, acquisition rules and registration requirements to engage in export contracts and sales, are still a burden (Carr and Outhwaite, 2009). These bureaucratic requirements can be time consuming, cumbersome and may distort important decision-making processes as well as the business environment.

(8)

needed to complete pre-shipment inspections or allowing illegal practices go unchecked. The second has an indirect impact on firms as it results in stealing and pilfering government resources. It thus worsens the business climate overall by raising the cost of credit, creating uncertainty and deteriorating public finances, infrastructure and public services (Gaviria, 2002; Carr and Outhwaite, 2009).

There are always two parties involved in a corrupt action, a bribe taker, usually a public official, and a bribe giver, a private agent. Bribery without theft occurs when public officials turn over the entire official proceeds to their government, only keeping the bribe (Kimuyu, 2006). Bribery with theft occurs when public officials conceal the transaction altogether, passing nothing to the government and allowing firms to pay a bribe that is lower than the official price. In this case, public officials have to create artificial shortages or stretch services. This form of corruption is beneficial to firms because it involves a lower price, thus promoting collusion between firms and officials (Kimuyu, 2006).

(9)

severe when institutions are bureaucratic. In order to test this relationship, the following hypothesis is formulated:

Hypothesis 1 (H1): Higher levels of bureaucracy result in higher levels of

corruption.

Exports  

In the past academics (De Groot et al., 2004; Huang, 2007) have discussed the influence of trade barriers and costs of trade on international export between countries. Their findings indicate that geographical, cultural and institutional differences are the main determinants for explaining potential trade and uncertainty. These determinants create additional costs that a firm would not experience if it were not exporting. On the other hand, export is a means of firm expansion, important for economic development. Some academics argue that reduced corruption can lead to more export growth and international trade (Kimuyu, 2006; Charoensukmongkol and Sexton, 2011).

Most academics look at export performance using a corruption-augmented gravity model based on a country level or firm level perspective (Anderson and Marcouiller, 2000; Dutt and Traca, 2010; De Jong and Bogmans, 2011; Zelekha and Sharabi, 2012). This paper looks from a firm level perspective at export performance and examines the export performance of firms by investigating the export intensity, which refers to the amount a firm is exporting.

(10)

pervasive nature of bribes. Firms that are unwilling to bribe experience difficulties such as delays in transaction processing time, which will discourage them from trade (Charoensukmongkol and Sexton, 2011).

De Jong and Bogmans (2011) distinguish corruption in international trade from corruption in general and corruption in an exporting economy from an importing economy. Corruption in general hampers international trade, especially exports, but the corruption and quality of institutions directly related to international trade, shows that bribes to customs enhance imports. This argument suggests that in countries with bad institutions, bribe payments would improve the situation for imports. No clear evidence is found that corruption in international trade enhances exports. According to some academics (Anderson and Marcouiller, 2000; Kimuyu, 2006; Dutt and Traca, 2010; Charoensukmongkol and Sexton, 2011), corruption tends to harm trade, particularly exports. Anderson and Marcouiller (2000), argue that corruption and imperfect contract enforcement create insecurity, thereby reducing international trade. They say that in countries with poor legal systems, corrupt officials create a hidden tax on trade by generating a price markup. Corruption creates additional costs, risk and uncertainty stemming from customs. Trade regulation officials may arbitrarily impose or adjust export rules and restrictions to garner bribes. Additionally, exporters may avoid trade with countries that exhibit high levels of corruption. Specific licenses and approvals from government agencies for international transactions are normally required. Thus, bureaucratic requirements and the complexity of trade regulations are barriers to internationalization for domestic firms. According to Dutt and Traca (2010), corruption creates a trade-taxing extortion effect because the amount of money the bribe costs increases with more corruption, leading to reduced incentives for exporters. However, they find evidence that in the case of high-tariff environments, corruption can have a trade-enhancing evasion effect, which means that corrupt officials allow exporters to evade tariff barriers.

(11)

exports depend on human capital effects such as organizational and operational management factors but no conclusive evidence is found (De Toni and Nassimbeni, 2001; Ganotakis and Love, 2012). Furthermore, the effects of the unpredictability of corruption and policies are expected to reduce trade even more, but the evidence is also inconclusive (De Jong and Bogmans, 2011). Besides, it could be possible that corruption and export intensity are sensitive to sample selection bias (Knack and Azfar, 2002). Kimuyu (2006) conducted research based on firm level data on African companies and demonstrates that firms experiencing corruption are less likely to export.

This study tests the hypothesis that export performance is influenced by corruption. The line of thinking that will be followed in this study is derived from the theory that states that corruption leads to lower export performance due to higher costs, uncertainty and a worsened business climate. Lee and Weng (2013) argue that home country bribery decreases firm exports. Therefore, the greater the amount of bribes that a firm pays to home country government officials, the lower the amount the firm exports. This will dampen the competitive position of the firm in the domestic and international market. More specifically, the expectation is that corruption influences export intensity negatively. Therefore, the following hypothesis is formulated:

Hypothesis 2 (H2): Higher levels of corruption will influence export intensity

negatively.

(12)

2010; Lambsdorff, 2011). However, Lambsdorff (2011) argues that intermediaries play a central role in arranging illegal transactions and spreading corruption. Based on case studies he argues that intermediaries do the dirty work for firms. In other words, the intermediary makes payments to clear goods at customs or to comply with regulatory or licensing requirements (Hasker and Okten, 2008; Lambsdorff, 2011). Drugov et al., 2011, also confirm this argument by using experimental data. They find that intermediaries, besides reducing uncertainty, increase corruption by reducing the moral or psychological costs of both public officials and private agents. Firms that are exporting indirectly are less exposed to corruption because they have less contact with regulations and public officials because the intermediary takes care of this process. The expectation is that intermediaries function as mediators that link the two parties involved in a corrupt deal (Hasker and Okten, 2008; Drugov et al., 2011; Lambsdorff, 2011). It is expected that corruption does affect the amount of exports in a firm exporting through an intermediary positively. In other words, corruption has a more negative impact in firms exporting directly. Therefore:

Hypothesis 3 (H3): The effects of corruption on exports are more prevalent for direct

exporters than indirect exporters.

