How does firm size matter when dealing with political instability? The business perspective.

How does firm size matter when dealing with political instability? The business perspective.

Master Thesis International Business and Management Faculty of economics and Business

2

Abstract

This study provides an insight into the effect of recent events around the world concerning political instability. The focus lies on discovering the relation between firm characteristics, and the way the firms‟ managers perceive political instability as an obstacle to doing business. Answering the main research question; how does a

firm’s size influence the perceived obstacle of political instability? Using a

3

Contents

Chapter 1 – Introduction... 4

Chapter 2 - Literature review ... 7

Chapter 3 - Methodology... 13

Chapter 4 – Results ... 16

4.1 - Descriptive Statistics of the variables ... 16

4.2 Results and Discussion ... 23

General - Multiple regression analysis ... 24

Country and sector specific - Multiple Regression Analysis ... 29

Additional findings ... 33

Chapter 5 Conclusions and limitations ... 34

5.1 Conclusions ... 34

5.2 Recommendations ... 37

References ... 39

4

Chapter 1 – Introduction

Usually, when you think of political instability in Europe, less developed countries are the first to think of. However, with the recent referendums causing great stir all over Europe with Great Britain leaving the European union, the consequences of political instability are felt everywhere in Europe. To top that, with the United States of America electing its new president, trading relations with countries all over the world are getting more and more stressed.

Not only is it felt, over 4.400 firm managers interviewed in the latest round of The Business Environment and Enterprise Performance Survey (BEEPS), indicated political instability as a key obstacle to its businesses. All countries in the Southern and Eastern Mediterranean faced challenges due to the spread of extremism, spill overs, regional turmoil, and external shocks (World Bank, 2015). With more and more political instability spreading across Europe, my aim is to find out whether firm size matters in doing business when dealing with the issue of political stability, or the lack of it. The research found significant evidence that in countries such as Egypt and Tunisia, the number one obstacle firm managers perceived, is political instability. It‟s stated that over 4,400 randomly selected firm managers from a range of countries in the southern and eastern mediterranean responded to that BEEPS survey. My research uses the latest BEEPS dataset, with interviews taken from 15,883 enterprises in 30 countries of Eastern Europe and Central Asia. With such an evident conclusion, the question that comes to mind is how firm managers across the whole of Europe view political instability as an obstacle to doing business. More precisely, how does a firm’s size influence the

perceived obstacle of political instability? “It is believed that one of the crucial

5

One of these differences for instance, is that they may be less affected by excessive regulations because they can more easiliy slip into informal arrangements (Weder, 2001). Meaning small firms can operate under the radar and stick to it‟s usual routines, less affected by changes in the law for instance. Other differences will be dealt with in the literature section. This thesis draws on the BEEPS enterprise survey covering countries in Eastern Europe and Central Asia to study the question whether the business obstacle called political instability is related to firm size. Valuable data can be found when discovering the perception of firm managers about this obstacle in the country of business.

I research this by questioning the relation between a firm managers‟ perception of political instability as an obstacle, and firm size. Concerning firm size, the literature has little to say about how the state of a country‟s financial and legal institutions affects firms of different sizes. One research studies the determinants of firm size by looking at only the largest firms around the world (Beck, Demirgu¸ Kunt, and Maksimovic, 2001). Therefore, the literature concerning political instability as an obstacle to businesses seems to be in a developing phase, where smaller companies are still left out, even when researching whether firm‟s size plays a significant role in dealing with political instability.

Research on the relation of firm size and political instability is important because from the literature, a consensus about the relation between firm size and political instability cannot be found.

The main research question I will answer is; Do smaller firms see political

instability as a bigger obstacle to its operations than bigger firms?

It concerns the firms‟ perception of political instability forming an obstacle to its operations in a country. I expect the results to provide a final clarity on whether small firms have a significant disadvantage when doing business in times of political instability, or instead achieve advantages because of their size. Moreover I think that it can predict and prepare small firms for the consequences of political

6

business in Great Britain due to extreme fluctuations in currency rates, policy changes, and tensions with other countries. Besides that, I will research firm age, and firm ownership structure have a significant influence on the perceived obstacle of political instability. I expect older firms to see political instability as less of an obstacle due to its relations and experience in the country. Moreover, firms partially owned by state are expected to see political instability as less of an obstacle as well. In the literature section, firm age and firm ownership structure are continued.

7

Chapter 2 - Literature review

Firm Size

One of the biggest differences between small and medium scale enterprises (SME), and the big multinational enterprises (MNE), is the high sensitivity of SMEs to external challenges such as political instability. A highly unpredictable external context creates a difference in innovativeness between SMEs and MNEs (Schwens, Eiche, & Kabst, 2011). For a SME it is more difficult to deal effectively with the country risks in terms of political, technological and institutional differences (Laufs & Schwens, 2014). Moreover, De Soto (1987), found in his research that small

entrepreneurs faced huge obstacles in terms of entry costs when trying to obtain a license to formally register as a company. A last characteristic of smaller firms compared to bigger firms is the ownership structure which effects decision making at an non rational level. The process of managing a small firm is usually different than in a large firm. Management in a small firm is more likely to also fulfil the role of owner, or stake holder. Therefore it can cause poor decision making, followed by inappropriate actions. (Jennings, Beaver, 1997) As stated by Demsetz (1985), large corporate firms often have a highly diffused ownership structure that separates ownership and management. Therefore, looking at firm managers‟ perception of an obstacle when researching its relationship is appropriate. Firm ownership structure is presented in the paper as a control variable and explained in the literature

review.

