THE EFFECT OF THE PERSONALITY TRAITS OF POLITICAL PARTY LEADERS ON VOTER PREFERENCES A FIELD RESEARCH IN KOCAELI, TURKEY

Transcription

1 THE EFFECT OF THE PERSONALITY TRAITS OF POLITICAL PARTY LEADERS ON VOTER PREFERENCES A FIELD RESEARCH IN KOCAELI, TURKEY A. Hamdi ISLAMOGLU, Professor Kocaeli University Umit ALNIACIK, M.A. Gebze Institute of Technology Volkan OZBEK, M.A. Gebze Institute of Technology Abstract Because of the increasing competition in social and business life, consumers became the decision makers in every stage of the professional life. This led the understanding of Marketing to change and develop, and to be implemented in different areas. One of those areas is the political arena. Marketing techniques are hired to orientate voter behavior in Turkey as well as the whole world. The issue of image and personality traits of political party leaders is extremely important on Turkish voter behavior. In this research, which had been conducted in Kocaeli just before the November 3, 2002 general elections; it is examined that how voters perceived the personality traits of Turkish political party leaders and how their preferences were affected by this perception. Keywords Political Marketing, Voter Behavior, Political Party Leaders, Personality Traits, Leader Image 1. INTRODUCTION Enormous changes in technology, communication and social life, and increasing level of competition resulted people having more choices and also to be more careful about the environmental issues. Though, Marketing Philosophy became more important and this increasing importance is recognized by various organizations. These organizations attached more importance to Marketing, as they understood the correlation between Marketing Philosophy and the success. Various non-profit organizations in size and subject, as well as the private companies are using marketing philosophy and techniques with an increasing density day by day. Political parties are also counted in these groups of organizations. Political parties understand that they have to renew their understanding of functioning in a revolutionary way, in order to be the party in power and also to stay there; and they have to meet the wants, needs, desires and expectations of the population and the state in the highest possible level. The explicit use of marketing techniques in politics dates back to 1920 s in Britain (Wring, 1994). Following applications were seen in the United States of America and France (Bongrand, 1991). And today, marketing techniques are used in politics all over the world. Several authors have discussed the use of marketing by political parties for over 40 years (Glick, 1967; McGinnis, 1969; Nimmo, 1970; Shama, 1973; Kotler, 1975; Kotler&Kotler, 1981; Mauser, 1983; Newman&Sheth, 1985; Smith&Saunders, 1990; Butler&Collins, 1994; O Cass, 1996).

2 1.1. POLITICAL MARKETING THEORY Political Marketing is something certainly different from just hunting the votes (Islamoglu, 2002). It is described in various ways. Political Marketing is one of the newest tools of political communication, which is needed because of the innovations in voting right, and changes in democracy and information technology (Bongrand, 1991). Political marketing is consist of a whole set of techniques that are used to make the party or the candidate favorable for the voters; to make the party or the candidate recognized by the most possible voter groups, and by every individuals in these groups; to put the party or the candidates differences forward; and to obtain the necessary amount of votes with the minimum expense to win the elections (Bongrand, 1991). Political marketing is the activities executed by the political organizations, to realize the necessary programs and services that the citizens need and to obtain support and organizational recognition (Kotler, 1972). A more complementary description of the Political Marketing can be made as follows: Political Marketing is a social system and a process for developing political programs to meet the needs, wants, desires and expectations of voters, the population and the state by using marketing techniques; and revealing candidates who has ability to implement those programs, promoting those candidates to voters; and also persuading the voters to vote for these parties and candidates by using marketing strategies. There have been many arguments considered if the marketing techniques could be used in the political arena. Former resources focused the Political Marketing theory mostly on campaigning, propaganda, advertising and establishing the communication between the voters and the parties, party leaders or the candidates (Franklin, 1992; Jones, 1995; Kavanagh, 1995; Maarek, 1995; McNair, 1995; Newman, 1994; Rees, 1992; Scammell, 1995). Some of the resources focused on lobbying activities (Dubs, 1979; Ellis, 1998; Miller, 1990; Wilson, 1993). Authors who approach from the view of marketing philosophy argued that marketing techniques could be used in nonprofit organizations (Kotler and Anderson, 1991; Levelock and Weinberg, 1984; Butler and Collins, 1994; Smith and Saunder, 1990). Those, who have not adequately adopted the marketing philosophy claim that, applying the marketing techniques into the politics is an ethical problem, and thus they oppose to this approach (Islamoglu, 2002). Table 1. Elements of Modern Marketing Approach Products and Services (Mainstream Marketing) 1. Consumer Based Approach a. A General Definition of Needs b. Defining the Target Groups c. Differentiated Goods d. Distinct Superiority Strategy e. Consumer Research 2. Consumer and Community Satisfaction 3. Integrated Marketing 4. Systematic Planning Politics (Political Marketing) 1. Citizen Based Approach a. Needs of Citizens b. Defining the Target Voter Groups c. Differentiated Political Products d. Distinct Strategies Among Competitors e. Research on Voter Behavior 2. Voter Satisfaction 3. Integrated Marketing 4. Systematic Planning Source: A. Hamdi Islamoglu, Siyaset Pazarlamasi Toplam Kalite Yaklasimi, 2.Ed., Beta Basim Yayim Dagitim A.S., Istanbul, 2002, s.24. 2