According to Almeida et al. (2008), governance efficiency is the key element to firms’ export intensity. Governance efficiency includes stability of government, efficiency of public institutions, the extent of regulatory activity, the effective rule of law and the level of corruption and accountability. Additionally, they argue that higher export performance is observed in bigger sized companies, older companies and companies that are foreign owned. Spencer and Gomez (2011) argue that a positive relationship exists between the host country’s corruption environment and the pressure subsidiaries face to engage in bribery locally. MNE subsidiaries face conflicting pressures in various national environments. On one hand, there is pressure from the national environment to engage in bribery, on the other hand, from the parent firm to comply with global anti-corruption standards (Rodriguez et al., 2005; Spencer and Gomez, 2011).

(13)

of firms owned by foreigners is partially determined by their need to defend themselves against threats from governments and the economic opportunities these governments create (Rodriguez et al., 2005). The expectation is that firms with a certain percentage of foreign ownership are less likely to engage in corruption due to afore mentioned pressure from the owners of the foreign market to comply with corruption practices. Foreign owned firms are more likely to comply with global anti-corruption measures due to the threat of reputational damage that could alter the firm’s overall performance. Firms that experience higher levels of corruption will try to reduce the risk of being exposed to corruption and will therefore decrease the amount of exports (Spencer and Gomez, 2011). Therefore, the following hypothesis is formulated:

Hypothesis 4 (H4): The effects of corruption on exports are more prevalent for

foreign owned firms.

Conceptual  framework  

The conceptual framework is depictured in Figure 1. This study will first examine the influence of bureaucracy and corruption. Second, the influence of corruption on export performance will be tested. Subsequently, the effects of corruption on exports in the case of direct exporting firms and of foreign owned firms will be examined.

(14)

Methodology  

Background  of  export  performance  and  corruption  in  Latin  America  

In the late 19th century, new Latin American democracies started to develop themselves independently. Latin America became a provider of raw materials for the rest of the world. The wars between the countries in Latin America and the creation of high interdependence with other countries led to difficulties in the development process. The years of American hegemony over the southern hemisphere, civil wars and revolutions mark the development in the 20th century (Gardini and Lambert, 2011). Still, Latin America did modernize in terms of industrialization, modern transportation and infrastructure. In the post-World War II period, import-substitution industrialization led to a shift toward manufacturing for domestic consumption and closing national economies. Governments in many Latin American countries held closed nationalistic policies that eventually led to capital flight and foreign exchange shortages. The debt crisis in the 1980s resulted in an export-oriented industrialization strategy. Subsequently, the introduction of trade reforms, tax reforms, financial liberalization and privatization made Latin American markets more attractive for foreign investors (Bértola and Ocampo, 2010; Bulmer-Thomas, 2010).

(15)

This creates opportunities for public officials to involve themselves in extortion in order to exploit the land and its natural resources. Other factors contributing to corruption in Latin America are low education and poverty. These create opportunities for public officials to compensate for low incomes. Additionally, social norms can impact attitudes toward bribery and corruption and thereby creating a barrier to enforcing anti-corruption policies (Charoensukmongkol and Sexton, 2011; Rehman and Perry, 2013).

Table 1. Exports per Country between 2005 and 2014.

2014 2013 2012 2011 2010 2009 2008 2007 2006 2005 Argentina -8.13 -3.97 -5.64 5.63 14.02 -9.40 1.09 7.88 5.50 12.54 Bolivia NA* NA NA NA NA NA NA NA NA NA Brazil NA 2.53 0.48 4.49 11.52 -9.12 0.55 6.20 5.04 9.33 Chile 0.66 3.37 0.11 5.52 2.32 -4.55 -0.70 7.19 5.10 2.83 Colombia -1.69 5.34 5.99 11.75 1.26 -2.85 4.48 6.91 8.60 5.71 Ecuador 6.16 2.35 4.70 5.67 -0.24 -4.79 2.98 0.02 7.13 8.63 Mexico NA 1.16 5.94 8.22 20.55 -11.78 -1.35 3.65 7.67 5.71 Paraguay 0.87 18.45 -6.72 6.23 19.89 -8.22 0.87 9.27 2.95 11.53 Peru -6.89 3.15 -0.26 6.55 0.75 0.84 3.03 6.83 3.04 13.11 Uruguay 1.90 0.18 3.10 5.80 7.18 4.45 8.53 4.78 5.62 15.99 Venezuela NA -6.17 1.59 4.66 -12.88 -13.68 -0.98 -7.55 -3.02 3.77 Costa Rica -1.66 3.64 9.26 5.52 5.54 -6.02 -2.01 9.91 10.28 12.79 El Salvador -0.95 4.83 -7.33 9.28 11.61 -15.97 6.86 7.08 5.88 0.90 Guatemala 7.55 6.66 1.75 3.01 6.13 -2.66 -0.16 9.42 4.79 -2.11 Honduras -1.33 1.63 9.85 8.36 15.68 -15.90 0.86 2.54 1.55 5.31 Nicaragua NA NA NA 5.95 14.10 3.48 11.95 8.33 12.49 7.67 Panama NA NA 10.15 22.83 1.61 -1.00 17.84 22.00 11.05 11.32

Note. Values are constant prices, year on year growth, in percentages. *NA = not available. Adapted from http://www.iadb.org/lmw. Copyright 2015 by

(16)

Table 2. Latin America Top 5 Main Exports to the World between 1981 and 2010. 1981-1990 1991-2000 2001-2010 2010 1 Petroleum and products Petroleum and products Petroleum and products Petroleum and products

2 Coffee, cocoa, spices tea, Electrical machinery Electrical machinery Electrical machinery 3 Non ferrous metals Transport equipment Transport equipment Transport equipment

4 Machinery, other than electric Machinery, other than electric Machinery, other than electric

Metalliferous ores and metal scrap

5 Iron and steel

Non ferrous metals

Metalliferous ores and metal scrap

Machinery, other than electric

Note. From: “Latin American Commodity Export Concentration: Is There a China

Effect?” by Fung, K. C., Garcia-Herrero, A., & Nigrinis Ospina, M., 2013, BBVA

Bank, Economic Research Department, p. 14.