8

Research done by Schiffer (2011) found that for political instability, there are no significant differences in the effects on firms on different sizes. So a consensus cannot be reached when studying previous literature. Smaller sized firms are likely to form pressure groups in order to bundle their influence. However, research (Olson, 1965) showed that groups containing a lot of members could prove to bring its own drawbacks if there is a free rider problem. Free riders are firms who profit from the pressure created by pressure groups representing them, without making costs. In short, this means that larger firms might be more successful in influencing politics and obtaining new rules in their favour, because they do not have to form a pressure group which at times can be inefficient. Thus gaining an advantage over smaller firms. Large firms might also be able to craft special deals with governments because of their power and their importance in the economy. For example, in a recession, they might threaten to lay off workers if they don‟t get tax reductions or for example influence a political decision. A clear example of this could be the U.S weapon industry, which threatens to lay off thousands of workers if the government cuts it‟s defence budget, keeping the congress in a chokehold.

Smaller firms‟ impact on an economy is less, therefore they have less influence in politics. Conversely there are several good arguments as to why larger firms may have more problems than smaller firms. For example informality; small firms can more easily slip into informal arrangements with governments, thereby avoiding taxes and regulations. Johnson, Kaufmann and Zoido-Lobaton (1998) have presented empirical evidence showing that a high level of corruption and weak institutions increases the size of the informal sector, hence which small firms can better adjust to, or even profit from.

9

rare, for the obvious reason that firm-level data are more difficult to assemble and too costly”, he found that costs are usually the biggest obstacle for researchers to perform such a detailed research. For political instability, firm level studies are lacking for the same reason. However, recent research by the World Bank made new firm level data available for public use (World Bank, 2014), paving the path for a whole different level of research on firm‟s performance and behaviour in relation to external threats.

Drawing upon the research in the field I have proposed the following hypotheses;

1a: Small firms see political instability as a bigger obstacle to its businesses than large firms do.

1b: Large firms see political instability as a bigger obstacle to its businesses than small firms do

I expect that small firms will see political instability as a bigger obstacle. Arguments that show that small firms suffer more than large firms are more common than arguments in the opposite direction, however it is not entirely clear which argument to vote for, therefore being a legitimate point of research. There are many reasons why small firms could be coping better than large firms. (Schiffer, 2001) In the methodology section I will explain how I will research this and what the numbers stand for.

To get a more detailed picture of the data and its meaning, I will not only look at the data as a whole, but the data will be analysed specifically for the biggest

countries in the dataset in order to find out if there are differences among countries in which the firms operate.

Differences in firm size may not be the only reason why firms may experience

10

Firm age is relevant since firms which have been active for a longer time have more experience in dealing with political instability in a country, that is, if there was

political instability in the past. Moreover, they have had time to time to build up a relationship with governments, and a reputation in the country. Therefore, older firms might experience political instability at a lower obstacle level than younger firms. An example of this is Shell‟s experience in Nigeria, a case study by Frynas (2010) showed that Shell adapted to political instability over time. Nowadays, it can even be seen as a competitive advantage. Shell managed to penetrate Nigeria‟s state structures, which is used to hedge political risk in Nigeria. Concluding that political instability is inherently harmful to business, but over time companies can adapt to political instability. However, evidence of a negative relationship between firm age and the severity of obstacles to doing business can be found for firms in formerly communist countries. Since the dataset contains firms in Eastern Europe and Central Asia this is particularly interesting. Fields (2011) writes “The Firms that were established before 1989- that is, firms from the communist era, are often heavily indebted and therefore might experience higher obstacle levels than firms that were launched in the post-communist era.” (Fields, 2011) One of these obstacles is political instability. Therefore, Hypothesis 2; The longer a firm is

active in a country, the smaller chance that they see political instability as an obstacle is formed.

Ownership structure I - Whether any government has a stake in the firm.

There are many reasons to believe that government participation in ownership has an influence on the level of obstacles for doing business. Firms partly controlled by government might be less exposed to regulations, or for example corruption than private firms. They might receive special treatment with regard to taxes and regulations, have easier access to infrastructure, be more satisfied with the

11

relevant when dealing with political instability. While it seems evident that firms with a government stake will report political instability as less of an obstacle to business, research is not conclusive. Dewenter (2001) and Boycko et al (1996), found that firms partly controlled by governments are less profitable than privatized firms. Kreuger (1990) argues that firms with a government stake might be

pressured into hiring workers connected to governments instead of the best

qualified workers. With firms in our sample being partially government owned, this causes great conflicts of interest.

The following hypothesis will be tested; Hypothesis 3; Firms with a government

stake in its ownership structure view political instability as a smaller obstacle.