3 In Political Marketing, consumers are replaced with the voters, the state, population and the citizens. Though, the subject of political marketing is developing and applying policies in order to respond the wants, needs, desires and expectations of people. According to this approach, we can compare the mainstream marketing and the political marketing in terms of the elements of modern marketing philosophy, as shown in Table 1. (Islamoglu, 2002). Andrew Lock and Phil Harris by referring to Wring 1994, say that the explicit use of marketing techniques in politics dates back through 1920 s. Today, Marketers believe that there is a direct transference of marketing concepts and tools to political arena (Lock and Harris, 1996). This is true to an extent. But there are some essential differences between using these techniques in mainstream (product and services) marketing and in the political arena (Lock and Harris, 1996; Egan, 1999; Kotler, 1991). The first difference is seen in terms of the objectives. Commercial organizations main aim is to make profit, while political parties aim to be the party in power. A second important difference is the time of the purchase. In political marketing, since voting is considered as purchase ; all voters make their choices on the same day. Another difference is that money has no effect on voter preferences. On the contrary to mainstream marketing, there is a collective action of selection in political marketing. Even though it is not his/her choice, the voter may have to accept the decision of the majority. On the other hand, in political marketing, the elected party may act as a monopoly until the next elections. Except coalition cases, the other parties which are not elected, have no power on managing the country; at most it may have a pressure on the leading party by opposing in different ways. When compared to mainstream marketing, it is much more difficult for a new brand to enter the market and to grow up in the political arena. (But it is not impossible ) When we look from the view of promotion, we again see some differences. For example, there can be negative advertising in political marketing, although it is forbidden in many countries to make negative advertising about products and services in the mainstream market. In mainstream marketing, the marketer defines and manages the media coverage of the product; but in political marketing, it is independent from the marketer or the product itself. Since it takes a heavy public attention; it is a main issue the media widely uses. Local campaigns are very important in political marketing and are much widely used comparing to mainstream marketing. When compared in terms of marketing mix; there are some differences and some similarities as well. The product of political marketing is the party, party program, candidate or the party leader. The price of these products, which the voters buy by their votes, may match with the tax rates, social programs and benefits, per capita GNP and average salaries. State or province offices or even the district and neighborhood offices of the political parties can be seen as the distribution channels of political marketing. Promotion activities are mostly consisted of election campaigns and propagandas; but not limited by only these. As it is not seen in the mainstream marketing, political party leaders may join TV (or radio) discussions together with their rivals, and try to influence voters choices by winning the discussion. These discussions may also help voters to increase their knowledge about the leaders and their programs. For this reason, TV and radio discussions are important parts of the campaigns (Denton and Woodward, 1990). Concepts such as market, selling and profit are also changed in political marketing (Bongrand, 1991). There is not a real market in political marketing. Instead, there are adult citizens who have a common sense, and cares about their future, playing the act of buying by voting; they are the voters. Selling in mainstream marketing matches transferring goods 3

4 or giving services to another party for an appropriate price. But, this term does not take any place in political marketing. Selling means an immoral action in political marketing. Instead of selling, political marketing tries to promote the ideas and the personality of the candidate, to build up reliance and to increase the field of acceptance. The profit of political marketing is winning the election, becoming the party in power and having the candidate elected. 2. RESEARCH SUBJECT and BACKGROUND Similar to mainstream marketing, voters as consumers are the focal points of political marketing activities. Buying and voting decision are quite similar behaviors (Newman and Sheth, 1987:5): A politician offers a service just as a lawyer or a doctor. A voter pays these services by taxes. A voter has specific motives, which make certain candidates more attractive than others. A voter seeks out information about a candidate, just as a consumer seeks information about goods and services. Finally, a voter experiences either satisfaction or dissatisfaction after the election, just as a consumer does after a purchase. It is essential to make a good analysis of the wants, needs, desires and expectations of voters in order to get the majority of the votes and to be the party in power. Here is an important point. A commercial organization may achieve its goals even by a small group of customers, but it is impossible for a political party to be the party in power with only a small group of voters. For this reason, political party administrators must carefully investigate and work on the voter groups who do not have the same ideological view with them, too. Party administrators have to try to persuade the highest number of voters. It means, they must be able to diffuse into other market sections. A political marketer who wants to control the voter behavior and wants to persuade the highest number of voters to vote for his party, must initially define a basic voter group which is consistent to party s philosophy and must determine how the voters perceive the political product (Islamoglu, 2002). After making this definition, he must develop correct programs and services in order to satisfy the voters. But, the motives that drive voters to vote for a political party is certainly different from the motives those drive a consumer to buy a personal computer. For this reason, the formation of voter behavior and factors effecting this formation are investigated in political marketing (Islamoglu, 2002). Voting as a sub-action of political behavior is formed in political culture and political socialization as it is affected by demographic (age, income, occupation, gender, education) and psychological factors. Since the first studies at Columbia University in 1940 s, factors and processes which plays role on voter preferences are tried to be explained by various models. Being able to explain voter preferences and make predictions on voter behavior is the reason which stands behind these studies. Most frequent elements of such models are party identification, issue positions, voter group membership, retrospective votes, situational contingency, and candidate/leader personality (Newman and Sheth, 1987; Boiney and Paletz, 1991). The key point in these models is that one or a combined set of these elements forms the voter preferences. 4