Table 3. Country Score Corruption Perceptions Index between 2005 and 2014.

2014 2013 2012 2011 2010 2009 2008 2007 2006 2005 Argentina 34 34 35 30 29 29 29 29 29 28 Bolivia 35 34 34 28 28 27 30 29 27 25 Brazil 43 42 43 38 37 37 35 35 33 37 Chile 73 71 72 72 72 67 69 70 73 73 Colombia 37 36 36 34 35 37 38 38 39 40 Costa Rica 54 53 54 48 53 53 51 50 41 42 Ecuador 33 35 32 27 25 22 20 21 23 25 El Salvador 39 38 38 34 36 34 39 40 40 42 Guatemala 32 39 33 27 32 34 31 28 26 25 Honduras 29 26 28 26 24 25 26 25 25 26 Mexico 35 34 34 30 31 33 36 35 33 35 Nicaragua 28 28 29 25 25 25 25 26 26 26 Panama 37 35 38 33 36 34 34 32 31 35 Paraguay 24 24 25 22 22 21 24 24 26 21 Peru 38 38 38 34 35 37 36 35 33 35 Uruguay 73 73 72 70 69 67 69 67 64 59 Venezuela 19 20 19 19 20 19 19 20 23 23

Note. Adapted from http://www.transparency.org/cpi. Copyright 2015 by

(17)

Data  

The data to test the hypotheses is obtained from the World Bank’s Enterprise Surveys. The Enterprise Survey is a firm level survey of a representative sample of an economy’s private sector that covers business environment topics, such as finance access, corruption and performance measures. Top managers and business owners are interviewed face-to-face. The number of interviews depends on the size of the economy’s population but minimum sample sizes are respectively 150, 360 and 1200-1800 for small-sized, medium-sized and large-sized countries. In 2006, the first round of interviews in Latin America took place. During this round 10,930 firms were interviewed in 15 different countries. During the second round in 2010, 12,855 firms were interviewed in 30 different countries in Latin America and the Caribbean. The data in the surveys refers to the characteristics of the firm in the last fiscal year 2005 or 2009. In Brazil the interviews were conducted in 2009, but referred to the fiscal year 2008. An overview of all Latin American countries can be found in Appendix A.

(18)

Variables  

Dependent  variables  

The dependent variable that measures export performance is export intensity. Export

intensity measures how much a firm was exporting in 2009. This is measured as a

percentage of the total sales that is exported either directly or through an intermediary. The indicators are described by the World Bank (2014), as a proportion of the total sales that are exported directly in percentages and a proportion of the total sales that are exported indirectly in percentages. These two indicators combined form the export intensity of a firm. The precise questions that the respondent had to answer was: ‘in fiscal year [insert last complete fiscal year], what percentage of this

establishment’s sales were: national sales; indirect exports (sold domestically to third party that exports products); direct exports?’ (World Bank, 2014). From this the

indicator for export intensity is derived. Higher values of export intensity indicate that there is more export.

Independent  variables  

The variable corruption expectation measures corruption, more specifically, it is the expectation firms have to hand over a gift or informal payment. This variable is the independent variable in model 2.1 and will function as the dependent variable in model 1.1 as well. This variable is measured as whether or not firms expect that firms with characteristics similar to theirs, hand over a gift or informal payment in order to achieve the desired results. This data is also obtained from the World Bank Enterprise Surveys (2014). The indicator is described by the World Bank (2014), as ‘the

percentage of establishments that believe that firms with characteristics similar to theirs are making informal payments or granting gifts to public officials to “get things done” with regard to customs, taxes, licenses, regulations, services etc.’. The

question that respondents had to answer was: ‘it is said that establishments are

(19)

outcome in percentages but only uses two values, namely 0 and 100. Higher values of

corruption expectation indicate higher levels of corruption.

The independent variable bureaucracy that will be used in model 1.1, is measured as the percentage of senior management time spent on dealing with requirements of government regulations. This indicator is described by the World Bank (2014), as the average percentage of senior management’s in a typical week time that is spent dealing with requirements imposed by government regulations (e.g. taxes, customs, labor regulations, licensing and registration), including dealings with officials and completing forms. The question that respondents had to answer was: ‘In a typical

week over the last year, what percentage of total senior management's time was spent on dealing with requirements imposed by government regulations? [By senior management I mean managers, directors and officers above direct supervisors of production or sales workers. Some examples of government regulations are taxes, customs, labor regulations, licensing and registration, including dealings with officials and completing forms]’ (World Bank, 2014). Higher amounts of

management time spent on dealing with requirements of government regulations indicate higher levels of bureaucracy.

Control  variables  

Prior research indicates that export performance is influenced by a number of firm and country characteristics. Derived from this prior research, there will be controlled for firm size, firm age, gross domestic product (GDP), industry and country characteristics. Again, these variables are obtained from the Enterprise Surveys, except the GDP, which is obtained from the World Bank Development Indicators.

The first control variable firm size accounts for the size of the firm. Firm size is a composite measure of permanent and temporary workers. The number of temporary workers is adjusted by the average number of months worked in a year (World Bank, 2014). There are three levels namely, small (5-19 workers), medium (20-99 workers) and large (100+ workers). The second control variable firm age is a firm’s age measured in years. The respondents had to answer the question: ‘in what year did this

establishment begin operations?’ (World Bank, 2014). Based on their answer, the

(20)

of the survey. The third control variable is GDP, which is measured as percent annual growth in GDP. The fourth control variable industry is showing the different industries in which the companies are operating. The last control variable country accounts for country in which the firm is located.

Interaction  variables  

The variables direct exports and foreign ownership will be added in an interaction term.

