Journal of Applied Business and Economics

North American Business Press Atlanta - Seattle – South Florida - Toronto

Editors Dr. Adam Davidson Dr. William Johnson

Editor-In-Chief Dr. David Smith

NABP EDITORIAL ADVISORY BOARD

Dr. Nusrate Aziz - MULTIMEDIA UNIVERSITY, MALAYSIA Dr. Andy Bertsch - MINOT STATE UNIVERSITY Dr. Jacob Bikker - UTRECHT UNIVERSITY, NETHERLANDS Dr. Bill Bommer - CALIFORNIA STATE UNIVERSITY, FRESNO Dr. Michael Bond - UNIVERSITY OF ARIZONA Dr. Charles Butler - COLORADO STATE UNIVERSITY Dr. Jon Carrick - STETSON UNIVERSITY Dr. Min Carter – TROY UNIVERSITY Dr. Mondher Cherif - REIMS, FRANCE Dr. Daniel Condon - DOMINICAN UNIVERSITY, CHICAGO Dr. Bahram Dadgostar - LAKEHEAD UNIVERSITY, CANADA Dr. Deborah Erdos-Knapp - KENT STATE UNIVERSITY Dr. Bruce Forster - UNIVERSITY OF NEBRASKA, KEARNEY Dr. Nancy Furlow - MARYMOUNT UNIVERSITY Dr. Mark Gershon - TEMPLE UNIVERSITY Dr. Philippe Gregoire - UNIVERSITY OF LAVAL, CANADA Dr. Donald Grunewald - IONA COLLEGE Dr. Samanthala Hettihewa - UNIVERSITY OF BALLARAT, AUSTRALIA Dr. Russell Kashian - UNIVERSITY OF WISCONSIN, WHITEWATER Dr. Jeffrey Kennedy - PALM BEACH ATLANTIC UNIVERSITY Dr. Dean Koutramanis - UNIVERSITY OF TAMPA Dr. Malek Lashgari - UNIVERSITY OF HARTFORD Dr. Priscilla Liang - CALIFORNIA STATE UNIVERSITY, CHANNEL ISLANDS Dr. Tony Matias - MATIAS AND ASSOCIATES Dr. Patti Meglich - UNIVERSITY OF NEBRASKA, OMAHA Dr. Robert Metts - UNIVERSITY OF NEVADA, RENO Dr. Adil Mouhammed - UNIVERSITY OF ILLINOIS, SPRINGFIELD Dr. Shiva Nadavulakere – SAGINAW VALLEY STATE UNIVERSITY Dr. Roy Pearson - COLLEGE OF WILLIAM AND MARY Dr. Veena Prabhu - CALIFORNIA STATE UNIVERSITY, LOS ANGELES Dr. Sergiy Rakhmayil - RYERSON UNIVERSITY, CANADA Dr. Fabrizio Rossi - UNIVERSITY OF CASSINO, ITALY Dr. Robert Scherer – UNIVERSITY OF DALLAS Dr. Ira Sohn - MONTCLAIR STATE UNIVERSITY Dr. Reginal Sheppard - UNIVERSITY OF NEW BRUNSWICK, CANADA Dr. Carlos Spaht - LOUISIANA STATE UNIVERSITY, SHREVEPORT Dr. Ken Thorpe - EMORY UNIVERSITY Dr. Robert Tian – SHANTOU UNIVERSITY, CHINA Dr. Calin Valsan - BISHOP'S UNIVERSITY, CANADA Dr. Anne Walsh - LA SALLE UNIVERSITY Dr. Thomas Verney - SHIPPENSBURG STATE UNIVERSITY Dr. Christopher Wright - UNIVERSITY OF ADELAIDE, AUSTRALIA Volume 16(3) Special Issue: International Perspectives ISSN 1499-691X

Authors have granted copyright consent to allow that copies of their article may be made for personal or internal use. This does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale. Any consent for republication, other than noted, must be granted through the publisher:

North American Business Press, Inc. Atlanta - Seattle – South Florida - Toronto

©Journal of Applied Business and Economics 2014

For submission, subscription or copyright information, contact the editor at: [email protected]

Subscription Price: US$ 360/yr

Our journals are indexed by UMI-Proquest-ABI Inform, EBSCOHost, GoogleScholar, and listed with Cabell's Directory of Periodicals, Ulrich's Listing of Periodicals, Bowkers Publishing Resources, the Library of Congress, the National Library of Canada. Our journals have been used to support the Academically Qualified (AQ) faculty classification by all recognized business school accrediting bodies.

This Issue

Financial Contagion: An Empirical Investigation of the Relationship Between Financial-stress Indexes of Australia and the US ...... 11 Sandra Mukulu, Samanthala Hettihewa, Christopher S. Wright

A key departure in this study from many earlier studies is that, on the continuum of financial stress from nil to very high, both very high levels of stress and very low levels are seen as being harmful and potential harbinger of a financial-market crisis. Specifically, a surfeit of stress can act as a tipping point into crisis and a dearth of stress can encourage hubris and increase a nation’s susceptibility to financial contagion from another nation; even one that is far removed by geographic and/or economic distance. This paper focuses on developing financial stress indices for the US and Australia using composite market indices, trade weight indices and yields on securities with different maturity dates. Monthly data from January 1989 to December 2011 was sourced from the Australian Bureau of Statistics (ABS), the Reserve Bank of Australia (RBA), the Federal Reserve Bank (FRB), the Bureau of Economic Analysis (BEA), the Federal Reserve Bank of St Louis website, Bank of Canada, Reserve Bank of New Zealand and Yahoo finance website. For purposes of this study the aggregate measures of stress consists of inverted yield spreads, volatility measures for market indices, volatility measures of trade weighted indexes, risk spreads, credit risk spreads and a measures of risk in the equity market.

Impact of Hedge Funds on Traditional Investment Products ...... 35 Kaouther Flifel

The purpose of this paper is to present the hedge fund industry in order to demonstrate that their primary interest in asset management is to diversify a standard portfolio because of their decorrelation with markets, and this by conducting research on two areas: First of all we will analyze the correlation that may exist between alternative and traditional investment products. Secondly we generate portfolio diversification thanks to the dynamic investment model in discrete time combined with the approach for evaluating the empirical probability. Our results emphasize the importance of the criterion of belonging to an investment strategy for a hedge fund.

Technology Transfer from MNCs to Host Country Enterprises: An In-depth Analysis Based on Game Theories ...... 51 Peihua Zhao, Weiguo Zhang, Robert Guang Tian

The problem of MNCs’ technology transfer is virtually the dynamic game problem between MNCs and host country enterprises. The paper introduces evolutionary game theory to the field of technology transfer. For the problem of technology transfer between MNCs and host country enterprises, Evolutionary game theory sets up the model of an evolutionary game according to host country enterprises whether they have research and development capability or not, and comparatively analyzes the equilibrium results. Finally, the paper draws a useful and enlightening conclusion.

Factor Decomposition of the Gender–Job Satisfaction Paradox: Evidence from Japan ...... 69 Shiho Yukawa, Yuki Arita

Previous studies found that although women have disadvantages in terms of wage and working conditions in labor markets, they derive more satisfaction from work than men do. This is called the “gender–job satisfaction paradox.” In this paper, we use a data set composed of company personnel data and employee survey data to examine whether such a paradox exists in Japan. Also, we use the Oaxaca– Ransom decomposition technique to reveal the main factors contributing to this paradox. We found a gender–job satisfaction paradox in treatment job satisfaction.

Inter-firm Relationships and the Creation of Social Capital ...... 90 Cameron Gordon, Shaun Cheah

The article explores inter-firm joint actions, both short and long term, using a social capital framework. The study reviews the literature on social capital generally and its application to inter-firm (or B2B) relationships specifically, finding these applications quite limited at present. The paper then conceptualizes a typology of joint actions firms typically engage in, their outcomes, and how they could contribute to the building of jointly owned social capital across firms. A more formal conceptual model of the creation, accumulation and use of inter-firm social capital is then constructed to be used in future empirical testing and managerial application.