Ownership Structure II - Whether foreign owners have a stake in the firm

Firms that are owned partly or fully by a foreign entity might find it more difficult to adapt to local customs and to the political system. Therefore, they might report higher obstacle levels. Moreover, because foreign-owned firms are likely to have higher import and export rates than the average firm, exchange rate obstacles could be worse for them than for others. As researched by Asidue (2006), political instability has a negative influence on foreign direct investment, thus indicating that a foreign stake in the firm might influence perception about political instability as an obstacle. But there are also arguments for a positive relationship between obstacles and foreign control. Multinationals may have very good relations with the

government and they may more easily and credibly threaten to exit and relocate. As seen in the Netherlands, where foreign multinationals account for 40% of the jobs in the private sector (Statistics Netherlands CBS, 2013). For a small country as the Netherlands these companies are crucial for its economy. Therefore,

12

Since previous research proved to have no evident results, we will take ownership structure into account in our model. Two hypotheses are formulated accordingly, Hypothesis 4a; Firms with a foreign ownership stake view political

instability as a bigger obstacle.

And hypothesis 4b; Firms that experience trade regulations as an obstacle to businesses view political instability as a bigger obstacle.

Financing

Beck (2006), researched the determinants of financing obstacles for firms. He found that there is a significant lower number of reports of older, larger, and foreign-owned firms on financing obstacles. Therefore he confirms that size, age, and ownership can be used to predict financing obstacles. Moreover, his research suggests that institutional development is the most important characteristic when explaining financing obstacles. This is interesting, because this indicates that financing obstacles not only relate to firm size, firm age, and firm ownership, but also to institutional development.

Moreover, firms in developing countries may face financing obstacles that prevent them from reaching their optimal size (Demirgüc¸-Kunt and Maksimovic, 1998). Thus, this obstacle could become even bigger when the level of political instability increases. Developing countries often tend to have a higher level of political

instability than developed countries (World Bank, 2012). Together with Beck‟s research, (2006) this leads to my belief that acces to financing could be an obstacle directly related to political instability. However, Bennet (1972) researched political Instability as a Determinant of Direct Foreign Investment. He found that political instability does not seem to discourage direct foreign investment.

13

Chapter 3 - Methodology

3.1 Procedure

This research will use data from the results of the Business Environment and Enterprise Performance Survey (BEEPS). It Concerns the fifth round of the BEEPS which covered 15,883 enterprises located in 30 countries. (Appendix I)

There are many reasons that BEEPS is selected as the source for our data. BEEPS is a highly professional organization, supported by the World Bank, which provides transparent research. To ensure reliable information, only formally registered

companies are eligible for an interview; there are no restrictions on a firms‟ age. In some larger economies such as Russia, Turkey and Ukraine, the survey design allows stratification by some of the sectors with the largest contribution to

employment and value added. Firms with 100% government/state ownership are, as opposed to previous BEEPS rounds, no longer eligible to participate in BEEPS. The world bank believes that this enhances the data quality because there is less state involvement in answering sensitive questions about the government. The survey covers firms in different sectors, including manufacturing, retail,

construction, and services sectors. Hence, data will be quantitative using an existing dataset of survey data.

14

Dependent variable – Perceived Political Instability

The dependent variable in this research is the extent to which firms see political instability as an obstacle, we will call it the obstacle of political instability. In the BEEPS survey all firms were asked to indicate to what degree political instability is an obstacle to the current operations of its establishment. (Appendix I) Responses

ranged from no obstacle, to minor obstacle, to moderate obstacle, to major obstacle, to lastly very severe obstacle. The responses were given a score of 0 to 4, respectively. Respondents could also respond with „don‟t know‟, or „does

not apply‟ these responses were given scores of -9

and -7, in the regression analysis, entries with these values were not selected because they were not relevant in the research.

Now that we know how we computed our dependent variable, it is clear that these scores reflect the way a firm sees the degree to which political instability forms an obstacle to the firm.

Independent variable – Firm Size

Drawing upon previous literature, the main independent variable which effects the political instability obstacle is a firms size. In the BEEPS survey, firms were

screened on firm size and then ranked accordingly. (Appendix I) Micro firms are

accounted as firms with less than 5 employees. Small firms are accounted for when a firm had 5 to 19 employees, medium sized firms are ranked as firms with 20 to 99 firms, and large firms are firms with 100 or more employees. I will use the ranking for firms as used in the survey. Again, these

firms were given a score from 0 to 3, where micro firms score 0, and small, medium and large firms score 1, 2, and 3 respectively.

15

The questionnaire covers firm age, in question B.5 (Appendix I) interviewees were asked in what year the establishment begin its operations. Using this data we can derive the firm age from it. As explained in the literature section we will use firm age as a control variable in this research. Chapter 4 provides a summary of the data collected for the firm age variable.

Control variable II A & B - Firm ownership structure

Next, firm ownership is covered. In the questionnaire (Appendix I), every interviewee is asked whether there is a stake in the firm by domestic parties, foreign parties, or government parties. To ensure data integrity, in case there is a 100% government or state stake in the firm, the interview was aborted. Chapter 4 provides a summary of the data collected for this variable. For this data, I will use two nominal variables. First, is there a foreign stake in the firm, yes or no.