The variable direct exports is measured as a percentage of the total sales that is exported directly. The indicator is described by the World Bank (2014), as a proportion of the total sales that are exported directly in percentages. The precise questions that the respondent had to answer was: ‘in fiscal year [insert last complete

fiscal year], what percentage of this establishment’s sales were: national sales; indirect exports (sold domestically to third party that exports products); direct exports?’ (World Bank, 2014). From this the variable direct exports is derived.

The variable foreign ownership accounts for a firm’s ownership structure, which is either domestic or foreign. A firm is considered to have foreign ownership if foreigners hold at least 10 percent of the ownership of the firm. The respondents had to answer the question: ‘what percent of this firm is owned by each of the following:

a. private domestic individuals, companies or organizations, b. private foreign individuals, companies or organizations, c. government/state and d. other’ (World

Bank, 2014).

Model  

The models derived from the conceptual framework simplified are as follows. Model 1.1 tests hypothesis 1.

𝑐𝑜𝑟𝑟𝑢𝑝𝑡𝑖𝑜𝑛  𝑒𝑥𝑝𝑒𝑐𝑡𝑎𝑡𝑖𝑜𝑛!!=

(21)

Whereby corruption expectation is a binary variable, a 0 means that a firm does not expect to grant a gift to an official and 1 if it does expect to grant a gift to an official in order to get things done. Bureaucracy is measured as the average percentage of senior management’s time in a typical week that is spent dealing with requirements imposed by government regulations. The control variable firm size is a dummy variable taking the value of 1 if a firm is smaller than 20 employees, 2 if a firm has between 20 and 99 employees and 3 if a firm has 100 employees or more. Firm age is measured in years. Furthermore, an industry dummy is added taking the values from 1 to 17 and a country dummy is added taking the values from 1 to 16. Lastly, 𝜀 is the error term.

Model 2.1 tests hypothesis 2, 3 and 4.

𝑒𝑥𝑝𝑜𝑟𝑡  𝑖𝑛𝑡𝑒𝑛𝑠𝑖𝑡𝑦!! = 𝛼

! +  𝛽1𝑐𝑜𝑟𝑟𝑢𝑝𝑡𝑖𝑜𝑛  𝑒𝑥𝑝𝑒𝑐𝑡𝑎𝑡𝑖𝑜𝑛!  +  𝛽2𝑓𝑖𝑟𝑚  𝑠𝑖𝑧𝑒!+

   𝛽3𝑓𝑖𝑟𝑚  𝑎𝑔𝑒!+  𝛽4𝐺𝐷𝑃!  +  𝛽5𝑖𝑛𝑑𝑢𝑠𝑡𝑟𝑦! +  𝛽6𝑐𝑜𝑢𝑛𝑡𝑟𝑦!+  𝜀! (2.1)

The export intensity is measured as a percentage of the total sales that is exported either directly or through an intermediary. Corruption expectation is a binary variable, a 0 means that a firm does not expect to grant a gift to an official and 1 if it does expect to grant a gift to an official in order to get things done. Firm size is a dummy variable taking the value of 1 if a firm is smaller than 20 employees, 2 if a firm is between 20 and 99 employees and equals 3 if a firm has 100 employees or more and

firm age is measured in years. Furthermore, industry is a dummy variable taking

values from 1 to 17. The variable country is also a dummy variable taking values from 1 to 16. Lastly, 𝜀 is the error term.

In order to test hypothesis 3, an additional interaction variable direct exports #

corruption expectation is added to the model, namely, direct exports, measured as a

percentage of the total sales that is export directly, multiplied by the variable

corruption expectation. For hypothesis 4 another interaction variable foreign ownership # corruption expectation is added, namely, foreign ownership, measured as

0 if domestically owned and 1 if foreign owned, multiplied by the variable corruption

(22)

Analysis  

(23)

Results  

Summary  statistics    

Table 4 summarizes the descriptive statistics and Table 5 the correlations of the variables used in this study.

Table 4. Descriptive Statistics

Variables Mean S.D. Min Max

1. Export intensity 9.93795 23.04283 0 100 2. Direct export intensity 7.185293 19.71892 0 100 3. Corruption expectation 0.1111111 0.3142866 0 1 4. Bureaucracy 15.47479 17.75484 0 100 5. Firm size 1.925926 0.7947439 1 3 6. Firm age 26.01422 19.78811 1 183 7. Foreign ownership 0.1228127 0.3282397 0 1 8. GDP -.6814707 2.543321 -4.7 4

The correlations in Table 5 show that corruption expectation and bureaucracy correlate negatively with export intensity and direct export intensity. Bureaucracy correlates positively with corruption expectation. Moreover, corruption expectation correlates negatively with all firm characteristics (firm size, firm age and foreign

ownership). All firm characteristics and GDP correlate positively with the export intensity.

Table 5. Correlation Matrix

Variables 1 2 3 4 5 6 7 8

1. Export intensity -

2. Direct export intensity 0.8396 -

3. Corruption expectation -0.0369 -0.0427 - 4. Bureaucracy -0.0081 -0.0195 0.1232 - 5. Firm size 0.2453 0.2526 -0.0435 0.0046 - 6. Firm age 0.0208 0.0400 -0.0377 0.0229 0.2471 - 7. Foreign ownership 0.2219 0.1986 -0.0077 0.0310 0.2580 0.0243 - 8. GDP 0.0424 0.0314 0.0384 0.0205 0.0076 0.0467 0.0250 -

(24)

The correlation matrix in Table 5 demonstrates that none of the independent variables are dependent on each other because none of the correlations is close to 1. Furthermore, in the Collin test, the rule of thumb is a mean VIF of 10. A higher score indicates multicollinearity. The test shows a mean of 1.05, thus indicating that multicollinearity is not a problem. Besides, none of the individual VIF values are close to 10. It is important to note that the condition number is 9.6, which is close to 10, which indicates instability. However, there is still no clear sign of multicollinearity. The Berja-Berque normality tests suggest that the variables are not normally distributed. The results are presented in Appendix C. However, as normality does not play a major role, the variables are not transformed to logarithms.