Tax Morale in Socio-Political Interactions: Insiders and Outsiders ...... 101 Savaş Çevik

This paper analyzes the importance of social and political contexts and individual value norms in tax morale. It introduces an approach to discuss tax morale with the notions of ‘insider’ and ‘outsider’ to capture socio-political context and personal value orientations. It constructs a general framework on tax compliance in socio-political context and analytically and statistically demonstrates the importance of identities and personal norms to explain the level of tax morale. Statistical results of an estimated logistics regression model from the World Value Survey data are generally consistent with hypothetical expectations. Results from the analysis indicate that social capital and political confidence are significant in order to estimate the level of tax morale. Moreover, individual reciprocal tendency, sensitiveness to expectations of others, collectivist orientation, and obedience tendency of authority are important determinants of tax morale.

Credit Search ...... 116 Simangaliso Biza-Khupe

There is a notable paucity of studies on the antecedents to consumer credit and debt behaviour, despite the importance of this subject matter. Moreover, there is anecdotal evidence of Data Smog within consumer credit markets, and yet this element remains relatively unexplored in the literature. This study proposes and empirically tests a comprehensive model of consumer credit search behaviour using structural equation modelling. Also, and simultaneously, the model tests for the antecedents of Data Smog, and its effect on Credit Search. The findings allude to the role that Data Smog plays in consumer financial markets and provide insights to the complexities of consumer financial decision-making processes. The paper concludes by discussing the implications of these findings to theory and policy, particularly as concerning rethinking financial information regulation in consumer financial markets.

Industrial Clustering Approach in Regional Development: The Case of Turkey ...... 135 Mustafa Cem Kirankabeş, Murat Arik

In this paper, a "3-Star Analysis,” commonly used in cluster mapping studies in the European Union (E.U.), was conducted, and manufacturing sectors with clustering potential in Turkey were determined across the 26 regions (NUTS 2). This study first introduces a novel concept of “cluster density index” for the manufacturing sectors in Turkey and then analyzes the relationship between the cluster density index and openness, economic development level and public incentives for investment. In this analysis, we used the non-parametric spearman’s rank correlation to test the relationships between the variables of interest.

Effectiveness of Corporate Social Responsibility in Enhancing Company Image ...... 152 Siphiwe P. Mandina, Christine V. Maravire, Victoria S. Masere

In recent times there has been much debate about whether corporations should be socially responsible or not and to what extent they should be responsible. This paper investigates the effectiveness of Corporate Social Responsibility (CSR) in enhancing company`s image, using Unki mine (UM) as a case study, with corporate philanthropy as center of focus. Research design was descriptive and exploratory. A sample size of 208 respondents was used. Stratified sampling technique was used and the population was divided into four stratums which are as follows: um management, UM employees, the local community and lastly Tongogara rural council employees. The research findings show that Philanthropic activities do enhance company image as well as relations between an organization and the community surrounding it. This paper recommends that um could put more investment on the other dimensions of CSR such as ethical responsibility, legal responsibility and economic responsibility.

Effect of Automated Teller Machine (ATM) on Demand for Money in Isolo Local Government Area of Lagos State, Nigeria ...... 171 Fatai Abiola Sowunmi, Zakariyah Olayiwola Amoo, Samuel Olasode Olaleye, Mudashiru Abiodun Salako

The study examined the effects of Automated Teller Machine (ATM) on demand for money. Primary data were analysed using difference of means and probit analyses. The study revealed that ATM has reduced queues in the banking hall significantly. The result showed that the frequency of demand for money to meet transactionary and precautionary motives is significant greater through ATM while average amount withdrawn is smaller compared to teller (p<0.05). Also the probability of a resident using ATM is 0.92. Investment in quality ATM and provision of alternative source of electricity are ways of improving the effectiveness of ATM.

Optimizing Patient Flow and Resource Utilization in Out Patient Clinic: A Comparative Study of Nkawie Government Hospital and Aniwaa Health Center ...... 181 John Mensah, David Asamoah, Akua Amponsaa Tawiah

The study is a comprehensive evaluation to explore current systems and practice regarding the Patient Flow and resource utilization in an out-patient clinic in both a government and private owned hospitals in Ghana. Currently we are witnessing unprecedented queues in these hospitals. the general information was analyzed by a single-phase queuing system. Findings indicated that the estimated mean arrival rate and the waiting time at the OPD for the public hospital were 23 and 0.5 hours respectively, and 25 and 0.5 hours for private hospital and this is directly opposite queuing situations found in developed countries (Ortola, 1993).

GUIDELINES FOR SUBMISSION

Journal of Applied Business and Economics (JABE)

Domain Statement

The Journal of Applied Business and Economics is dedicated to the advancement and dissemination of business and economic knowledge by publishing, through a blind, refereed process, ongoing results of research in accordance with international scientific or scholarly standards. Articles are written by business leaders, policy analysts and active researchers for an audience of specialists, practitioners and students. Articles of regional interest are welcome, especially those dealing with lessons that may be applied in other regions around the world. This would include, but not limited to areas of marketing, management, finance, accounting, management information systems, human resource management, organizational theory and behavior, operations management, economics and econometrics, or any of these disciplines in an international context. Focus of the articles should be on applications and implications of business, management and economics. Theoretical articles are welcome as long as their focus is in keeping with JABE’s applied nature.

Objectives

. Generate an exchange of ideas between scholars, practitioners and industry specialists

. Enhance the development of the Business and Economic disciplines

. Acknowledge and disseminate achievement in regional business and economic development thinking

. Provide an additional outlet for scholars and experts to contribute their ongoing work in the area of applied cross-functional business and economic topics.

Submission Format

Articles should be submitted following the American Psychological Association format. Articles should not be more than 30 double-spaced, typed pages in length including all figures, graphs, references, and appendices. Submit two hard copies of manuscript along with a disk typed in MS-Word.

Make main sections and subsections easily identifiable by inserting appropriate headings and sub-headings. Type all first-level headings flush with the left margin, bold and capitalized. Second-level headings are also typed flush with the left margin but should only be bold. Third- level headings, if any, should also be flush with the left margin and italicized.

Include a title page with manuscript which includes the full names, affiliations, address, phone, fax, and e-mail addresses of all authors and identifies one person as the Primary Contact. Put the submission date on the bottom of the title page. On a separate sheet, include the title and an abstract of 200 words or less. Do not include authors’ names on this sheet. A final page, “About the authors,” should include a brief biographical sketch of 100 words or less on each author. Include current place of employment and degrees held.

References must be written in APA style. It is the responsibility of the author(s) to ensure that the paper is thoroughly and accurately reviewed for spelling, grammar and referencing.

Review Procedure

Authors will receive an acknowledgement by e-mail including a reference number shortly after receipt of the manuscript. All manuscripts within the general domain of the journal will be sent for at least two reviews, using a double blind format, from members of our Editorial Board or their designated reviewers. In the majority of cases, authors will be notified within 60 days of the result of the review. If reviewers recommend changes, authors will receive a copy of the reviews and a timetable for submitting revisions. Papers and disks will not be returned to authors.

Accepted Manuscripts

When a manuscript is accepted for publication, author(s) must provide format-ready copy of the manuscripts including all graphs, charts, and tables. Specific formatting instructions will be provided to accepted authors along with copyright information. Each author will receive two copies of the issue in which his or her article is published without charge. All articles printed by JABE are copyrighted by the Journal. Permission requests for reprints should be addressed to the Editor. Questions and submissions should be addressed to:

North American Business Press 301 Clematis Street, #3000 West Palm Beach, FL USA 33401 [email protected] 866-624-2458

Financial Contagion: An Empirical Investigation of the Relationship Between Financial-stress Indexes of Australia and the US

Sandra Mukulu University of Ballarat, Victoria, Australia

Samanthala Hettihewa University of Ballarat, Victoria, Australia

Christopher S. Wright Burgundy School of Business, France

INTRODUCTION

Recent global financial crisis (GFC) and the European Credit Crunch (ECC) demonstrated how countries that are different in their economic structure, with location differences and even without substantial economic links can be influenced in different degrees by the financial crisis in other countries pressurizing the global investors and policy makers searching for solutions to mitigate/avoid crisis forming due to the contagion effects. The GFC have enflamed fears of how financial contagion (see Allen and Gale 2000 for details on financial contagion) can rapidly inflict panic in even apparently vibrant economies and transform them from prosperity to crisis. While financial contagion risk constitutes a legitimate fear, its path and effects are so convoluted that they are often unpredictable a priori and may be driven more by the attributes of potential destination nations than those of the source nation.