Secondly, is there a government or state stake in the firm, yes or no. This way the data will be less complicated and it makes for better testing.

Control variable III – Access to finance as an obstacle

Firm managers were asked whether they thought access to finance as an obstacle. Responses ranged from no obstacle, to minor obstacle, to moderate obstacle, to major obstacle, to lastly very severe obstacle. The responses were given a score of 0 to 4, respectively. Appendix I shows the questions asked with the answer

possibilities.

Control Variable IV – Trade regulations as an obstacle

The literature points out that whether firm managers see trade regulations as an obstacle are closely related to firm ownership structure, thereby correlating with our dependent variable. Interviewees were asked whether trade regulations and

16

Chapter 4 – Results

4.1 - Descriptive Statistics of the variables

In this chapter, the data will be explained and descriptive tables will give a clear overview on the data of the 15.883 firms interviewed.

First, we have the descriptive of the BEEPS data. It shows that the mean of firm size is 1,5564, this means that the average firm in our sample is between medium and small sized. Political obstacle scores a mean of 1,2887, meaning that the average firm in our dataset sees political instability as some kind of obstacle. Non responses and invalid answers are left out, hence there are 15474 observations.

Overall Descriptive Statistics

Mean Std. Deviation Observations

Political Obstacle 1,2887 1,40176 15474

Size 1,5564 0,74393 15474

Table 1

Before testing the data, we will look at the data more specifically. Filtering the data into country sets will show if there are the same findings when looking at countries individually. Using our three biggest countries, Russia, Ukraine and Turkey, the biggest datasets will be available, still providing reliable results. First, looking at the descriptive statistics shows a very interesting fact. We see that the mean score for the political instability obstacle in the Ukraine is 1,6267. Whereas Russia and Turkey have surprisingly lower mean scores. This points out that the severity problem itself differs among countries.

Country Specific Descriptive Statistics

Political Obstacle Mean Std. Deviation Observations

Russia 1,2974 1,34883 4220

Ukraine 1,6267 1,35815 1002

Turkey 1,1094 1,36298 1344

17

Second, the countries represented by the data are shown. As table 3 on the next page shows, the data is spread evenly, with the biggest countries in the dataset, Russia (4220), Turkey (1344), and Ukraine (1002) being represented by the most companies. Separately using the biggest country datasets first, will tell if there is a difference between countries.

18

Country Descriptives

19

Our main variable is explained in the obstacle table, it shows that there are 7014 firms seeing political instability no obstacle, accounting for 44.2%, and 53.2% of the firms see political instability as an obstacle. It is interesting to see that 9.2% of the interviewees see political instability as a very severe obstacle, thus crucial for doing business. Moreover, 2.6% of the responses were not valid. Again, the data is spread evenly across the options. As the table shows, the data is scaled from 0 (No Obstacle), to 4 (Very Severe Obstacle).

Political Obstacle Severity

Frequency Percent Invalid response 405 2,6 No obstacle 0 7014 44,2 Minor obstacle 1 2114 13,3 Moderate obstacle 2 2675 16,8 Major obstacle 3 2207 13,9

Very severe obstacle 4 1464 9,2

Total 15884 100

20

The independent variable firm size is spread with 49,2% of the firms accounting as a small firm. Medium sized firms take 35,5% of the total responses and there are 1884 large firms, 11,9% of the total firms. Micro firms are, with 3,5% the smallest group, but still represent 549 firms. This means that we have sufficient data in each size category to test with.

Firm Size Distribution

Size Frequency Percent

Large 1884 11,9 Medium 5631 35,5 Small 7819 49,2 Micro 549 3,5 Total 15884 100,0 Table 6 Firm Age

Respondents were asked in what year the establishment start its operations,

deducting that from 2016 will give us firm age of that establisment. In the table, the statistics for firm age are shown. With the 160 non-responses left out, firm age has a minimum of 3 years, and the eldest establishment is 178 years old. The mean age is 18 years (17,82 rounded up).

Firm Age

Valid 15724 Missing 160 Mean Firm Age 17,82

21

As explained in the literature section, to test the influence of political instability on a firm, it is clear that partial state ownership influences this. Therefore the following table shows the descriptive statistics, I have modified the data into a nominal

variable. Moreover, the data covers whether there is a foreign interest in the firm. The nominal variable shows the following.

Ownership Structure

Partial State owned Partial Foreign Owned Percentage

Invalid response 163 162 1 1 No (0) 15441 14632 97,2 92,2 Yes (1) 279 1089 1,8 6,8 Total 15883 15883 100 100 Table 8 Acces to Finance

Table 9 shows the descriptives of the fourth variable, a firms‟ access to finance. It shows that 45,7 percent of all firms report that access to finance is not an obstacle, whereas 5,6 percent of the 15883 firms see financing as a very severe obstacle. This indicates that there is a wide spread across the our sample which will provide reliable results when testing the data in our regression analysis

Finance Obstacle

Frequency Percent Cumulative Percent

Invalid 313 2,0 2,0

No obstacle (0) 7254 45,7 47,6

Minor obstacle (1) 2692 16,9 64,6

Moderate obstacle (2) 2782 17,5 82,1

Major obstacle (3) 1906 12,0 94,1

Very severe obstacle (4) 936 5,9 100,0

Total 15883 100,0 100,0

22

Table 10 shows whether firms see trade regulations as an obstacle to business, as explained in the literature this is relevant when testing our results. We see that 64,1 percent of firms see trade regulations as no obstacle, whereas roughly 25 percent of the firms see trade regulations as an obstacle, varying from a minor o bstacle to very severe obstacle.