Model  estimation  

The results of the regressions are presented in Table 6 and Table 7. Table 6 presents the results of the probit estimations for model 1.1, which tests hypothesis 1. Table 7 shows the results of the tobit estimations for model 2.1, which tests hypothesis 2, 3 and 4.

Table 6. Probit Estimation of Corruption Expectation

Variables Model 1 Model 2

controls controls + bureaucracy

Firm age -0.000 -0.000 (0.00) (0.00) Firm size -0.086*** -0.085*** (0.02) (0.02) GDP -0.162*** -0.164*** (0.03) (0.03)

Industry dummies Yes Yes

Country dummies Yes Yes

Bureaucracy - 0.005*** (0.00) Constant -0.856*** -0.955*** (0.09) (0.10) AIC 5921.280 5637.756 BIC 6163.645 5885.460 Chi square 567.366 594.843 p(chi2) 0.000 0.000 Psuedo R-square 0.088 0.097 Number of observations 9213 8754

Note. Standard errors appear in parentheses.

(25)

Table 6 shows a significant positive relationship between bureaucracy and corruption

expectation. Column 1 shows the results when the control variables are added. The firm age is not statistically significant. Firm size and GDP have a negative coefficient.

Column 2 shows that if bureaucracy increases, the corruption expectation increases (β = 0.005, p < 0.001). This supports H1 that higher levels of bureaucracy result in higher levels of corruption. However, the coefficient is relatively small, which demonstrates that the influence on corruption of an increase in bureaucracy is small.

From the theoretical arguments presented in the previous chapters, it follows that corruption affects export intensity negatively. Furthermore, the effects of corruption on export intensity are more prevalent when firms are foreign owned or are exporting directly. In Table 7, Column 1 only the control variables are included. Column 2 includes the corruption expectation variable. Column 3 shows the results if the interaction term direct export # corruption expectation is added. Lastly, Column 4 shows the results if the interaction term foreign ownership # corruption expectation is included. The first Column of Table 7 shows that all control variables are significant at a 1 percent level. The age of a firm has a negative relationship with export intensity. As shown in Column 2, it is found that when corruption increases, firms export intensity increases (β = 0.034, statistically insignificant). This contradicts H2, which states that higher levels of corruption will influence export intensity negatively. However, the obtained result is not statistically significant.

Column 3 in Table 7 depicts the result that the interaction term direct export #

corruption expectation has a small positive effect on export intensity (β = 0.029,

statistically insignificant). This does not support H3, which states that the effects of corruption on export intensity will be more prevalent in firms that are exporting directly. Again, the result is not statistically significant. The firm age does not turn significant in this model.

Column 4 in Table 7, shows that the interaction term foreign ownership # corruption

expectation has a small effect on export intensity (β = 0.880, statistically

(26)

Table 7. Tobit Estimation of Export Intensity

Model 1 Model 2 Model 3 Model 4

controls controls + controls + controls +

Variables corruption expectation interaction direct exports interaction foreign ownership Firm age -0.104** -0.104** -0.013 -0.078* (0.04) (0.04) (0.02) (0.04) Firm size 29.361*** 29.362*** 9.377*** 25.373*** (1.01) (1.01) (0.55) (1.01) GDP 15.264*** 15.266*** 5.512*** 15.309*** (2.48) (2.48) (1.29) (2.47)

Industry dummies Yes Yes Yes Yes

Country dummies Yes Yes Yes Yes

Corruption expectation - 0.034 1.668 0.220

(2.38) (1.41) (2.52)

Direct exports - - 1.356*** -

(0.02)

Interaction direct exports # corruption expectation - - 0.029 -

(0.06)

Foreign ownership - - - 27.679***

(2.01)

Interaction foreign ownership # corruption expectation - - - 0.880

(6.43) Constant -63.215*** -63.223*** -32.005*** -59.628*** (3.61) (3.65) (1.97) (3.57) AIC 33649.726 33651.726 29361.261 33448.988 BIC 3.899.219 33908.347 29632.139 33719.866 Chi square 2051.663 2051.663 6346.127 2258.400 p(chi2) 0.000 0.000 0.000 0.000 Psuedo R-square 0.058 0.058 0.178 0.063 Number of Observations 9213 9213 9213 9213

(27)

Robustness  

In order to establish robustness additional tests are performed. First, the regressions will be tested on robustness by adding extra explanatory variables or replacing them by other variables.

The results for the probit estimation are presented in Appendix D and for the tobit estimation in Appendix E Column 1. These results of the probit and tobit estimations remain the same if GDP per capita is used as control variable instead of the GDP. It is important to note that in the probit model the coefficient for GDP per capita does turn positive and in the tobit model the coefficient for GDP per capita does turn negative compared to the GDP variable in the previous estimations.

Other explanatory variables for the corruption variable are added to test if the results are altered. When the corruption expectation variable is replaced by the corruption

obstacle variable the results remain the same for the control variables. The coefficient

of the corruption obstacle variable remains statistically insignificant. This result contradicts H2 that states that corruption will influence export intensity negatively. It is important to note because the corruption expectation coefficient suggests the influence of corruption on export intensity is positive. This thus confirms previous findings and the results are not altered. The variable corruption obstacle is the perception firms have of corruption. This is measured as the percentage of firms that indicate corruption as a major obstacle in current operations. The indicator is described by the World Bank (2014) as the percentage of establishments that consider corruption to be their biggest obstacle. The question that was asked was: ‘By looking

at card [CARD 25] can you tell me which of the elements of the business environment included in the list, if any, currently represents the biggest obstacle faced by this establishment? (CARD 25: 1. Access to finance; 2. Access to land; 3. Business licensing and permits; 4. Corruption; 5. Courts; 6. Crime, theft and disorder; 7. Customs and trade regulations; 8. Electricity; 9. Inadequately educated workforce; 10. Labor regulations; 11. Political instability; 12. Practices of competitors in the informal sector; 13. Tax administration; 14. Tax rates; 15. Transport)’ (World Bank,