Journal of Applied Business and Economics vol. 16(3) 2014 11 The Objective of the Study This paper accepts that, as discussed later, there is clear incontrovertible evidence of the existence of financial contagion; the issue being reviewed is, what conditions of. Domestic/trade and/or financial links with the source nation can make financial contagion a significant issue and/or an ancillary issue); what and when should national policy makers act to deflect, mitigate, or ameliorate the harm caused by financial contagion. The central aim of this study is to find out the use of financial factors to gauge the financial stability of an economy with particular emphasis on the risk factors for the contagion of financial distress from one country to another examining Australia and US in this empirical study. A key departure in this study from many earlier studies is that, on the continuum of financial stress from nil to very high, both very high levels of stress and very low levels are seen as being harmful and potential harbinger of a financial-market crisis. Specifically, a surfeit of stress can act as a tipping point into crisis and a dearth of stress can encourage hubris and increase a nation’s susceptibility to financial contagion from another nation; even one that is far removed by geographic and/or economic distance. The rest of this paper is organized with: A literature review that gives special emphasis to foundation theories on the fundamentals of asset pricing and market performances; A review of financial stress in Australia and the US; A summary of the data gathering approach, research methodology/design and initial findings; A discussion of the index building approach and technique; A review of Granger Causality and this study’s Granger outcomes; and A concluding discussion of the issues raised by this study’s findings and how the issues raised might be addressed by future research.

LITERATURE REVIEW

Theoretical Background Modern Portfolio Theory (Markowitz, 1952), Efficient Market Hypothesis (EMH) developed by Fama (1970), Capital Asset Pricing Model (Sharpe, 1964) and the other theories that have, until recently, dominated the way investors and researchers evaluate market performance were unable to suggest the causes and potential remedies to these often contagion-driven panics. As the foundation theories of financial-market performance are based on assumptions that: risk is defined by volatility; investors act rationally; all available information is costless to gather and is incorporated into decisions effortlessly and seamlessly. Such assumptions have led to the axiom that security prices reliably and fairly reflect intrinsic value and, as a result, unexpected asset returns always follow a random walk. As a result, above-normal returns are always unpredictable. These long-run notions are confirmed by the Theory of Arbitrage (Fama, 1965; Ross (1976)) where short-run price deviation is quickly identified by rational investors who, in seeking to profit from it, rapidly normalizes it and returns the market to the long-run equilibrium. As a result, investment markets are expected to be efficient in the long run and the price of assets in those markets are expected to reflect their intrinsic value (Fama, 1965). Nonetheless, irrationality and repeated errors-in-judgment appeared periodically, mostly every 10-20 years, stressed financial markets, but the timing and magnitude of those bouts are difficult to predict (Ball, 2009). The financial market meltdown in 1987, the 1990 Japanese economic crisis; the 1994-1995 Mexican crisis (Kindleberger & Aliber, 2005; Mazumder & Ahmad, 2010); Asian financial crisis in 1997 and a subprime mortgage crisis in 2007 which led to a global financial crisis are among the many significant bouts of irrationality and recurring errors in judgment that continue to recur and that many researchers (including Fama & French, 1993 and 1996) have difficulty in explaining them within the framework promulgated by the foundation-theories of finance. Keynes (1936, p. 148), in explaining equivalent recurrent events, noted that the “...practice of calmness and immobility, of certainty and security [of investor forecasts], suddenly breaks down ... [and is then] subject to sudden and violent changes”.

Behavioral Finance and Prospect Theory As researchers identified difficulties with the traditional paradigm (including investor behaviors seen as irrational) and searched for ways to explain those difficulties and irrationalities, they initiated the new sub-disciplines of Behavioral Finance (De Long, et al, 1990; Shelifer & Vishny, 1997) and Prospect

12 Journal of Applied Business and Economics vol. 16(3) 2014 Theory (Kanheman & Tversky, 1979). Given that over the last few decades, repeatedly, many hundreds of billions of dollars were lost as the hubris of a Bull Market collapsed into the overly timorousness of Bear Market, it is important to draw from the new perspectives in finance, ways to identify the symptoms and patterns within the financial stress stages so that either the wild market swings can be moderated or prudent investors can avoid being dragged into losses by less prudent investors, or even profit by identifying the coming financial storm.

Extant Indices of Financial Stress Financial stress is a feeling of unease that arises from fear of what may occur if one’s income does not at least cover outgo; a fear, also, known as the Micawber Principle.1 A financial stress index is a measure of the aggregated financial stress within a nation and is on a continuum ranging from a low indicative of a pending “bubble crisis”, a high indicative of an imminent “crash crisis” (Illing & Liu, 2006), and a propensity to swing rapidly and with little warning from a bubble to a crash. The goal of this study is to explore the use of economic and financial factors to gauge the financial stability of a country with particular emphasis on the risk factors for the contagion of financial distress from one country to another. The evidence of financial market volatility, and episodes of financial crises (distress), suggests that financial markets are either inefficient or that the semi-strong market- efficiency exists in the long run and is punctuated by periodic bouts of inefficient hubris that can quickly switch to transpose inefficiency of being overly timorous. This study concentrates on identifying what factors affect macro-finance stress and the risk of contagion between Australia and one of its longest and most prominent trade partners (the USA). This study uses the review of extant studies, the indices and Granger analysis to answer the following research questions: i) What is the economic impact of financial stress in Australia? ii) What are the risk factors for financial contagion? iii) What stress indicators should be part of a comprehensive index of financial stress in Australia? iv) What is the value of using the proposed index to predict periods of financial distress in Australia? v) What are the limitations of using a financial-stress index to forecast financial distress?

In order to achieve macroeconomic stability it is essential to closely monitor the economic environment and implement policies that safeguard against: on one hand, financial stress escalating into financial distress; and on the other, a dearth of stress that encourages investor hubris and leads to a bubble and collapse that precipitates financial stress. Wolf (2009) and Fox (2010) worry that economists, overly focused on the ideals and market conditions needed to achieve the elusive but notionally ideal laissez- faire of a strong-form efficient-market hypothesis (EMH), have failed to consider other more-workable options. A second-best, but highly achievable focus would be on the economic analysis needed to identify the what and the when of an intervention, to prevent/mitigate the distress of a pending financial crisis. Siegel (2009) asserts that the EMH nurtured an environment where asset bubbles, poor policy controls, and engineering of malicious and complicated instruments thrived and subsequently led to the subprime mortgage crisis of 2007. Similarly, Krugman (2009) asserts that economists overlooked the reality of market imperfections and that securities can and often are incorrectly priced. Further, if securities were overpriced traders in the financial markets are human and investors are not always as rational as what is proclaimed by many economists. Behavioral models use other ways to explain stock- price anomalies including over and under reactions, herding behaviors, momentum strategies, investor overconfidence and firm-size bias (Barberisb et al., 1996; Daniel et al., 1998; Dremen & Lufkin, 2000; Chan, 2001). EMH model, with its laissez faire ideology, does not encouragement policy makers to intervene, even when they realize that a market is slipping into distress. Specifically, this ideologically driven policy of leaving markets to self-correct inevitably increases the occurrence, length and depth of financial crises and associated distress.