Trade Regulations Obstacle

Frequency Percent Cumulative Percent

Invalid 1703 9,5 9,5

No obstacle (0) 10176 64,1 73,6

Minor obstacle (1) 1578 9,9 83,5

Moderate obstacle (2) 1424 9,0 92,5

Major obstacle (3) 779 4,9 97,4

Very severe obstacle (4) 414 2,6 100,0

Total 15883 100,0

23

4.2 Results and Discussion

Before analyzing the regression results, a test is done to check whether there is collinearity among variables. In other words, it is a check for the phenomenon in which two or more predictor variables in a multiple regression model are highly correlated, meaning that one can be linearly predicted from the others with a substantial degree of accuracy. The higher the variance inflation factors (VIF), the more variance of the estimated regression coefficients are inflated as compared to when the predictor variables are not linearly related. Researchers tend to keep a score of 5 (Rogerson, 2001) or 4 (Pan & Jackson, 2008) as a maximum when

deciding to remove a variable from the research. All variables in the general dataset as a whole score lower than 2 so therefore I will keep all variables in the analysis. However, when analysing the statistics for our country specific, and sector specific results. It shows that the VIF score on Foreign Ownership and State Ownership in the Ukraine are 5,642 and 5,620. This indicates that in the Ukraine these variables have to be removed when selecting Ukraine since they express too much collinearity with the political instability as an obstacle.

Collinearity table

Total Country Sector

Variable Collinearity Statistics Ukraine Russia Turkey Manufacturing Service

Tolerance VIF VIF VIF VIF VIF VIF

Size 0,997 1,003 1,003 1,001 1,002 1,000 1,008

Foreign Ownership 0,514 1,946 5,642 2,332 3,614 2,081 1,844

State Ownership 0,516 1,94 5,620 2,328 3,613 2,074 1,839

Firm Age 0,999 1,001 1,001 1,001 1,002 1,001 1,002

Financing Obstacle 0,993 1,007 1,008 1,000 1,041 1,005 1,009

24

First we will test the hypotheses using a multiple regression analysis on the complete dataset. Including all countries and both sectors in the research, adding the control variables discussed in the literature will provide the best reliable results since the dataset is at its biggest. The Pearson-correlation test will provide us the option to test the correlations between variables. The correlations and the

coefficients (table 12) show that there is a 0,011 value on the Pearson correlation score when looking at political instability as an obstacle, and firm size respectively. This means that, by looking at this figure alone, a firm in a higher size category will likely have a 1,1% increase in the chance it sees political instability as an obstacle. This is an extremely small figure. When looking at the Model Summary (table 14), the significance level is 0,084. At the confidence interval of 1% we can say it is significant, however with such a small relationship it is a very small effect.

Translated in a more simple subject, for instance education and earnings, it would mean that the relation to your level of education and your earnings would be just a few cents. That brings to question why there is such a strong variance in the

literature about this topic. Where researchers could not find an evident conclusion, the results are also not evident. It seems that the pros and cons of different sized firms in this matter square out each other, hence there is no consensus in the literature. Moreover, when looking at table 13, the regression analysis shows that there is a 0,022 coefficient value between our firm size and political instability. This value however, is not significant. This tells that there is no significant relationship, with the coefficients value not being significant, I will dismiss both Hypothesis 1a; Small firms see political instability as a bigger obstacle to its businesses than large firms do. And Hypothesis 1b Large firms see political instability as a bigger obstacle to its businesses than small firms do.

25

Obstacle Political Size Ownership Foreign Ownership State Firm Age Obstacle Finance

Trade Regulations Obstacle Political Obstacle 1,000 Size ,011** 1,000 Foreign Ownership ,005 -,019*** 1,000 State Ownership -,014*** ,007 ,695*** 1,000 Firm Age -,004 -,010 ,019** ,003 1,000 Financing Obstacle ,016*** -,012* -,040*** -,040*** ,002 1,000 Trade Regulations Obstacle -,007 -,039*** ,047*** -,002 ,029*** ,069*** 1,000 * = <0.1 significance ** = <0.05 significance *** = <0.01 significance Table 12 Coefficientstable

Unstandardized Coefficients Standardized Coefficients

t Sig. B Std. Error Beta Size ,022 ,015 ,012 1,453 ,146 Foreign Ownership ,111*** ,040 ,031 2,747 ,006 State Ownership -,149*** ,047 -,035 -3,142 ,002 FirmAge ,000 ,001 -,004 -,512 ,609 Finance Obstacle ,013** ,006 ,017 2,074 ,038