(28)

The second variable added to the model in order to replace corruption expectation is

corruption value, which is the experience firms have with corruption. This is

measured as a percentage of contract value that needs to be handed over as a gift or informal payment expected to give in order to secure a government contract. This indicator is described by the World Bank (2014) as ‘the percentage of contract value

expected as a gift to secure government contract. Only firms that have confirmed that they have secured or attempted to secure a government contract in the last 12 months were allowed to answer this question’ (World Bank, 2014). The following question

was asked: ‘When establishments like this one do business with the government, what

percent of the contract value would be typically paid in informal payments or gifts to secure the contract?’ (World Bank, 2014). After this variable is added to the

estimation, the number of observations decreases substantially to 1747 observations. This is due to the large number of missing observations for this variable. Moreover, Column 3 of Appendix E shows the results that suggest that the variable corruption

value alters the results of the model in overall. The variable firm age turns statistically

insignificant and the coefficient positive. Additionally, the coefficient for corruption

value is small and negative but statistically insignificant. It is important to note that

this contrary to result of the coefficient of corruption expectation, and would thus confirm H2 that states that corruption will influence export intensity negatively. However, this variable is not strong due to the substantial decrease in the number of observations.

The results presented in Column 1 Appendix F show the test results when an ordinary least square regression is performed with export intensity as the dependent variable. In this case the dependent variable is not treated as censored data. The results show that this regression the coefficient for corruption expectation becomes negative, which means that when corruption increases, the export intensity decreases. The variable is statistically significant at a 10 percent level (β = -1.439, p < 0.10). The control variables firm age and firm size are statistically significant. Only GDP is statistically insignificant in contrast to the tobit estimation. Furthermore, the coefficients of the control variables firm size (β = 7.371, p < 0.001), firm age (β = -0.066, p < 0.001) and

(29)
(30)

Conclusion  

Discussion  

The aim of this study was to identify the impact of corruption on a firm’s export performance. Prior research has shown opposing results. This study followed the line of reasoning that corruption has a negative impact on export performance. It has argued that the effects of corruption are more prevalent in the case of foreign owned and direct exporter firms.

This study presents contradicting results to the line of reasoning followed in this research. The results do not support the hypotheses put forth in the literature review, except in the case of bureaucracy, where the expectation was that bureaucracy results in more corruption. This expectation is supported by the empirical results. Contrary to what was expected, the findings of this study suggest that when corruption increases, the firm’s export performance increases as well. Although this does not support the hypotheses build in this paper, the results are not unexpected if the other line of reasoning in the theory is followed. This theory argues that in high-tariff environments corruption can have a positive effect because it allows exporters to circumvent tariff barriers. Corruption could lower the price for exporters and provides opportunities for exporters to evade the large numbers of controls and authorizations required in their operations. Furthermore, exporters that bribe might be able to speed up custom processes. In this way firms will save the costs regarding inspections, regulations and delays. Therefore, this study still adds value to the academic literature that already exists. However, it is important to note that the empirical evidence is not conclusive because the corruption variable does not turn statistically significant.

(31)

thus suggests that higher amounts of exports can have a positive effect on fighting corruption.

Implications  

This study examined the effect of corruption on export performance based on firm-level data. After analyzing the sample of Latin American countries in fiscal year 2009, the results show that corruption has a positive effect on export performance. In other words, firms that are expecting corruption in the domestic market have a higher export performance. The findings presented in this paper do not confirm the hypotheses built in accordance with the literature, but this study has still provided insights into the impact of corruption on export performance.

Bureaucracy does influence corruption and could be seen as a fuel for corruption. The most striking finding is that, contradictory to the proposed hypotheses; corruption has a positive effect on a firm’s level of export. This means that corruption improves a firm’s export performance. However, the empirical evidence is not conclusive.

Moreover, the results of this study have implications for managers of exporting firms. As the results suggest that higher levels of corruption will lead to increased export, firms might be willing to engage in corruption in order to improve export performance. However, it has to be taken into account that corruption has its costs. These are higher than the amount paid as a bribe alone. A bad reputation caused by suspicion or allegations of corruption might lead to extra costs. Furthermore, corruption leads to less transparent business operations due to its secretive nature. This study did not look at the impact of corruption on overall firm performance. Managers have to be aware of the complications that are associated with corruption before they choose to improve their export performance in this way.

(32)

operational practices. Corruption by its nature makes processes less transparent and more complicated. Thus, it is debatable whether corruption will improve export or firm performance in general and consequently the development of the economy on the long run.

Limitations  and  future  research  

This research has certain limitations that provide opportunities for future research. First, the data used in this research is survey data. Therefore, bias may be present in responses to the questions, especially in those about corruption. Corruption is a sensitive subject and therefore will remain sensitive to discuss for many interviewees in the future. However, more sophisticated surveys, focusing on one firm, may gain additional information about corruption.

Second, this study only takes into account the impact of domestic market corruption on export performance. However, foreign market corruption could also affect export performance. Zelekha and Sharabi (2012), argue that the level of corruption rather correlates to the trading partner’s absolute level of corruption instead of the home country’s level of corruption. The data used in this research does not provide insight on export destinations. Hence, foreign corruption levels are not taken into account. Future research could further explore the possibilities to account for both domestic and foreign corruption.

(33)

Fourth, the sample consists of Latin American countries. It might be interesting to see if the same results will be obtained in other regions in the world. Nevertheless, this is beyond the scope of this research and leaves possibilities for future research.

Fifth, the robustness checks show evidence of reverse causality. Future research could explore this in more detail based on more theoretical background to support the causality.

(34)

Bibliography    

Almeida, J. P., Borges, M. T., Cabral, J., & Ferreira, F. J. (2008). The Propensity And Intensity Of Export Activities of Companies Operating In Europa. The Journal of

Applied Business Research, Vol. 24, Nº 4.

Anderson, J. E., & Marcouiller, D. (2000). Insecurity and the Pattern of Trade: An Empirical Investigation. Revision of NBER Working Paper #7000.