Journal of Applied Business and Economics vol. 16(3) 2014 13 Currently, little is known about the appropriate structure and content of a financial- stress index, or where the tipping-point(s) occur(s), what creates a risk of contagion, and, how to modify/regulate the financial markets to avoid or ameliorate the descent into the financial distress of a financial crisis.

The Relationship Between Micro- and Macro-Financial Stress Financial stress can be experienced at a microeconomic or macroeconomic scale. The microeconomic school focuses on financial stress at the household and individual businesses scale; whereas, the macroeconomic school relates to financial stress at the national or regional economy scale. While these two scales can be interrelated, not every micro-level stress accumulates into a macro-scale stress. However, uncontrolled micro financial stress can have spill-over effect which often causes the stress at macro level vice versa. Given that these stresses accumulate on a systematic basis, the final outcome can be unexpected and unpredictable in both direction and magnitude. Micro-finance stress: is “...the adverse economic or social outcomes associated with a household’s [/business’s] financial situation, including debt repayment problems, delinquency, bankruptcy and lack of discretionary income” (Worthington, 2006, pp. 2-3). Breunig and Cobb-Clark (2006) and Wesley Mission (2006) explain the different conditions that fuel the micro-level stress. Ehrhardt and Brigham (2011) explain the micro-stress with administrative, economic and financial factors. Lakonishok and Smidt, (1988) and French (1980) state how tax benefits can encourage excessive debt financing, which adds to micro-level financial stress. Dun and Bradstreet Credibility Corp (2012) show how extensive episodes of stress often culminate in bankruptcy, receivership or liquidation. Macro-finance stress: definitions focus on how financial stress impacts on a national economy. This kind of stress can be defined in several ways, depending on the factors that triggered an episode of distress. Since these factors vary, depending on (among other things) political and economic characteristics, settling on a definition that will incorporate all characteristics of historical episodes of stress is difficult. Nevertheless, Hakkio and Keeton (2009) maintain that regardless of the origin of stress, financial stress generally results in the “...interruption of the normal functioning of financial markets”. As noted earlier in this study, moderate levels of financial stress can act as a moderating influence on markets by discouraging the investor hubris that can lead to bubbles, contagion, and the distress of a financial crisis. A more specific definition (proposed by Illing and Liu (2006)), suggests that macro-economic financial stress is the anxiety experienced by producers or consumers due to increased uncertainty and changing expectations of economic losses in the financial markets and institutions. Similarly, an episode of financial stress is understood as a period when a country’s financial system is under pressure and the country lacks adequate resources to facilitate a quick transition out of an economic slump. Typically, countries under excessive financial stress can experience significant changes in commodity prices, a rapid increase in risk and/or uncertainty, limited liquidity, and fears about the health of the banking system (Balakrishnan, Danninger, Elekdag, & Tytell, 2009). Interdependence: On one hand, Hakkio and Keeton (2009) point out that when financial markets are in distress, the savers are less willing to lend money unless they are provided with more security and/or a premium to compensate for increased risks of default. As a result, increased uncertainty at a macroeconomic level can contribute to a credit crunch at the microeconomic level. On the other hand, Gramlich and Oet (2011) suggest that structural fragility in key financial and regulatory agencies can lead to and/or potentiate a crisis, as seen in the subprime mortgage crisis. Thus, while the notion of financial stress and financial crisis are closely related, they are not the same—an important distinction between the terms is that financial crises are found at the extremes of the financial stress continuum (Illing & Liu, 2003).

The Importance of the Study There is extensive research on the impact of financial stress at the microeconomic level (Bray, 2001; Breunig & Cobb-Clark, 2006; Commonwealth Department of Family and Community Services, 2003; Marks, 2007; Wesley Mission, 2006; Worthington, 2006). However, the current high volatility in financial markets and the rising risk of contagion as globalization intensifies, make it ever more important

14 Journal of Applied Business and Economics vol. 16(3) 2014 to identify and measure macro-level financial stress. As a result, building and calibrating an appropriate financial stress index, at the nation level, is an ever more vital element of maintaining the well-being of financial markets. Although there is much research on the factors that may contribute to financial stress and the distress of a financial crisis, there is little research on combining those factors into an appropriate and useful financial stress index.

The Nature and Origin of Financial Stress Origin of the financial stress: Not all periods of stress are harmful as certain economic disturbances are linked to economic recessions or downturns while others have little impact. Consequently, financial stress can be used as a warning of an impending crisis; if the financial stresses are building from rising forces that result from a loss of confidence or other disturbances that are consistent with an impending down-turn and/or recession. Nevertheless, while an episode of financial stress often leads to a correction, it does not always translate into a financial crisis. The main issue of concern for analysts is the nature and origin of the financial stress and the volatility of the markets (e.g. a spark that is harmless in most times and/or places can lead to a conflagration, if it occurs in a tinder-dry forest). While researchers have identified many different contributing factors, their influence individually and in combination is not as well understood. The IMF states that the probability of an economic recession depends on the degree to which house prices or aggregate credit rose before the episode of financial stress (International Monetary Fund (2008)) and identified a positive relationship between financial stress and large increases in housing prices or credit. The larger the increase in credit and house prices the more the financial stress and vice versa. Hakkio and Keeton (2009) assert that the depth and length of a subsequent recession depends on the extent to which firms and households reduce spending and cut costs. Moreover, the added push from structural weaknesses in the banking sector can often generate more severe economic downturns than the effects of securities or foreign-market-related stress. Illing and Liu (2006) suggest that countries with weak financial systems are a fertile ground for economic shocks to germinate into financial stress and on into financial crises. Furthermore, Misina & Tkacz (2009) suggest that an absence of financial crisis in a country does not mean that it is safe from future episodes of financial distress. For example, seeming tranquility in an economy may mask rising stress levels, either via the slow, steady increase in financial imbalances over time or via rapid contagion from other economies. Thus, there is an onus on regulatory authorities to keep watch and intervene before an episode of rising stress matures into a full-blown crisis.

Contagion Kaminsky and Reinhart (2000) see contagion as process by which “financial difficulties spread from one economy to another in the same region and beyond” via trade and financial linkages (p. 51). In an empirical investigation of the contagion process Hettihewa and Mallik (2005) examined co-integration between eight countries. It is difficult to determine if contagion at a regional level arises from mostly financial or from trade links because countries usually co-establish regional trade agreements and corresponding interbank linkages to facilitate the trade agreements (Caramazza, Ricci, & Salgado, 2004). The 1997 Asian financial crisis is an example of regional contagion that originated in Thailand and spread to neighboring Indonesia, South Korea, Philippines and Malaysia. Aggressive dealings by speculators convinced foreign lenders to cease all loans denominated in the rupiah, the ringgit and the won to minimize the impact of speculation. The ultimate effect of such Draconian actions was to slow the adjustments in the region, to starve the affected countries of foreign reserves, and to reduce their ability to service their foreign debt (Kindleberger & Aliber, 2005). Another variant of contagion, suggested by Calvo and Mendoza (1998), can develop even if linkages are absent or fully controlled. Specifically, it afflicts financial markets because of a herding behavior that may not be fully rational (Caramazza et al., 2004).