5 The term of party identification is defined as a psychological loyalty through a certain party (Uztug, 1999). Starting multi-party democracy in Turkey dates back to In 1950 general elections, voters decided between two different political parties for the first time: Republican People s Party and Democrat Party. This situation turned into a basic polarization and even it became a tradition in Turkish political life (Kardam and Tuzun, 1998). Most of the Turkish voters vote for the political party who is representing the right polar at the time of elections. Party identification, even though it is a central factor effecting the voting decision, has a declining importance (Boiney and Paletz, 1991). This is also valid for Turkey. Rational measures mean beneficial expectations of voters such as economical, social and foreign policy that can be obtained when the party comes into power (Islamoglu, 2002). Before 2002 general elections, since he knew the unemployment issue was considered very important by the voters, Erdogan, the leader of Justice and Development Party (AKP) was promising to start up a road construction project throughout the whole country, which would create new job opportunities for thousands of people. Similarly, Mesut Yilmaz, the leader of Motherland Party (ANAP) indicating the risk of a devastating earthquake that might damage Istanbul, was promising a large-scale reconstruction and retrofitting project in Istanbul. He argued that this project would create lots of job opportunities while helping to decrease the possible earthquake damage. Since the voter lives in different social groups, and is in interaction with these groups; his political preferences is formed up, improved and even changed by these groups. The level and qualification of the feeling of belonging to a certain racial, religious or economical group may also effect the voting decision (Uztug, 1999). A person who belongs to a certain group may adopt group s (or the group leader s) norms as well as its political choices. It is well known in Turkey that, some people who are the members of some certain Islamic groups adopt their life style as well as their political choices to these certain groups norms. Recent studies conclude that past votes also influence the individual s voting decision. Similarly, some temporary and unpredicted social and/or economical events may effect voting decision. These events may create opportunities for some parties, although it may cause threats for some others. For example, arresting the leader of a terrorist organization-pkk just before the 1999 general elections boosted Bulent Ecevit s Democratic Left Party (DSP). Similarly, deep economical crisis of 2001 and old Prime Minister Bulent Ecevit s sudden health problems in the first half of 2002, has an important role which must not be underestimated on Tayyip Erdogan s victory in 2002 elections. Another important factor on Erdogan s victory was that Turkish people was really tired of living with unstable coalitions for more than 15 years. No matter how strong or weak a country lives the democracy, political leaders are extremely important on voter preferences. The party leader is crucial aspect of the party product, even in a non-presidential system (Lees-Marshment and Rudd, 2003). In Turkey, their leaders identify even the political parties. When asked, voters usually declare their political preference by saying the name of the party leader, not the party. Leaders are the most visible indication of what the party stands for and what it is likely to deliver should voters elect it to government (Parchment and Rudd, 2003). The idea of candidate evaluation was first proposed by Campbell et al (Campbell Gurin and Miller, 1954). This study was restricted candidates personal attributes. By the mid 1950 s and early 1960 s the term of candidate image started to be used in articles to describe 5

6 candidate personality (Nimmo and Savage, 1976). According to Nimmo, images are creations of the media and are perceived by voters as artificial and invented. Various studies in the USA showed that party identification does not have an important role on voting decision while there are numerous studies concluding that candidate image has an important role on voter preferences. Shama determined that voter response was based on the image of the candidate. (Shama, 1975). In a review of election results from 1952 to 1964, Natchez found candidate image to be the best single predictor of voter behavior (Natchez, 1970). Miller, Wattenberg and Malanchuk, analyzing American elections data, found that perceiving the candidates is focused on candidate s personality traits rather than political issues (Miller, Wattenberg and Malanchuk, 1985). Glass (1985) claimed that in some cases more educated voters are more interested in the personality traits than others (Quoted: Uztug, 1999). Lau s research in 1986 shows the importance levels of different factors on voting behavior as follows: a) candidate/leader personal characteristics b) primary issues c) group relations d) party identification (Quoted: Yildiz, 2002). Boiney and Paletz s study in 1991 concludes similar results. Various studies after 1980 support these findings; candidates personality traits and in general their images play a central role on voting decision (Yildiz, 2002). Nimmo used the concept of image to bridge the gap between what political scientists know about electoral behavior and the notion of the voter as consumer, defining political image as the voter s subjective appraisal of the candidate (Nimmo, 1975). Shama determined that, voter response was based on the image of the candidate (Shama, 1975). Kelby and Miner were able to predict with an 87,7% accuracy how a voter would behave by evaluating the voter s likes and dislikes of a candidate s personality (Kelby and Miner, 1974). Fishbein predicted behavior from intention with 91,2% accuracy by measuring intention based on variables that included, among others, the candidate s personal characteristics (Fishbein and Ajzen, 1980). Although these findings are valid for the USA, recent research shows that Turkey is experiencing similar changes (Yildiz, 2002). Caha, Toprak and Dalmis investigated party members levels of concern about politics, levels of political participation, the reasons of preferring the party they vote, their approach to other parties and their channels of obtaining political information in their 1996 research in Kirikkale (a province in Central Anatolia Region of Turkey). They found that 21,3% of the respondents said they vote for the party because of its leader in their first choice (Quoted: Uztug, 1999). TUSES s research in 1990 s shows that there is/will be a rapid change from party to leader in terms of affecting voter behavior (Yildiz, 2002). Before April 1999 general elections, the results of Media and Elections research, which was conducted by Ankara University and Konrad Adenauer Association, also support this claim (Quoted: Yildiz, 2002). In researches about candidate personality, especially in USA based ones, it is found that trustworthiness, economical and political competence, intelligence, decisiveness, honesty, being open minded, sincereness, effectiveness and being hard working are important characteristics. In Turkey, the characteristics of the leaders of political arena can be examined in three main categories. First group includes personality traits, character, good speech style, honesty, being well educated, trustworthiness etc. Second group is consisting of the leader s political characteristics: patriotism, Kemalism (following the principles of Mustafa Kemal Ataturk), modernism etc. Third group is mostly consisted of issue positions such as unemployment, economic programs, health and education policies, membership of European Union etc. A new dimension as management which is consisted of managing the party, 6