Asiedu, E., & Freeman, J. (2009). The Effect of Corruption on Investment Growth: Evidence from Firms in Latin America, Sub-Saharan Africa, and Transition Countries. Review of Development Economics, Vol. 13, Nº 2, pp. 200-214.

Bannon, I. (1999). The Fight Against Corruption, A World Bank Perspective. Central

America Country Management Unit, Latin American and the Caribbean Region.

Stockholm. Retrieved from:

http://www.iadb.org/regions/re2/consultative_group/groups/transparency_workshop6. htm

Bértola, L., & Ocampo, J. (2010). Desarrollo, vaivenes y desigualdad. Una historia

económica de América Latina desde la independencia. Madrid: Secretaria-General

Iberoamericana.

Bulmer-Thomas, V. (2010). La historia económica de América Latina desde la

independencia. Mexico: Fondo de Cultura Económica.

Carr, I. M., & Outhwaite, O. (2009). Investigating the impact of anti-corruption strategies on international business: an interim report. Retrieved from: surrey.ac.uk.

(35)

Costas, R. (2015, February 5). Petrobras scandal casts long shadow over Brazil. BBC. Retrieved from http://www.bbc.com/news/world-latin-america-31155317

De Groot, H. L., Linders, G. J., Rietveld, P., & Subramanian, U. (2004). The institutional determinants of bilateral trade patterns. Kyklos, Vol. 57 Nº 1, pp. 103-123.

De Jong, E., & Bogmans, C. (2011). Does corruption discourage international trade?

European Journal of Political Economy. Vol. 27, Nº 2, pp. 385-398.

De Toni, A., & Nassimbeni, G. (2001). The export propensity of small firms: A comparison of organisational and operational management levers in exporting and non-exporting units. International Journal of Entrepreneurial Behaviour & Research, Vol. 7, Nº 4, pp 132-147.

Dutt, P., & Traca, D. (2010). Corruption and Bilateral Trade Flows: Extortion or Evasion? The Review of Economics and Statistics, Vol. 92, Nº 4, pp 843-860.

Drugov, M., Hamman, J., & Serra, D. (2011). Intermediaries in Corruption: An Experiment.

Estrin, S., Meyer, K. E., Wright, M., & Foliano, F. (2008). Export Propensity and Intensity of Subsidiaries in Emerging Economies. International Business Review, Vol. 17, Nº 5, pp. 574-586.

Fung, K. C., Garcia-Herrero, A., & Nigrinis Ospina, M. (2013). Latin American Commodity Export Concentration: Is There a China Effect? BBVA Bank, Economic

Research Department, Working Paper, Nº 1306. Retrieved from:

https://www.bbvaresearch.com/KETD/fbin/mult/WP_1306_tcm348-370671.pdf?ts=1752013

(36)

International Business Studies, Vol. 43 Nº 8, pp 693-718. Retrieved from:

http://core.ac.uk/download/pdf/11304700.pdf

Gardini, G. L., & Lambert, P. (2011). Latin America foreign policies: between

ideology and pragmatism. New York: Palgave Macmillan.

Gaviria, A. (2002). Assessing the Effects of Corruption and Crime on Firm Performance: Evidence from Latin America. Fedesarrollo, Colombia, Bogota.

Hasker, K., & Okten, C. (2008). Intermediaries and corruption. Journal of Economic

Behavior & Organization, Vol. 67, pp. 103-115.

Hessels, J., & Terjesen, S. (2010). Resource dependency and institutional theory perspectives on direct and indirect export choices. Small Business Economics, Vol. 34, Nº 2, pp 203-220.

Hill, R., Griffiths, W. E., & Lim, G. C. (2011). Principles of Econometrics, 4th editition. John Willey & Sons Inc.

Huang, R. R. (2007). Distance and trade: Disentangling unfamiliarity effects and transport cost effects. European Economic Review, Vol. 51, Nº 1, pp. 161-181.

Inter-American Development Bank. (2014). Latin Macro Watch Database. Retrieved from: http://www.iadb.org/lmw

Kimuyu, P. (2006). Corruption, Firm Growth and Export Propensity in Kenya.

Institute for Development Studies, University of Nairobi, Working Paper, Nº 54.

Knack, S., & Azfar, O. (2002). Trade intensity, country size and corruption.

Economics of Governance, Vol. 4, p. 1-18.

Lalountas, D. A., Manolas, G.A., & Vavouras, I. S. (2011). Corruption, globalization and development: How are these three phenomena related? Journal of Policy

(37)

Lambsdorff, J. G. (2011). Corrupt intermediaries in international business transactions: between make, buy and reform. European Journal of Law and

Economics, Vol. 35, Nº 3, pp. 349-366.

Lee, S. H., & Weng, D. H. (2013). Does bribery in the home country promote or dampen firm exports? Strategic Management Journal, Vol. 34, pp. 1472-1487.

Majocchi, A., Bacchiocchi, E., & Mayrhofer, U., (2005). Firm Size, Business Experience, and Export Intensity in SME: A Longitudinal Approach to Complex Relationships. International Business Review, Vol. 14, pp. 719-738.

Mauro, P. (1995). Corruption and growth. The quarterly journal of economics, Vol. 110, Nº 3, pp. 681-712.

McArthur, J., & Teal, F. (2002). Corruption and Firm Performance in Africa. Retrieved from: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.196.7084 &rep=rep1&type=pdf

Rodriguez, P., Uhlenbruck, K., & Eden, L. (2005). Government corruption and the entry strategies of multinationals. Academy of management review, Vol. 30, Nº 2, pp. 383-396.

Rehman, S. S., & Perry, F. V. (2013). Corruption, Constitutions and Crude in Latin America. Retrieved from: http://works.bepress.com/fredrick_perry/1

Søreide, T. (2014). Drivers of Corruption: A Brief Review. World Bank Studies. Washington, DC: World Bank.

(38)

Sterlacchini, A. (2001). The Determinants of Export Performance: A Firm-Level Study of Italian Manufacturing. Weltwirtschaftliches Archiv, Vol. 137, Nº 3, pp. 450-472, retrieved from: http://www.jstor.org/stable/40440853.