Journal of Applied Business and Economics vol. 16(3) 2014 15 Trade Theory, Globalization and Contagion2 International trade has played a key role in the economic growth and development of any economy. For instance, Dornbusch, Fischer, and Samuelson (1977) developed a Ricardian model for a continuum of goods, based on the notion that comparative advantage is mainly driven by differences in technology across nations and that labor is the most relevant factor to consider in the analysis of comparative trade. Feenstra and Hanson (1996) used the Ricardian model to explore the relationship between wages and international trade in a capital based economy (the US), verses a labor based economy (Mexico). The study found that when the US outsourced labor, relative demand for non-production labor increased while demand for production labor decreased. Correspondingly, there was a drop in the wages of production workers and an increase in the labor rates of non-production workers. 3 Moreover, the relative demand for production workers rose in the labor intensive country with a corresponding upward pressure on labor rates. The resulting outcome was mixed with: On one hand, winners included Mexico (gaining from jobs creation), US companies (that minimized their production costs with cheaper offshore labor) and non- production staff in US (who realized an increase in their wages); On the other hand, losers included production workers in the US (who missed-out on jobs which were outsourced to other countries with employers preference shifting from the local market to the international ones unless employees were willing to take a significant pay-cut in order to keep their jobs) and sectors of the US market that service the US production workers. Theoretically, a utopian country with no trade or financial linkages with the rest of the world is safer from financial-stress contagion, but misses-out on the benefits of international trade. Conversely, a high dependence (i.e. an open small economy) means that a country is more susceptible to events in other countries. (see the United Nations Conference on Trade and Development (2008); Oxfam, 2002) for details). Thus, globalization multiplies the pathways for greater economic growth and for increased risk of economic recession. In effect, policies reducing trade barriers create opportunities for benefit to an economy and risks of harm (Hettihewa & Wright 2010; Stiglitz & Charlton, 2007). For instance, Gramlich and Oet (2011) state that the Greek crisis spilled into neighboring European countries via the increased structural fragility arising from linkages such as the interbank lending relationships, credit lines and solvency issues. Financial globalization now involves a network that facilitates quick transmission of economic shocks from one country to another. The dilemma that most countries find themselves in is optimizing the degree of financial integration so that they can enjoy an ease of trade, but minimize their exposure to contagion stress (Stiglitz, 2010). Interdependent relationships are regarded as a technique of insuring a country against excluded from the benefits of foreign market changes (Daniels & Radebaugh, 1995). However, financially interdependent countries are more likely to share financial crises and countries who are more dependent than interdependent are more of a one-way pipeline (i.e. having to suffer any imported crises alone rather than passing some of the pain on to others). Thus, higher levels of dependence are associated with increased vulnerability to contagion. The vulnerability of a country to contagion of financial stress also depends on the degree to which a country depends on trade with other countries.

AUSTRALIA’S TRADE RELATIONSHIP WITH THE US AND ITS EXPERIENCE WITH THE GFC

The U.S. has for decades been one of Australia’s leading trading partners. According to the Ehrhardt and Brigham (2011), US was Australia’s largest bilateral trading partner in 2010-11; US exported 36.3 billion dollars’ worth of goods and services to Australia and imported 14.2 billion dollars’ worth of products and services from Australia. “In 2010–11, the United States was also Australia's largest services trading partner and third largest merchandise trading partner” (p. 192). The main non-trade related links between the countries relate to national security through sharing of technological knowledge, training techniques and intelligence as per the 1951 ANZUS treaty. Since the establishment of the treaty the two countries have continued to engage in trade agreements that encourage liberal trade not only in the two countries but also in the Asia Pacific region (Ehrhardt & Brigham, 2011). Also, the current episode of GFC indicates that the impact of US subprime crisis had mammoth flow on effect on many countries

16 Journal of Applied Business and Economics vol. 16(3) 2014 including Australia making it is vital for Australia to investigate the contagion effect between these two countries. Australia has experienced a few episodes of volatility in its financial markets in the past few decades. Indeed many such episodes experienced in the Australian economy over its history can be seen as contagion driven financial stress. In a recent example, the GFC started in the US and affected global markets, including those in Australia. In recent decades, Australia has for a variety of reasons, been very effective in resolving economic downturns. Specifically, Australia has stable financial institutions with strong prudent regulatory measures that shielded it from the worst depredations of the GFC. The Australian Prudential Regulation Authority (APRA) facilitated strong regulation, close supervision and effective risk management of Australian banks which nurtured a stable banking sector in Australia, as compared to many overseas countries.4 Also, the Australian government took pre-emptive measures to ensure Australian banks had sufficient foreign currency at their disposal to prevent bank runs. Further, even after the GFC was well established, Australia enjoyed an extended economic boom, due (in large part) to the exports of its mining industries (Perlich, 2009). A more interesting question is: with its strong regularity standards and strong mining sector, how strong is Australia’s capacity to resist global financial crisis? While, in answering this key question, it is essential to find out the contagion relationship with all the trading partners, this study takes a first step by looking into the contagion effect between the USA and Australia.

DATA AND METHODOLOGY

Background for Contagion Measures Determining the impact of contagion related stress between two countries compared to that between few countries is a less complicated task. This study focuses on financial stress, and contagion of stress transmitted from one country to another. The main countries of interest are Australia and the United States. The US and Australia have a long history of trade and non-trade links that are aimed at cooperatively enhancing economic prosperity and national security. While in recent years, Australia has shifted much of its trade focus away from the US toward Asian countries (China in particular), it takes time for institutions that have developed over many decades to refocus. According to the Ehrhardt and Brigham (2011), “...the United States is ... [still] Australia's largest services trading partner and third largest merchandise trading partner”. Also, the non-trade related links between the countries (national security via sharing of technological knowledge, training techniques and intelligence go back to the 1951 ANZUS treaty (Ehrhardt & Brigham, 2011). This study uses a quantitative approach to estimate financial stress in the USA and Australia and to examine whether financial stress in USA can be used to forecast financial stress in Australia. Generally, a country has several trade partners and deciphering the impact of contagion related stress caused by each country could prove difficult. In order to do this, a researcher would not only have to construct a suitable financial stress index for each country but also provide a criterion for checking validity of the financial stress index. Illing and Liu (2003) suggest that a comprehensive measure of financial stress can be designed using a two-pronged approach that incorporates quantitative and qualitative analysis. The former entails the use of empirical criteria to measure the degree of financial stress experienced by each country while the latter utilizes expert surveys aimed at correctly diagnosing episodes of stress in a particular country. Once both factors have been considered the financial stress index of each trading partner can be estimated and the causality relationship examined in order to determine the degree of contagion stress transmitted from one country to another. This technique would ensure that empirical analysis for episodes of stress is supported by prevailing perceptions of economic stress in a particular country. The practicality of this technique is subject to availability of time and resources to conduct surveys in the countries of interest (if no secondary information is available) and construct financial stress indices for all trading partners. Researchers used a more pragmatic approach while examining financial stress in several countries (Cardarelli, Elekdag, & Lall, 2009; Duca &

Journal of Applied Business and Economics vol. 16(3) 2014 17 Peltonen, 2011). Both constructed stress indices and checked whether they signaled historical episodes of financial stress; a similar method was employed in this research.

Data Development and Variable Choice in the Model This paper focuses on developing financial stress indices for the US and Australia using composite market indices, trade weight indices and yields on securities with different maturity dates. Monthly data from January 1989 to December 2011 was sourced from the Australian Bureau of Statistics (ABS), the Reserve Bank of Australia (RBA), the Federal Reserve Bank (FRB), the Bureau of Economic Analysis (BEA), the Federal Reserve Bank of St Louis website, Bank of Canada, Reserve Bank of New Zealand and Yahoo finance website. For purposes of this study the aggregate measures of stress consists of inverted yield spreads, volatility measures for market indices, volatility measures of trade weighted indexes, risk spreads, credit risk spreads and a measures of risk in the equity market. The econometric packages used in the estimation process are: (1) Eviews 7 and (2) IBM SPSS Statistics 19.5