7 organizing the team etc. can also be added to these groups (Oktar, 1996). A research conducted by Strateji Mori in February 1995 found the characteristics of a good leader in order of their importance levels as follows: honesty/trustworthiness, patriotism, humility, modernism/creativity, foreseeing the future and intelligence, experience in politics, claiming the religion, knowledge about economy, bringing solutions to problems of human rights, Kemalism (Quoted: Uztug, 1999). A field research conducted in Ankara (capital city) by Nermin Abadan Unat, Ahmet Taner Kislali and Dogu Ergil just before 1975 senate elections, found Bulent Ecevit s most admired characteristics in order as follows: proficiency in foreign affairs (Cyprus issue), humility (being a public man), trustworthiness, being honorable, intelligence, courage, proficiency in solving economical problems, patriotism and obtaining benefits for local people (Milliyet Newspaper, 10 October 1975). We tried to find out the most admired personality traits, which are expected to be carried by Turkish political party leaders. And investigated how existing political leaders are perceived by the voters regarding those personality traits. 3. RESEARCH OBJECTIVE, GOALS AND THE RESTRICTIONS The objective of this study is to find out the importance levels of certain personality traits which were expected to be carried by Turkish political leaders, to check whether this importance levels vary regarding voter demographics or not, and to investigate if Turkish political leaders are perceived same or different regarding those traits. The expected outcome of this study is to guide political leaders to find out which personality traits should be put in the front place, and where should they put the emphasis. It is believed that this would help leaders to create more effective strategies for their image and communication efforts. The study is limited with the sample population (the voters living in Kocaeli-a province in north-west part of Turkey, 100 km east of Istanbul) and by the time when the survey was conducted (May 2003). Although the research findings are valid for Kocaeli and can not be scientifically generalized throughout the whole Turkey, they give significant clues for Turkish voter behavior. Research data was collected from primary resources. Lack of an up to date and detailed list of the voters living in the region put difficulties on sampling procedure. It was also seen that the 10 point scale, which was used to get the respondent s evaluation about the political leaders personality traits, increased the total time of filling the survey which resulted uncompleted questionnaires in some cases. This problem is missed during the field test of the questionnaire. It should be better to use a 7-point scale instead. 4. METHODOLOGY 4.1. RESEARCH PROBLEM We tried to answer two main problems; 1. What are the main personality traits and their importance levels which are expected to be carried by political party leaders? 2. How Turkish political party leaders are perceived in terms of these personality traits? 7

8 4.2. RESEARCH METHODOLOGY Both primary and secondary data resources are used for gathering necessary data to answer the research problems. First, a literature review was done about voter behavior and past researches on the subject. After the literature review, a field research was also conducted in Kocaeli. The research data was collected by a questionnaire. The sample was randomly selected among the voters who were living in Kocaeli. The respondents filled the questionnaires by face-to-face interviews. Data collection (survey application) was done by 10 interviewers (7 male, 3 female) between April-May Collected data was analyzed by using SPSS 9.0. After the literature review for this study and investigating past researches, we found 10 different personality traits of political leaders, that the Turkish voters attach the most importance: knowledge, honesty, being consistent, humility, power, courage, reformism, democraticism, religionism, patriotism. In the first part of the survey questionnaire after the questions asking their demographics, respondents were asked to choose the most important 5 traits out of these 10, and to put them (selected 5) in order of importance levels. After that, it is analyzed whether the order varies or not regarding respondent demographics. In the second part of the questionnaire, respondents are asked to rank 8 political party leaders (6 of them were on their duty by the time of the field research: Recep Tayyip Erdogan-Justice and Development Party (AKP), Deniz Baykal-Republican People s Party (CHP), Bulent Ecevit-Democratic Left Party (DSP), Necmettin Erbakan-Felicity Party (SP) [In fact, Erbakan was the de facto leader of SP since he was proscribed from politics], Devlet Bahceli- Nationalist Action Party (MHP), Cem Uzan-Young Party (GP); and 2 of them were not on duty since they had withdrawn after 2002 elections: Tansu Ciller-Right Path Party (DYP), Mesut Yilmaz-Motherland Party (ANAP)) for each of those 10 personality traits separately, by using a 10 point scale (1 for the least, 10 for the most). Analysis of Variance (ANOVA) is used to test if there was a significant difference between the average points that the leaders get for each trait SAMPLING It was calculated that a sample of 609 randomly selected respondents should be representing Kocaeli, for this survey. We considered Which party did you vote in 2002 elections? as the key question for defining the sample size. We accepted that a 2,5 percent variance (in 95% confidence interval) for the votes that AKP received in Kocaeli (43%), was tolerable. (The variance of the population was estimated as 0,1). According to Kocaeli election council, there were voters in Kocaeli by the time of November 3rd 2002 elections of these people voted in the elections, and there were valid votes at the end. To get a more representative sample, we used a combined sampling method. To do so, first we stratified the population (the whole province) into 7 subdivisions (Kocaeli s districts: Derince, Gebze, Golcuk, Izmit, Kandira, Karamursel, Korfez). We planned to get the samples from each districts proportional to the actual numbers of voters in each district. We got the neighborhood (mahalle) lists of each district and than we clustered the districts into their neighborhoods. We calculated the #neighborhoods that would be included in the sample for each district. To find out which neighborhoods would be included in the sample for each district, we used the following protocol: 8