Trade Regulations Obstacle -,005 ,004 -,009 -1,156 ,248

Dependent Variable: Political Instability Obstacle Table 13

Model Summary

R R Square Adjusted R Square Std. Error of the Estimate Durbin-Watson

Complete dataset ,034a _{,001*** 0,001 1,401 1,733} Ukraine ,044a _{0,002 -0,004 1,361 1,73} Russia ,041a _{0,002 0 1,349 1,991} Turkey ,068a _{0,005 0 1,363 1,982} Manufacturing ,031a _{0,001 0 1,399 1,777} Service ,041a _{0,002*** 0,001 1,403 1,726}

Predictors: TradeRegulationsObstacle, StateOwnership, FirmAge, Size, FinanceObstacle, ForeignOwnership Dependent Variable: PoliticalObstacle

26

When studying the literature, firm age seemed like a relevant control variable which could possibly influence our dependent variable, political instability.

Therefore, instead of running firm age as a control variable, I decided to test whether firm age as an independent variable shows a correlation to political instability. The results are shown in the correlations table (12) and in the

coefficients table (13). Looking at the data, again we see that there is a Pearson-correlation score of only -0,004 between firm age and political instability as an

obstacle to business. This indicates that there is very little relation between firm size and political instability as an obstacle. The significance level is 0,590. Moreover the coefficients table shows a 0,000 value.

Therefore, with the relationship being that small and non-significant, we can dismiss our hypothesis 2; The longer a firm is active in a country, the smaller

chance that they see political instability as an obstacle.

Firm Ownership Structure

Two other important variables included in the research concern firm ownership structure. The literature gave several hints that whether the firm has any part of government ownership, or foreign ownership could influence a managers view on political instability as an obstacle. When looking at the significance level of

government ownership related to political instability as an obstacle we see a significance level of 0,037. This means it passes our confidence interval, and the relation is therefore significant, be it very small with -0,014. There is a side note. In the descriptive statistics we see that there are only 279 firms state owned.

27

Looking at table 12, we see a 0,005 on the relation between political instability, and foreign ownership. This is a small value since the score can range from -1, to +1. The small figure tell us that these variables have a low relation. However, when looking at the coefficients table (13), a 0,111 value at the highest significance level shows that in fact there is a significant relationship with the severity of political instability as an obstacle. The positive value indicates that there is a higher change of foreign owned firms seeing political instability as an obstacle.

Due to the significant values in the coefficients table, it is evident that we can confirm Hypothesis 4; Firms with a foreign ownership stake view political instability as a bigger obstacle.

Trade Regulations Obstacle

Regarding trade regulations, the correlations table (12) shows very little relation with the dependent variable. Only a -0.007 value which is insignificant. Moreover, the coefficients table (13) shows insignificant values. Therefore hypothesis 4b; Firms that experience trade regulations as an obstacle to businesses view political instability as a bigger obstacle, can confidently be dismissed.

Access to finance Obstacle

The variable access to finance is added in order to find whether firms who lack access to finance, tend to see political instability as more of an obstacle. The table shows that there is a correlation of 0.016, at the highest significance level (<0.01). With this much significance, the low score cannot be ignored, looking at the

coefficients table, a value of 0,013 can be found at a significance level o f <0.05. With both values being significant, hypothesis 5; Firms that experience access to finance as an obstacle to business view political instability as a bigger obstacle, is confirmed.

28

big. A big sample size automatically decreases the R-squared values. This effect can be seen in the table by looking at the country, and sector specific values. Where the sample size decreases, R-squared values increase. However, the significance drops as we can see that only the R-squared value for the sector specific sample is

significant. Due to the big sample size, the model is very strong, yet it does bring a nuance to the results. Therefore we will look at the samples in more detail by

29

To find out if the results can be generalized, the research gets more complex. First, we have looked at the data in general. Now, filtering the data into country sets will show if there are the same findings when looking at countries individually. Using our three biggest countries, Russia, Ukraine and Turkey, the biggest samples will be tested, still providing reliable results. Moreover, the data is divided into two sectors, manufacturing and services. The results are discussed below.

Using selection variables to filter Ukraine, Russia and Turkey as the countries, and distinguishing between the manufacturing sector and the services sector. The results of the multiple regression analysis are shown in table 15 and 16. Since this research focuses on the effect of the independent variables on the obstacle of political instability, the first column provides valuable results. We see that in Russia, the correlation score between firm size and political instability as an obstacle is significant with a value of 0.034.

Turkey shows a significant value of 0.054, whereas in the Ukraine no significant correlation is found.

Moreover, in the manufacturing sector a significant correlation of 0.018 is found, whereas the services sector does not hold a significant correlation. This indicates that there are differences not only across countries, but also across different sectors. Therefore the results cannot be generalized. Further, we can see that

financing as an obstacle is a significant predictor of political instability as an obstacle in only the services sector. When looking at the model summary in table 14 we can see that the variables predict the dependent variable with and R-square of 0.002 in the services sector with significance. The other specific analyses did not provide significant R-square values.