Stress Index and the Variables Developing a plausible starting point for selection of stress variables requires careful consideration of historical episodes of stress or crises. Illing and Liu (2006) suggest that historical crises mainly originated in the banking sector and financial markets that comprise the foreign exchange, debt and equity markets. None of these sectors can be considered in isolation since problems in one sector can be transmitted to other sectors of the economy. Banking related crises are characterized by banking failures and often result from poor regulation or transmission of stress from other sectors. For instance, a steady increase in real estate prices in Japan led to a bubble that eventually burst in the 1990s leaving Japanese banks with large losses. Moreover, the currency crisis in 1997 left many banks in South Korea, Thailand and Malaysia bankrupt (Kindleberger & Aliber, 2005). These bank failures resulted from a sudden drop in real estate prices and depreciation of Asian currencies. The main issue of concern for policy makers is: whether it is possible to anticipate banking crises using macroeconomic and financial sector data and avert a crisis or intervene before an episode of stress translates to a crisis in the banking sector. Demirguc-Kunt and Detragiache (1998, 2005) suggest that low GDP growth rates, high real interest rates and high inflation indicate increased vulnerability of an economy to a banking crisis. Furthermore, increased interbank linkages can help a developed countries hedge against credit risk associated with operating in a country. Demirguc-Kunt and Detragiache (1998) argue that increased integration of banks may be key to strengthening banks worldwide since entry of foreign banks encourages healthy competition and adoption of better banking regulation practices (p. 103). However, setting up these linkages will also increase the vulnerability of a country to financial- stress contagion, so countries should be careful when encouraging increased financial integration across borders. Hardy and Pazarbasioglu (1999) explored the further use of financial and economic indicators in 50 different countries with the aim of identifying suitable banking-stress indicators (e.g. the use of ratios that measure the relationship between banking deposits and GDP). However, many of the proposed variables are not provided at monthly frequency. For example, GDP data is only available on quarterly and annual basis. Originally, the difference between the Eurodollar interest rate (ED) and the Treasury bills interest rate (T) of a country is defined as the TED Spread however, recent studies also define the TED spread as the difference between the London Interbank Offered Rate (LIBOR) rate and treasury bills (Hammoudeh et al., 2011; Lee et al., 2007). This study focuses on the latter definition of the TED spread. The Australian TED spread is the difference between the 3-month LIBOR in Australian dollars and the 90-day bank accepted bills. Rising TED spreads indicate that banks are unwilling to lend to each other for fear of default loss (Hammoudeh, Chen, & Yuan, 2011). Figure 1 shows the TED spreads for Australia and US. Both spreads indicate an increased reluctance by banks to lend to each other especially during the 2007 subprime crisis. Understandably, international banks were less inclined to lend to US banks compared to Australian banks as indicated by a larger spike in the TED spread.

18 Journal of Applied Business and Economics vol. 16(3) 2014 FIGURE 1 TED SPREADS FOR AUSTRALIA AND US

Equity Market Stress Equity market crashes are characterized by plummeting of share prices. For instance, in the 1987 Black Monday share prices in the New York Stock exchange fell by about 33.33% over five trading days in October (Patel & Sarkar, 1998). Speculative forces that intensify during a financial crisis suggest that a country can never be immune to large drops in the share prices regardless of whether a country is an emerging economy or developed countries such as US and Australia. However, share indexes in emerging markets are more fragile and bound to experience larger share price losses compared to developed countries. Illing and Liu (2006) propose the use of a share value at t, indexed to the maximum share value for the review period (CMAX) to measure share volatility in a financial market. Following Illing and Liu (2006), this research adopts a CMAX measure for a period of two years.6 The two year review period is a trade-off between using a longer period which would give better statistics and a shorter period which tends to minimize the change in share prices due to long-term trends.

The CMAX calculation can be expressed as shown in equation 1.

= (1) ( | , , … ) 퐼푡

퐶푚푎푥푡 max�퐼 ∈ 퐼푡−푗 푗=푛 푛+1 푛+2 푇 � CMAX t = share value at t, indexed to the maximum share value for the review period. It = stock index value at time t. t = moving time window during j. j = review period (n to n+T). n = start of review period. T = end of review period.

Hence, the CMAX calculation for both, Australia and US financial stress indexes considers the value of the stock index at time t compared to the maximum value over the past two years.

Journal of Applied Business and Economics vol. 16(3) 2014 19 Figure 2 shows the graphical representation of the computed CMAX for the Dow Jones and the All Ordinaries. As expected, the largest drop in the share indexes was experienced between 2008 and 2009; during the GFC, with the epicenter of the crisis (US) recording higher loss in value of the share index.

FIGURE 2 CMAX GRAPHS FOR AUSTRALIA AND US

Yield Spreads Consistent with Illing and Liu (2006), this research used the inverted spread between a long-term security and a short term security to measured interest rate shocks. Interest rates on a 10-year government bond were used to represent long-term securities in both countries, while yields on a 3-month treasury bill and a 90-day accepted bills were chosen as short-term securities in US and Australia respectively. The inverted spread was calculated by deducting the interest rate of the long-term security from the interest rate of the shorter-term security. The logic being long-term rates are the equilibrium rate and stress is experienced when the short-term rate surpasses the long-term ones. In addition to the yield spreads, this study will measure uncertainty in Australia and US debt markets using risk spreads. Following Illing and Liu (2006), based on the compatibility and risk of the neighboring trading partners, this study evaluates the uncertainty of Australian market considering the difference between yields on 90-day bank accepted bills in Australia and New Zealand. The uncertainty in the US market was calculated as the difference between US and Canadian 3 month treasury bills.

Volatility in Stock Prices and Exchange Rates Stock prices and exchange rates tend to fluctuate more during periods of crisis compared to periods of no crisis. For instance, in 2008, Gujarati (2011) states that the US Dow Jones Index oscillated due to rising oil prices and the 2007 subprime mortgage crisis. More specifically, on 29 September 2008, the Dow Jones lost 777.7 points and subsequently swung upwards and downwards by more than 300 points for most of October 2008 (p.238). Such volatility in stock prices can be incorporated into a financial- stress index using a Generalized Autoregressive Conditional Hetero-skedasticity (GARCH) process (Bollerslev, 1986). In this study, GARCH models are used to capture volatility clustering exhibited in

20 Journal of Applied Business and Economics vol. 16(3) 2014 stock prices and exchange rates using stock indexes and trade weighted indexes of both countries. The stock indexes for the Australian and US market include the All Ordinaries (AOrds) and the Dow Jones Industrial Average (DJ) indexes respectively. Both indexes were deemed representative, as they contained historical information of companies with the largest market capitalizations in the respective countries. The descriptive statistics of four series are contained in Table 1.

TABLE 1 DESCRIPTIVE STATISTICS FOR STOCK AND TRADE WEIGHTED INDEXES

LNDJ LNAORDS LNATWI LNUTWI Mean 8.384316 7.625621 4.132516 4.538750 Median 8.566813 7.712085 4.094345 4.525038 Std. Dev. 0.902329 0.694669 0.170700 0.155443 Observations 384 384 384 384

The natural log of Dow Jones stock index (LNDJ) reports a higher standard deviation in returns from the mean compared to the natural log of the All Ordinaries index (LNAORDS). Similarly, the natural log of US trade weighted index (LNUTWI) has higher standard deviation compared to the natural log of Australian Trade weighted index (LNATWI). This suggests that for the selected period the Australian market has enjoyed lower fluctuation in the share prices and exchange rates compared to the US. Figure 3 shows the graphs for natural logarithm of each series from January 1980 to December 2011. To confirm this observation, the Augmented Dickey Fuller (ADF) tests developed by Dickey and Fuller (1979, 1981) were conducted on all series.

FIGURE 3 GRAPHS OF STOCK INDEXES AND TRADE WEIGHTED INDEXES

Journal of Applied Business and Economics vol. 16(3) 2014 21 Because the stock index series trend upwards, a trend term is included in the ADF tests for these series. Therefore, the estimating regression for the ADF test7 is as shown in equation 2.