9 1. We defined coefficients for each neighborhood regarding their population. In each district, the smallest neighborhood, in terms of #actual voters, got the 1 coefficient. 2. The other neighborhoods got coefficients according to their #actual voters in proportion to the smallest neighborhood s. By this way, each neighborhood had a coefficient interval. 3. We had a cumulative sum of intervals for each district. (By adding up each neighborhood s coefficients together in the district) 4. By using a computer, we randomly selected the predefined# of neighborhoods in the interval of each district. After defining 42 neighborhood names by using this protocol, we got the household address lists from the headman (muhtar) of each neighborhood. Than we randomly selected 620 households from these lists and visited each address for the interview. The interviewers applied the questionnaire by face to face interviews, to the oldest member of each household, who voted in the latest election (November 3rd, 2002) and who accepted to participate the survey. Table 2 shows the distribution of #of actual voters in Kocaeli, and the voters in our sample by district names. Table 2. Distribution of Actual Voters in Kocaeli and Respondents by District Names DISTRICTS #VOTERS ACTUAL SAMPLED (November 3, 2002) # % # % Izmit , ,7 Gebze , ,0 Golcuk , ,8 Kandira ,2 27 4,4 Karamursel ,3 26 4,2 Derince ,1 53 8,5 Korfez ,0 52 8,4 Total , ,0 The first section of the survey questionnaire has 7 questions regarding respondent demographics. Table 3 shows the frequency tables of respondent demographics (age, gender, education level, income level, working status, marital status, and household size) Table 3. Respondent Demographics AGE Frequency % EDUCATION LEVEL Frequency % 25 or Less ,9 Illiterate 11 1, ,6 Primary&Secondary School , ,7 High School ,9 46 or More ,2 University and Post Graduate Total ,4 Total ,2 Missing 10 1,6 Missing 5 0,8 TOTAL TOTAL

10 Table 3. Continued WORKING STATUS Frequency % HOUSEHOLD SIZE Frequency % Working for Someone Else ,8 1-3 People Self Employed ,7 4-5 People Not Working (Student, Housewife, Retired, Unemployed) ,2 6 People and More 98 15,8 Total ,7 Total ,8 Missing 8 1,3 Missing 1 0,2 TOTAL TOTAL INCOME LEVEL (Monthly) Frequency % MARITUAL STATUS Frequency % Less Than $ ,3 Married ,3 $350 - $ Not Married ,2 $700 $ Widowed/Divorced 22 3,5 More Than $ ,3 Total Total ,6 GENDER Frequency % Missing 21 3,4 Male ,1 TOTAL Female ,9 Total The difference between the sample s distribution of age groups and that of actual voters (people who are older than 18 and who are settled in Kocaeli) is less than 10% (actual distribution according to 2002 census of the Turkish State Institute of Statistics (DIE) is as follows: 18-25: 26,3%; 26-35: 26,1%; 36-45: 20,6%; 46: 26,8% ) (DIE, 2002). The mode of household size of the sample indicates the 4-5 People category. The percentage of married respondents is closed to actual number (66,4%) (DIE, 2002). The percentage of male respondents (71%) is high, since the refusing the interview rate was high in female respondents. Mostly, male members of the households participated to the survey. Similarly, members with a university or a postgraduate degree were more likely to participate the survey than the others. This caused another anomaly in respondent demographics when compared to actual numbers HYPOTHESIS According to the model that is described above, we developed 2 hypotheses to solve the research problem. Ho1 = Voter groups with different demographics behave indifferent while they put certain personality traits of political leaders in order of importance levels Ho2 = Voters perceive political leaders indifferent regarding their certain personality traits. 5. DATA ANALYSIS AND TESTING THE HYPOTHESIS After completing the interviews answers were coded on papers, and those codes were transferred into computers. Research data is tabulated and analyzed by using SPSS 9.0. As the respondents chose 5 out of 10 personality traits, and put them in order of importance levels (#1 most important; #5 least important), we tabulated the frequencies of the order number for each trait. After finding the top five personality traits which are expected to be carried by 10