Examining the coefficients table, we see that for Russia and Turkey there actually is a significant relation between our dependent variable political instability as an obstacle, and our main independent variable firm size. This is a particularly

30

relationship. Therefore we can say that hypothesis 1, which on grounds of the general analysis is dismissed, is open for discussion.

Furthermore we can clearly see that decreasing the sample size has led to less significant observations, since only in the services sector there are significant relations left. We see that ownership structure is significantly related. There is a 0.121 coefficient score on foreign ownership, which tells us that the more foreign ownership a firm has, the more likely it is the firm experiences political instability as an obstacle. Moreover, government ownership has a significant negative

31

Correlations Table Ukraine

32

Coefficients Table

Ukraine Russia Turkey Manufacturing Services

B B B B B Size -0,032 0,072** 0,102** 0,034 0,013 Foreign Ownership -0,137 0,097 -0,093 0,093 0,121** State Ownership 0,141 -0,055 -0,03 -0,13 -0,16** Firm Age -0,001 -0,002 0 0,001 -0,001 Finance Obstacle -0,021 0,005 0,008 0,001 0,021***

Trade Regulations Obstacle -0,003 0,003 0,001 -0,005 -0,005

Dependent Variable: Political Obstacle * = <0.1 significance

33

By doing a multiple regression analysis we can also find whether our independent variables are correlated to each other. The correlations table (12), shows many significant correlations. This is partly due to the big dataset, but we can see some interesting figures when looking at both the general data, and the correlations table (16) which contains country and sector specific results. It shows that size is

significantly correlated to political instability as an obstacle in Russia, Turkey, and in the manufacturing sector.

Furthermore we see that state ownership and foreign ownership are highly

correlated among both sectors, as well as Russia and the Ukraine. Whereas Turkey does not hold this correlation. This could indicate there are strict policies in turkey on firm ownership structures.

A quite obvious finding is that firms which are partly foreign owned, are likely to experience trade regulations as an obstacle. We can see that this holds among all countries and both sectors. Moreover, we see that financing as an obstacle is negatively related to foreign ownership, again this is what you would expect. Since foreign owned firms have easier access to finance as described shortly in the

literature section.

For state owned firms, the statement that they see access to finance as less of an obstacle does not hold true in every country. We see that in Russia and the Ukraine, no significance can be found, whereas in both sectors and in Turkey there is a

significant negative relationship.

34

Chapter 5 Conclusions and limitations

5.1 Conclusions

This research clearly has some evident conclusions, moreover it provides interesting insights into further research using hints derived from various correlations.

Starting with Hypothesis 1a: Small firms see political instability as a bigger obstacle to its businesses than large firms do, and 1b: Large firms see political instability as a bigger obstacle to its businesses than small firms do. The correlations significantly

state that large firms see political instability as a bigger obstacle to business than small firms do. However, when looking at country specific statistics, the results show that we cannot generalize the findings for every country. In the Ukraine for example, there is no relation between firm size and political instability as an obstacle. Whereas in Turkey and Russia there are significant relationships in the correlations table. More importantly, looking at the coefficients table, the sector specific results show that in the services sector the relationship between firm size and political instability is significant, as opposed to the manufacturing sector. On grounds of the general results both hypothesis 1a and hypothesis 1b are dismissed. Noting that the results vary across sectors, which qualifies the hypotheses to be tested even more specifically in further research.

Secondly, Hypothesis 2; The longer a firm is active in a country, the smaller chance

that they see political instability as an obstacle. Is dismissed because the

correlations are extremely low, let alone insignificant. The general, the country specific, and the sector specific results all show no significant correlation. Therefore I can conclude that there is no relation between firm age and our dependent

variable.

Thirdly, we researched whether government stake in a firm correlates with our dependent variable. The results show significant values over the general dataset. The country specific dataset does not show significant values, and the services sector shows a significant relationship. Therefore Hypothesis 3; Firms with a

35

is not a significant value within each sample.

Concerning foreign ownership stake in firms, I conclude there is a relation between foreign ownership and our dependent variable. Since the general, and the country and sector specific results show significant correlations, confirming Hypothesis 4;

Firms with a foreign ownership stake view political instability as a bigger obstacle, is

justified. Trade regulations seemed closely related to political instability, therefore the variable whether trade regulations were seen as an obstacle is analyzed. Results are however, insignificant. With -0.007 correlation and no significant results in the coefficients tables, we dismiss hypothesis 4b; Firms that experience trade

regulations as an obstacle to businesses view political instability as a bigger obstacle.

Whether firms that experience financing as an obstacle to businesses, would find political instability as an obstacle to business was researched as well. Over the whole dataset we can conclude that indeed there is a significant relationship between the two variables. Moreover the country and sector specific results show that in the services sector this relationship is most significant, whereas

manufacturing firms do not show a significant relationship. On grounds of the findings I have confirmed hypothesis 5; Firms that experience access to finance as

an obstacle to business view political instability as a bigger obstacle.

Like previous research, it does not clearly answer the question: how does firm size play a role in dealing with political instability. On grounds of my research, there is a small, but significant relation between a firm managers‟ perception of political

instability as an obstacle, and firm size.