= + + + + + + (2)

푡 푡−1 1 푡−1 푝 푡−푝 푡 = 1st∆푦 difference훼 of훽푡 the stock휌푦 index훾 ∆푦 ⋯ 훾 ∆푦 푒 α = a constant term, ∆푦β푡 = coefficient of the trend term t = the trend term ρ = the coefficient of the lagged stock index, st γ1 = coefficient of the 1 difference of the first lag of the stock index, st th γp = coefficient of the 1 difference of the p lag of the stock index, et = error term

The trade-weighted indexes follow a random walk. the equation for conduction the ADF tests for these series excludes the trend term in equation 1:

= + + + + + (3)

푡 푡−1 1 푡−1 푝 푡−푝 푡 The ∆푦number훼 of lags휌푦 (p) in 훾equation∆푦 2 and⋯ 3 훾were∆푦 determined푒 using the Modified Akaike Information Criterion (MAIC) as proposed by Ng and Perron (2001). Table 2 contains the results of the ADF tests. As expected, all series are non-stationary at the level as each series contains a unit root. By contrast, the first difference of each series proves stationary at any level of significance. Therefore, the first difference of each series was used to estimate the GARCH models using the GARCH process developed by Bollerslev (1986). TABLE 2 UNIT ROOT TESTS USING ADF TEST (FOR MONTHLY DATA)

Variables ADF test Level First difference lnAOrds -2.660 -9.878*** lnDJ -1.218 -6.819*** lnUTWI -1.239 -5.045*** lnATWI -1.897 -12.968*** * indicates that the Dickey-Fuller tau statistic is significant at the 10% (*), 5% (**) or 1% (***) level.

We start with a simple GARCH(1,1) or an AR(1)-GARCH(1) model before considering other GARCH models such as a GARCH (2,1) model and IGARCH models. The estimated AR(1)-GARCH(1) model can be written in equation form as shown in equation 4a & 4b.

= + + (4a)

푡 0 1 푡−1 푡 푦 =휙� +휙� 푦 +푒̂ (4b) 2 2 2 Table 3 shows휎�푡 the훼�0 estimated훼�1 푒̂푡−1 GARCH훽휎�푡−1 (1, 1) models for the four series using the ordinary least squares approach. The GARCH (1, 1) model proves sufficient for the stock indexes as indicated by the highly significant GARCH parameter estimates. On the other hand, the GARCH (1, 1) model is a poor fit for the

22 Journal of Applied Business and Economics vol. 16(3) 2014 trade-weighted indexes as the β coefficients associated with the variance are highly insignificant. In this case the use of a GARCH (2, 1) may be more suitable8.

TABLE 3 GARCH (1, 1) MODELS FOR STOCK AND EXCHANGE INDEXES

LNAORDS LNDJ LNATWI LNUTWI 0.0055*** 0.0069*** 0.0008 0.0005 (0.0017) (0.0022) (0.0015) (0.0004) 흓ퟎ 0.1887*** 0.0036 0.0373 0.3696*** (0.0471) (0.0632) (0.0702) (0.0312) 흓ퟏ 8.66E-5* 8.86E-5* 0.0006*** 0.0003*** (4.59E-5) (5.06E-5) (0.0001) (7.68E-5) 휶ퟎ 0.1890*** 0.0974*** 0.2983*** -0.0950*** (0.0301) (0.0319) (0.0598) (0.0101) 휶ퟏ 0.7714*** 0.8659*** 0.0354 -0.0356 (0.0442) (0.0395) (0.1366) (0.2581) Robust휷 standard errors in parentheses. * indicates that the coefficient is significant at the 10% (*), 5% (**) or 1% (***) level.

The GARCH (2, 1) process was used to estimate volatility in exchange rate. The estimated GARCH (2, 1) model is shown in equation 5a & 5b and its output shown in Table 4.

= + + + (5a)

푡 0 1 푡−1 2 푡−2 푡 푦 =휙� +휙� 푦 +휙� 푦 +푒̂ (5b) 2 2 2 2 휎�푡 훼�0 훼�1 푒̂푡−1 훼�2 푒̂푡−2 훽휎�푡−1TABLE 4 GARCH (2, 1) MODELS AUSTRALIAN AND US TRADE WEIGHTED INDEXES

LNATWI LNUTWI 0.0015 -0.0007 (0.0015) (0.0008) 흓ퟎ 0.0440 0.3706*** (0.0695) (0.0386) 흓ퟏ -0.0019 -0.0979*

(0.0534) (0.0521) 1.46E-5*** 3.14E-5 흓ퟐ (3.44E-6) (2.54E-05) 휶ퟎ 0.2896*** -0.0863*** (0.0581) (0.0077) ퟏ 휶 -0.2697*** 0.1328*** (0.0581) (0.0290) 휶2 0.9657*** 0.8436*** (0.0140) (0.1003) Robust standard휷 errors in parentheses. * indicates that the coefficient is significant at the 10% (*), 5% (**) or 1% (***) level.

Journal of Applied Business and Economics vol. 16(3) 2014 23 From the results, both indexes show highly significant variance coefficients. Notably, the Australian TWI shows persistent variance with + + = 0.9856 compared to just 0.8901 for the US TWI. Hence, the GARCH (2, 1) model is deemed a suitable fit for the US trade weighted index and not the Australian trade weighted index. Rather,훼1 훼an2 IGARCH훽 process is applicable for modeling persistent volatility in the Australian case (Robert & Bollerslev, 1986). An IGARCH (2, 1) model is estimated by excluding the constant term ( ) from equation 5a and 5b and restricting the sum of + + =1. 0 훼 훼1 훼2 훽 TABLE 5 IGARCH (2, 1) MODEL FOR AUSTRALIAN TRADE WEIGHTED INDEX

LNATWI 0.0036***

(0.0004) 0.0139 ퟎ 흓 (0.0477) -8.78E-5 ퟏ 흓 (0.0343) 0.2798*** ퟐ 흓 (0.0310) -0.2079*** ퟏ 휶 (0.0376) 0.9281*** 2 휶 (0.0136) Robust standard errors in parentheses. * indicates that the coefficient is significant at the 10% (*), 5% (**)휷 or 1% (***) level.

Table 5 shows the resulting IGARCH (2, 1) which is a better fit for modeling exchange volatility compared to the GARCH (2, 1) model. Moreover, the IGARCH model has highly significant GARCH coefficients. Therefore, the IGARCH (2, 1) model is used to model the volatility of the trade weighted index.

INDEX BUILDING TECHNIQUE

Selecting a particular index building technique can prove difficult, as different researchers prefer different techniques to form a composite index. Some common techniques of index construction include Principal Component Analysis method (PCA), variance-equal weights technique, credit weights technique and transformation using sample Cumulative Distribution Functions (CDFs) (Federal Reserve Bank of St Louis, 2010; Hakkio & Keeton, 2009; Holló, Kremer, & Duca, 2012; Illing & Liu, 2006; Louzis & Vouldis, 2011). This study explores the use of the transformation of variables to CDFs method.

Transformation of Variables to CDFs According to Illing and Liu (2006) the transformation of variables to CDF’s method involves converting all variables to their sample CDF before constructing the index. Each variable is expressed in terms of a rank percentile that ranges from 1 to 100. The lowest values of a particular variable are assigned the value one and the highest are assigned the value a hundred. The composite index is computed by taking the arithmetic mean of the transformed variables.

24 Journal of Applied Business and Economics vol. 16(3) 2014 FSI of Australia The stress index for the Australia is computed by calculating the arithmetic average of the variables in equation 6.

= ( + + + + + )/6 (6)

Where, AFSI퐴퐹푆퐼 = Australian퐼푌푆푝 퐿푛퐴푂푅퐷푆 financial stress퐿푛퐴푇푊퐼 index. 푅푆푝 퐶푀퐴푋퐴푂푅퐷푆 푇퐸퐷 IYSp = inverted yield spread. LnAords = volatility for the natural logarithm of the All Ordinaries index. LnATWI = volatility for the natural logarithm of the Australian trade weighted index. RSp = Australian-New Zealand risk spread. CMAXAORDS = risk of equity market loss based on the All Ordinaries share index. TED = credit risk spread.

The resulting financial stress index (FSI) is shown in figure 4 and it depicts that for the three periods between 1989 and 1990, between 1998 and the early 1999 and from late 2008 to 2011 the FSI is greater than 70 with the highest level of historical stress estimated at about 80th percent in 1989 to 1990. A plausible cause of the indicated stress during this period is the Japanese crisis that occurred between 1989 and 1990 (Kindleberger & Aliber, 2005). Stress in 1998 is possible due to the contagion of the 1997 Asian crisis to Australian markets (Corsetti, Pesenti, & Roubini, 1999). The final episode of stress is possibly due to the 2007 GFC and the recent 2010 European Economic crisis.