11 political leaders, we created cross tables for each of these 5 personality traits across respondent demographics in order to see whether the order varies or not according to the demographics of the respondents. By doing so, we tested our first hypothesis. In the next section, we analyzed the ratings of 8 political party leaders for given 10 specific personality traits. We calculated the average scores for each trait of each leader, and run analysis of variance to see if there were significant differences between the leaders in terms of these average scores. In other words, to test our second hypothesis. 6. RESULTS 6.1. THE ORDER OF THE IMPORTANCE LEVELS OF PERSONALITY TRAITS WHICH ARE EXPECTED TO BE CARRIED BY POLITICAL LEADERS Respondents were asked to choose 5 out of 10 given personality traits, and to put them in order of importance levels according to their own evaluation. #1 meant most important, and #5 meant least important. Table 4 shows the frequency distribution of the ratings given to those 10 personality traits marked by A,B,C,D,E,F,G,H,I,J (A= Knowledge, B= Honesty, C= Being Consistent, D=Humility, E= Power, F= Courage, G= Reformism, H=Democraticism, I=Religionism, J=Patriotism) Table4. The Order of the Importance Levels of Personality Traits Which Are Expected To Be Carried By Political Leaders FREQUENCIES ORDER Knowledge Honesty Being Consistent Humility Power Courage Reformism Democratici sm Religionism Patriotism I II III IV V The most frequent trait in each row is accepted as the row-trait. It is seen that the most important personality trait, which is expected to be carried by the political party leaders is having knowledge, since it is the most frequent trait in row 1 (219 respondents put this trait in the first place). The second most important one is honesty. Being innovative is the third one. Democraticism is the fourth most important trait. The fifth trait that the respondents attach importance is patriotism. We checked whether the sequence of the order is random or not by Wald-Wolfowitz Runs Test. After seeing the order is random, we created the following cross tables showing this order of importance against respondent demographics. 11

12 GENDER Table 5. The Order of the Importance Levels of Personality Traits, Which Are Expected To Be Carried By Political Leaders by Gender of Respondent GENDER ORDER MALE FEMALE 1. Knowledge Knowledge 2. Honesty Honesty 3. Patriotism Reformism 4. Democraticism Democraticism 5. Reformism Patriotism Male respondents put knowledge in the first row. Than comes honesty, patriotism, democraticism and reformism. Similarly, female respondents put knowledge and honesty in the first two rows. But different from male respondents, they put reformism in the third row. This shows that male and female voters behave different, while they put the personality traits in order of importance levels. Female respondents give more importance to reformism than the male respondents. Male respondents give more importance to patriotism than female respondents AGE GROUPS Table 6. The Order of the Importance Levels of Personality Traits, Which Are Expected To Be Carried By Political Leaders by Respondents Age Groups AGE GROUP 25 or Less and More ORDER 1. Knowledge Knowledge Knowledge Honesty 2. Honesty Honesty Honesty Knowledge 3. Reformism Democraticism Democraticism Patriotism 4. Democraticism Reformism Reformism Reformism 5. Patriotism Patriotism Patriotism Democraticism Respondents who are younger than 46 years of age, put knowledge in the first row and honesty in the second row. Respondents in 46 years or more age group puts honesty in the first row. This shows that respondents who are in different age groups behave different, while they put the personality traits in order of importance levels. The oldest respondents (46 years or more age group) give more importance to the trait of patriotism than all other respondents. On the other hand, the youngest respondent group (25 or less) considers the trait of reformism more important than the other respondents do EDUCATION LEVEL Table 7. The Order of the Importance Levels of Personality Traits, Which Are Expected To Be Carried By Political Leaders by Respondents Education Levels EDUCATION LEVEL Illiterate Primary & High School University & Post ORDER Secondary School Graduate 1. Honesty Knowledge Knowledge Honesty 2. Knowledge Honesty Honesty Knowledge 3. Patriotism Patriotism Democraticism Reformism 4. Democraticism Reformism Reformism Democraticism 5. Reformism Democraticism Patriotism Patriotism 12

13 As seen on Table 7, illiterate respondents and respondents with a university or a post graduate degree think that honesty is the most important personality trait for political leaders, while other respondents put knowledge in the first row. Illiterate respondents and respondents whose education level is primary or secondary school put patriotism in the third row. Respondents with a high school degree put democraticism in the third row, and other respondents who have university or postgraduate degree put reformism here. This shows that respondents with different education levels behave different, while they put the personality traits in order of importance levels. Respondents in the lowest and the highest education level groups attach the most importance to honesty for political leaders WORKING STATUS Table 8. The Order of the Importance Levels of Personality Traits, Which Are Expected To Be Carried By Political Leaders by Respondents Working Status WORKING STATUS Working for Self Employed Not Working ORDER Somebody Else 1. Knowledge Honesty Knowledge 2. Honesty Knowledge Honesty 3. Democraticism Patriotism Reformism 4. Reformism Reformism Democraticism 5. Patriotism Democraticism Patriotism Not working respondents, and respondents who works for somebody else puts knowledge in the first row. Self-employed respondents give the most importance to honesty. The third most important personality trait is different in all of the groups. We can say that respondents with different working status behave different, while they put the personality traits in order of importance levels MARITAL STATUS Table 9. The Order of the Importance Levels of Personality Traits, Which Are Expected To Be Carried By Political Leaders by Respondents Marital Status MARITUAL STATUS Married Not Married Divorced/Widowed ORDER 1. Knowledge Knowledge Knowledge 2. Honesty Honesty Honesty 3. Democraticism Reformism Patriotism 4. Reformism Democraticism Reformism 5. Patriotism Patriotism Democraticism The first two most important personality traits are the same for all of the respondents regarding their marital status. But the third most important personality trait is different for all of the respondent groups. Married respondents put democraticism in the third row, while not married respondents put reformism and the respondents who are divorced or widowed put patriotism. This shows that respondents with different marital status put the personality traits that they expect a political leader to carry, in different orders of importance levels. 13