36

when dealing with political instability cannot be confirmed, results show the opposite is true. This means that on grounds of this research, only supporting smaller firms when facing political instability as a government is not justified. The literature points out that there are advantages for being a smaller firm in times of political instability, but at the same time there are disadvantages. Hence, our results being small, but significant.

Finding significant correlations also held for other variables. We can say that firms with a government stake are likely to perceive political instability as less of an obstacle. Moreover, firms which experience financing obstacles are likely to see political instability also as a bigger obstacle. Results held true fo r the general

37

Since the analysis in general and the empirical model have been constructed as basic and as comprehensive as possible, there are some limitations, causing

suggestions for further research and improvements to the existing research I have done in this paper.

This quantitative approach showed small relationships between firm size and

political instability as an obstacle to a firms business in the correlations table. Since the model predicted a small part of the dependent variable, I strongly recommend not only looking at the factors I suspected of influencing the dependent variable like firm size, firm age, and ownership structure, access to finance, and trade

regulations as an obstacle. Using more control variables, creating framework of all external factors and it‟s relationships the detailed effect of the variables can be studied. Keeping countries and sectors separately showed different results, therefore the data should be defined even better into smaller parts. By doing so, this research might improve, these suggestions serve as a base for further research. The suggestions also point out the limits of this thesis. Although other variables are accounted for, there are many more variables which might correlate with the

dependent variable. For example the interviewees‟ heritage, age, experience and political views are not accounted for. This is a missed opportunity when asking for the interviewees personal views on political instability. Due to the dataset providing limited data, other aspects are, even though very interesting and helpful to give more detailed answers to our questions not researched.

Reflecting on my own research, the data is basic and in depth research might

provide clearer insights. Since the analysis of data is always bound to the limitations of the data sources , Interpretation of the BEEPS data is crucial and could be

38

39

Asiedu, E. (2006). Foreign Direct Investment in Africa: The Role of Natural

Resources, Market Size, Government Policy, Institutions and Political Instability. The

World Economy. 29 (1), p63–77.

Beck, T, et al.(2005). Financial and Legal Constraints to Growth: Does Firm Size Matter?. THE JOURNAL OF FINANCE. LX (1), p137-177.

Beck, T. (2006). Small and medium-size enterprises: Access to finance as a growth constraint. Journal of Banking & Finance. 39 (1), p2931–2943.

Boycko, Maxim; Shleifer, Andrei and Vishny, Rob- ert W. "A Theory of Privatization." Economic Journal, March 1996, 106(435), pp. 309-19.

Calof, J.L. (1994) The Relationship between firm size and export behavior revisited. Journal of International Business Studies, 25 (2) p. 367 - 387

De Soto, Hernando (1987). The Other Path: The Economic Answer to Terrorism. New York: Harper & Row Publishers. p271.

Demzets, H. Lehn, K.. (1985). The Structure of Corporate Ownership: Causes and Consequences. Journal of Political Economy. 93 (6), p1155.

Dewenter, L. Malatesta, H.. (2001). State-Owned and Privately Owned Firms: An Empirical Analysis of Profitability, Leverage, and Labor Intensity. The American

EconomicReview. 91 (1), p320-334.

EBRD-World Bank Business Environment and Enterprise Performance Survey (BEEPS). (2014). The Business Environment and Enterprise Performance

40

Fields, G (2013). Pathways Out of Poverty: Private Firms and Economic Mobility in

Developing Countries. London: Springer Science & Business Media. p218.

Huntington, S. 1968. Political Order in Changing Societies (New Haven, CT: Yale University Press,).

Jennings, P. Beaver, G.. (1997). The Performance and Competitive Advantage of Small Firms: A Management Perspective. International Small Business Journal. 15 (2)

Kotey, B. (2005) Are performance differences between family and non-family SMEs uniform across all firm sizes? International Journal of Entrepeneurial Behaviour & Research, 11 (6), p. 394-340.

Krueger, Anne 0. "Government Failures in De- velopment." Journal of Economic Perspec- tives, Summer 1990, 4(3), pp. 9-23.

Laufs, K. Schwens, C. (2014). Foreign market entry mode choice of small and medium-sized enterprises: A systematic review and future research

agenda. International Business Review. 23 (6), p1109–1126.

Rugman, M and Verbeke, A. (2007). Liabilities of Regional Foreignness and the Use of Firm-Level versus Country-Level Data: A Response to Dunning et al.. Journal of International Business Studies. Vol. 38 (1), p200-205.

Schwens, C. Eiche, J and Kabst, R. (2011). The Moderating Impact of Informal Institutional Distance and Formal Institutional Risk on SME Entry Mode

Choice. Journal of Management Studies. 48 (2), p330–351.

Schumpeter, J. A. (1943). Capitalism, Socialism and Democracy. USA: George Allen & Unwin.

Schiffer, M. Weder, B. (2001). Firm Size and the Business Environment: World Wide

41

Weder, B. (2001). Firm Size and the Business Environment: Worldwide Survey Results. IFC Discussion Paper . 43 (1), p215-225.

Wei, Shang-Jin, “Corruption in Economic Transition and Development: Grease or Sand?” Manuscript. Brookings Institution (2001).

42

Appendix I