Tanzi, V. (1998). Corruption Around the World, Causes, Consequences, Scope, and Cures. Staff Papers-International Monetary Fund, Vol. 45, Nº 4, pp. 559-594.

The Economist (2015, February 7). Brazil’s oil industry, Pitfalls at Petrobras. The

Economist Newspaper Limited. Retrieved from http://www.economist.com/printedition/2015-02-07

The World Bank. (1997). Helping Countries Combat Corruption. The Role of the

World Bank. Washington, DC: World Bank.

The World Bank. (2014). Enterprise Surveys 2010. Washington, DC: World Bank Retrieved from: http://www.enterprisesurveys.org.

The World Bank. (2014). World Development Indicators 2014. Washington, DC: World Bank. Retrieved from: http://data.worldbank.org/data-catalog/world-development-indicators

Transparency International. (2015). Corruption Perceptions Index 2005-2014. Retrieved from: www.transparency.org/cpi

Tuaño, P., Manzano, G., & Villamil, I. (2014). Determinants of export intensity and propensity among small and medium-sized enterprises: The case of the Philippines.

ARTNeT, Working Paper, Nº 140. Retrieved from: http://hdl.handle.net/10419/103855.

Watts, J. (2015, February 4). Oil giant boss quits as Brazil corruption scandal puts pressure on president. The Guardian. Retrieved from http://www.theguardian.com

Zelekha, Y., & Sharabi, E. (2012). Corruption, institutions and trade. Economics of

(39)

Appendices  

Appendix  A.  Country  Overview  Latin  America  and  the  Caribbean  

Medium-sized countries: Bolivia, Costa Rica, Ecuador, El Salvador, Guatemala, Honduras, Nicaragua, Panama, Paraguay, and Uruguay.

Large countries: Argentina, Brazil, Chile, Colombia, Mexico, Peru, and Venezuela.

(40)

Appendix  C.  Results  Berja-­‐Berque  Normality  Test  

Appendix  D.  Robustness  Checks  -­‐  Probit  Estimation  

Variables Model 1 Model 2

Controls Controls + bureaucracy

Firm age -0.000 -0.000 (0.00) (0.00) Firm size -0.086*** -0.085*** (0.02) (0.02) GDP per capita 0.000*** 0.000*** (0.00) (0.000)

Industry dummies Yes Yes

Country dummies Yes Yes

Bureaucracy - 0.005*** (0.00) Constant -3.316*** -3.442*** (0.54) (0.55) AIC 5921.280 5637.756 BIC 6163.645 5885.460 Chi square 567.366 594.843 p(chi2) 0.000 0.000 Psuedo R-square 0.088 0.097 Number of observations 9213 8754

Note. Standard errors appear in parentheses.

+p < 0.1; *p < 0.05; **p <0.01 ***p<0.001 (two-tailed tests)

   

Variables Outcome Normal (yes/no)

1. Export intensity 27611.89 no

(41)

Appendix  E.  Robustness  Checks  -­‐  Tobit  Estimation  Export  Intensity  

Variables Model 1 Model 2 Model 3

Firm age -0.104** -0.103** 0.045 (0.04) (0.04) (0.05) Firm size 29.362*** 29.193*** 16.235*** (1.01) (1.02) (1.52) GDP - 15.215*** 11.717** (2.49) (3.63) GDP per capita -0.024*** - - (0.00)

Industry dummies Yes Yes Yes

Country dummies Yes Yes Yes

Corruption expectation 0.034 - - (2.38) Corruption obstacle - 0.273 - (1.49) Corruption value - - -0.008 (0.32) Constant 168.689*** -62.709*** -33.387*** (37.92) (3.74) (5.76) AIC 33651.726 33270.392 6999.911 BIC 33908.347 33526.514 7191.208 Chi square 2051.663 2015.140 362.281 p(chi2) 0.000 0.000 0.000 Psuedo R-square 0.058 0.057 0.050 Number of Observations 9213 9086 1747

Note. Stand errors appear in parentheses.

+p < 0.1; *p < 0.05; **p <0.01 ***p<0.001(two-tailed tests)

(42)

Appendix  F.  Robustness  Checks  –  OLS  Estimation  Export  Intensity  and  Probit   Estimation  Corruption  Expectation  

OLS Estimation Probit Estimation

Variables Export Intensity Corruption Expectation

Firm age -0.066*** -0.000 (0.01) (0.00) Firm size 7.371*** -0.074** (0.30) (0.02) GDP 0.097 -0.160*** (0.39) (0.03)

Industry dummies Yes Yes

Country dummies Yes Yes

Export intensity 0.034 -0.002* (2.38) (0.00) Corruption expectation -1.439+ (0.74) AIC 82829.079 5918.634 BIC 83078.572 6168.127 Chi square - 572.012 p(chi2) - 0.000 Psuedo R-square - 0.089 Number of Observations 9213 9213

Note. Stand errors appear in parentheses.

Referenties

GERELATEERDE DOCUMENTEN

This could nevertheless support H2 and indicate that the positive effect of CSR performance on the disclosure of firm-level corruption is particularly strong in

Multiple determinants for corrupt firm behaviour are considered, including: country-level corruption, length of applications, gender of the owner, gender of the

insignificant. With the Hosmer and Lemeshow goodness of fit test stating the model was a poor fit, the same dependent and independent variables should be studied again under a

The regression coefficients in Model 2 show a clear positive, significant relationship between institutional corruption and firm performance.. When adding the squared term in Model

Master Thesis – MSc BA Small Business &amp; Entrepreneurship.. University

A lack of trust does affect firm performance in the sense that it takes longer to gain trust when crossing subnational borders, and to gain trust companies need to put in more

Overall it can be concluded that there is a clear statistical negative at the 5 percent significant effect of corruption on the firm performance when the

(2003) only used a sample which covered data from large firms, this sample also covered small and medium sized firms. Therefore with the results of this research it is