14 HOUSEHOLD SIZE Table 10. The Order of the Importance Levels of Personality Traits, Which Are Expected To Be Carried By Political Leaders by Respondents Household Size MARITUAL STATUS or More ORDER 1. Honesty Knowledge Knowledge 2. Knowledge Honesty Honesty 3. Democraticism Democraticism Reformism 4. Reformism Reformism Patriotism 5. Patriotism Patriotism Democraticism As seen in Table 10, respondents who live in smaller households with 1-3 people put honesty in the first row, while others prefer knowledge as the most important personality trait for political party leaders. Therefore, it is seen that respondents with different household size behave different, while putting the personality traits in order of importance levels INCOME LEVEL Table11. The Order of the Importance Levels of Personality Traits, Which Are Expected To Be Carried By Political Leaders by Respondents Income Level INCOME LEVEL Less Than $350 ORDER $350-$700 $700-$1000 More Than $ Honesty Knowledge Honesty Honesty 2. Knowledge Honesty Knowledge Knowledge 3. Patriotism Reformism Democraticism Reformism 4. Democraticism Patriotism Reformism Democraticism 5. Reformism Democraticism Patriotism Patriotism Respondents living in middle class families with $350-$700 monthly income put knowledge in the first row, while others think that the most important personality trait should be honesty. Respondents living in families with different income levels, put the personality traits that they expect a political leader to carry, in different orders of importance levels. The Ho1 hypothesis is rejected. Respondents with different demographics (gender, age group, education level, working status, marital status, household size, family income level) put certain personality traits that they expect political party leaders to carry, in different orders. 14

15 6.2. COMPARING THE AVERAGE SCORES OF THE POLITICAL PARTY LEADERS FOR CERTAIN PERSONALITY TRAITS Table 12. The Average Scores That the Political Party Leaders Get for Ten Different Personality Traits Which Voters Expect Them to Carry N=620 Knowledge Honesty Being Consistent Humility Power Courage Reformism Democraticism Religionism Patriotism OVERALL MEAN (Leader) R.Tayyip Erdogan (AKP) 6,27 6,07 5,88 6,35 6,60 6,65 5,92 5,87 8,23 6,92 6,48 Deniz Baykal (CHP) 6,08 5,22 4,83 5,14 5,32 5,55 5,77 6,11 4,01 6,15 5,42 Bulent Ecevit (DSP) 5,15 5,23 4,30 5,10 3,47 3,97 3,91 4,85 3,49 5,49 4,50 Necmettin Erbakan (SP) 5,48 3,92 3,90 3,70 4,42 4,31 3,12 3,43 6,94 4,83 4,40 Devlet Bahceli (MHP) 5,48 5,64 5,13 5,50 5,33 5,18 4,65 4,92 5,73 6,96 5,45 Cem Uzan (GP) 4,67 3,58 3,98 4,03 5,32 5,65 5,54 4,83 3,99 4,83 4,64 Tansu Ciller (DYP/F) 4,91 3,29 3,33 3,66 3,72 4,11 4,08 4,15 3,55 4,31 3,91 Mesut Yilmaz (ANAP/ F) 4,22 3,12 3,13 3,19 3,42 3,42 3,50 3,73 3,38 3,97 3,51 OVERALL MEAN (Trait) 5,28 4,51 4,31 4,58 4,70 4,85 4,56 4,74 4,92 5,43 4,79 It is seen from the table that there are two personality traits, whose mean is more than 5 points: 1) Patriotism (5,43) had the highest overall average score among the traits. 2) Knowledge (5,28) had the second highest overall average score. All of the remaining personality traits got average scores less than 5,00 (mid-point). Being Consistent (4,31) has the least overall average score among all the traits. The second least overall average score is 4,51 for Honesty. The overall average score for all of the traits is 4,79 points. When we look across the leaders; it is seen that Recep Tayyip Erdogan, the leader of Justice and Development Party (AKP), the winner of 2002 elections and existing prime minister, has the highest overall average score (6,48). The leader of Nationalist Action Party (MHP), Devlet Bahceli follows him with an overall average of 5,45. This can be counted as evidence showing that the image of the party leader is not the only factor that shapes voter behavior (MHP could not pass the 10% national threshold). Deniz Baykal, leader of the main opposing party in the parliament, Republican People s Party (CHP) got an average score of 5,42. The overall average scores of remaining 5 leaders are less than the overall average score of 4,79. Mesut Yilmaz, former leader of Motherland Party (ANAP) got the least overall average score. Tansu Ciller, former leader of True Path Party (DYP) got the second least point. These two leaders withdrew after the elections. 15

16 Table 13. Analysis of Variance, the Average Scores That the Leaders Got For Ten Different Personality Traits ANOVA Table N=620 Sum of Squares Df Mean Square F Sig. Between Groups Combined 1803, ,645 29,59 0,000 Within Groups 37106, ,706 Knowledge Total 38910, Between Groups Combined 4968, ,764 91,4 0,000 Within Groups 32725, ,766 Honesty Total 37693, Between Groups Combined 3240, ,968 63,71 0,000 Being Within Groups ,267 Consistent Total 33806, Between Groups Combined 4472, ,928 85,2 0,000 Within Groups 31547, ,499 Humility Total 36019, Between Groups Combined 4822, ,93 95,01 0,000 Within Groups 30506, ,251 Power Total 35328, Between Groups Combined 4358, ,675 79,75 0,000 Within Groups 32760, ,808 Courage Total 37119, Between Groups Combined 4318, ,981 80,81 0,000 Within Groups 32028, ,635 Reformism Total 36347, Between Groups Combined 3350, ,712 61,14 0,000 Within Groups 32883, ,829 Democraticism Total 36234, Between Groups Combined 12817, , ,000 Within Groups 29850, ,097 Religionism Total 42668, Between Groups Combined 4894, ,226 73,43 0,000 Within Groups 40270, ,522 Patriotism Total 45164, As seen on Table 13; according to the one-way analysis of variance, in 95% confidence interval, there is significant difference among the leaders regarding the average scores they got for ten different personality traits. The Ho2 hypothesis is rejected. Respondents perceive the political party leaders different regarding their certain personality traits (knowledge, honesty, being consistent, humility, power, courage, reformism, democraticism, religionism, patriotism). 16

17 7. CONCLUSION This study explored the personality traits that the voters expect the political party leaders to carry. Respondents are asked to put a set of given personality traits in order of their importance levels. First five traits, according to the frequency distribution, are treated as five most important personality traits that the voters expect the leaders to carry. It is found that the most important personality trait is knowledge for the respondents. Honesty, reformism, democraticism and patriotism come after, regarding their importance levels defined by the respondents evaluation. We think that Turkish political party leaders may have benefits if they consider this finding when they are positioning themselves. It is found that respondents with different demographics (e.g. gender, age group, education level, working status, marital status, household size, family income level) sort the most important five personality traits different according to their importance levels. Political party leaders may also consider this finding while they try to position their image. They may benefit if they divide the respondents into sub groups according to their demographics, and develop image strategies by giving the necessary consideration to the trait that each group attach the most importance. Respondents are also asked to evaluate Turkish political leaders regarding their appearance in terms of personality traits. It is found that there is significant difference among the leaders regarding how they were perceived by the voters. The leaders may also benefit if they take this finding into consideration. By doing so, they can find out their strong and weak points in the eyes of the voters and can redefine the image and communication strategies. It is found that respondents gave the least scores to the traits of being consistent and honesty when they evaluate a given set of personality traits of Turkish political party leaders. This can be considered as the expression of the lack of confidence to the politicians. We think that being careful about this point may bring some advantages, too. But, all of these statements and suggestions are limited by the limitations of the research. Similar to mainstream marketing, the usage of marketing research is obviously beneficial for political arena. More research will give us more generalized statements. 17

18 REFERENCES BOINEY, John, David PALETZ. (1991). In Search of The Model: Political Science versus Political Advertising on Voter Decision Making, N.J.: Hillsade Erbaum Assc. BONGRAND, Michel (1991).Politikada Pazarlama, Cev: Fatos Ersoy, Iletisim Yayinlari, Istanbul. BUTLER, P. and N. COLLINS. (1994). Political Marketing: Structure and Process, European Journal of Marketing, Vol.28, No.1. CAMPBELL, Angus, Donald GURIN and Warren E. MILLER (1954).The Voter Decides, Row, Peterson and Company, New York. DENTON, Robert, Garry Woodward. (1990). Political Communication in America, Prager Series in Political Communication. EGAN, John. (1999). Political Marketing: Lessons From The Mainstream, Journal of Marketing Management, Vol.15. ERDER Necat, Sezgin TUZUN, Ahmet KARDAM, Filiz KARDAM. (1995).Turkiye de Siyasi Parti Secmenleri ve Sosyal Demokrasinin Toplumsal Tabani, Turkiye Sosyal, Ekonomik, Siyasal Arastirmalar Vakfi (TUSES) Yayinlari, Ankara. FISHBEIN, Martin and Icek AJZEN. (1980). Predicting and Understanding Voting in American Elections: Effects of External Variables in I. Ajzen and M. Fishbein (eds.), Understanding Attitudes and Predicting Social Behavior, Prentice-Hall, New Jersey. FRANKLIN, M. et al, (1992).Electoral Change: Responses to Evolving Social Attitudes in Western Countries, Cambridge University Press: Cambridge. GLICK, E. (1967).The New Methodology, American Institute for Political Communication, Washington. ISLAMOGLU, A.Hamdi (1996). Pazarlama Yonetimi ve Uygulamalari, Kocaeli,. ISLAMOGLU, A.Hamdi Siyaset.(2002). Pazarlamasi Toplam Kalite Yaklasimi, 2. Baski, Beta Basim Yayim Dagitim A.S., Istanbul. JONES, N. (1995). Soundbites and Spin Doctors: How Politicians Manipulate the Media And Vice Versa, Casell, London. KARDAM, Ahmet, Sezgin TUZUN. (1998). Turkiye de Siyasi Kutuplasmalar ve Secmen Davranislari, 1. Baski, Veri Arastirma, Ankara. KAVANAGH, Dennis. (1995). Election Campaigning the new marketing of politics, Oxford: Blackwell. KELBY, Stanley and Thad W. MINER. (1974). The Simple Act of Voting, American Political Science Review,