FIGURE 4 A COMPOSITE FINANCIAL STRESS INDEX FOR AUSTRALIA

FSI of US Similarly, the financial stress index for US can be obtained by calculating the arithmetic mean of the variables as shown in equation 7.

Journal of Applied Business and Economics vol. 16(3) 2014 25 = + + + + + /6 (7)

퐷퐽 Where:푈퐹푆퐼 UFSI� 퐼푌푆푝=the US퐿푛퐷퐽 financial퐿푛푈푇푊퐼 stress index;푅푆푝 IYSp퐶푀퐴푋 = inverted푇퐸퐷 yield� spread; LnDJ= the volatility for the natural logarithm of the Dow Jones; LnUTWI= the volatility for the natural logarithm of the US trade weighted index; RSp= the US-Canada risk spread; CMAX =the risk of equity market loss based on the Dow Jones share index and TED=the credit risk spread. DJ The resulting financial stress index is shown in figure 5. From the graph indicates financial stress during late 1998 and late 2000 when the index exceeds the 70th percentile. The indicated distress in Dec/98 may be as a due to contagion of the Asian currency crisis. Alternatively, the indicated distress may also be due to the contagion of 1998 Russian crisis that led to the controversial but necessary bail out Long-Term Capital Management, a hedge fund company that lost “$550 million on August 21 2000” (Jorion, 2000; Pinto & Ulatov, 2010). The 2000 crisis may be indicative of the high-tech or dot-com crash, whereby the value of the Nasdaq fell by 40 percent (Illing & Liu, 2006). Contrary to what was expected, the constructed index reports mild levels of stress during the GFC, with the 65 percent in Mar/09 being the highest level recorded. This is lower than the stress levels recorded in the Australian market.

FIGURE 5 A COMPOSITE FINANCIAL STRESS INDEX FOR UNITED STATES

A plausible explanation for this is overreaction of the Australian financial market to news of the subprime mortgage crisis. Unfortunately, there is no way of confirming the exact cause of stress in the different markets using empirical means.

GRANGER CAUSALITY RELATIONSHIP

The financial-stress indexes were assessed to determine if the US-financial stress index is useful for forecasting financial stress in Australia. Specifically, we sought to determine whether: past values of the estimated financial stress in USA can be used to estimate the level of financial stress in Australia? This

26 Journal of Applied Business and Economics vol. 16(3) 2014 Granger-causality relationship is of interest, given the historical episodes of financial stress crossing boarders (e.g. the GFC had global impacts) including those on developed countries (such as Australia). This question can be answered by performing Granger-causality tests on the estimated financial stress indexes (Granger, 1969). By definition, Granger-causality tests were designed for stationary series. Yang (2000) suggests that if two series are non-stationary, the first difference can be used to transform them to the covariance stationary form required in Granger-causality tests. Furthermore, it is important to assess the co- integrative relationship between two non-stationary series. Dakurah, Davies, and Sampath (2001) state that using two non-stationary series that are integrated of the same order yet are not co-integrated could yield spurious causality regressions—a problem that could be fixed by taking the first difference of each series. Otherwise, if two series are non-stationary and co-integrated then the causality tests should incorporate the co-integrated relationship between the two variables (Dakurah et al., 2001; Yang, 2000). Graphical representations of both index in figure 4 and 5 indicate that both series are non-stationary. Unit root tests were conducted using the OLS equation 3 and the results of ADF tests are shown in Table 6. At a 5% level of significance, both stress indexes were found to contain a unit root. Conversely, the first difference of each index is stationary; thus each series is integrated of order one.

TABLE 6 UNIT ROOT TESTS USING ADF TEST THE FINANCIAL STRESS INDEXES

Variables ADF test Level First difference -1.677 -21.683*** -2.658* -24.304*** The Dickey-Fuller퐴퐹푆퐼 tau statistic is significant at the 10% (*), 5% (**) or 1% (***) level. 푈퐹푆퐼

The co-integrated relationship between the Australian financial stress and the US financial stress can be evaluated by regressing UFSI on AFSI. The estimated OLS regression is:

= 37.384 + 0.256 (8)

The 푈퐹푆퐼�residuals푡 ( ) for equation 퐴퐹푆퐼8� follow푡 a random walk as shown in figure 6 whereby the residuals are integrated of the first order ( ~ 1(1)). Hence, the two stress indexes are not cointegrated. 푡 Accordingly, the first푒 difference of each series was used for the Granger-causality testing procedure. 푒푡

Journal of Applied Business and Economics vol. 16(3) 2014 27 FIGURE 6 LINE GRAPH OF RESIDUALS

Usually, the bi-variate VAR model (shown in equations 9a and 9b) is used to test for Granger- causality when both series are stationary (Granger, 1969).

= + + (9a) 푘 푘 푡 푖=1 1푖 푡−푖 푖=1 2푖 푡−푖 푡 퐴퐹푆퐼 = ∑ 훽 퐴퐹푆퐼 +∑ 훽 푈퐹푆퐼 +푢 (9b) 푘 푘 However,푈퐹푆퐼 푡given∑ 푖=1that훼 the1푖푈퐹푆퐼 two stress푡−푖 indexes∑푖=1 훽 2푖are퐴퐹푆퐼 integrated푡−푖 푣 of푡 order one and are not co-integrated, Toda and Yamamoto (1995) recommend the use of a VAR model of variables in the first difference form. Consequently, the Granger-causality equations applicable for this study are as shown in equations 10a and 10b.

= + + (10a) 푘 푘 푡 푖=1 1푖 푡−푖 푖=1 2푖 푡−푖 푡 ∆퐴퐹푆퐼 = ∑ 훽 ∆퐴퐹푆퐼 +∑ 훽 ∆푈퐹푆퐼 +푢 (10b) 푘 푘 As Granger∆푈퐹푆퐼푡-causality∑푖=1 훼tests1푖∆ 푈퐹푆퐼are sensitive푡−푖 ∑ 푖=1to the훽2푖 number∆퐴퐹푆퐼푡−푖 of lags푣푡 (k) included in the test, 5 tests were conducted to determine the appropriate number of lags to include in the VAR model. These five tests are the modified Likelihood Ratio (LR) test, the final prediction error (FPE), the Akaike Information Criterion (AIC), Schwarz Information Criterion (SIC) and the Hannan-Quinn Information Criterion (HQIC). All tests were conducted at a 5% level of significance and the results of the tests are included in Table 7.

28 Journal of Applied Business and Economics vol. 16(3) 2014 TABLE 7 VAR LAG SELECTION TESTS

Lag LR FPE AIC SIC HQIC 0 NA 1958.998 13.25594 13.28311 13.26686 1 53.74490 1642.373 13.07965 13.16114 13.11240 2 13.02638 1609.748 13.05958 13.19540 13.11416 3 7.290453 1612.900 13.06152 13.25167 13.13794 4 17.52852 1551.885 13.02293 13.26741 13.12118 5 4.138048 1573.830 13.03692 13.33574 13.15701 6 3.232229 1601.699 13.05441 13.40755 13.19633 7 3.636400 1627.275 13.07017 13.47764 13.23392 8 8.272690 1622.242 13.06696 13.52876 13.25254 9 6.864191 1626.189 13.06924 13.58537 13.27666 10 0.698703 1671.877 13.09678 13.66723 13.32603 11 3.182262 1701.177 13.11393 13.73872 13.36502 12 1.846032 1740.614 13.13660 13.81571 13.40952 Note: * indicates the lag length selected by the test criterion

The LR, FPE and AIC criterion recommend the inclusion of four lags while the SIC and the HQIC criterion recommend the use of one lag in the VAR framework. Accordingly, VAR models for Granger- causality were constructed using 4 lags; since most tests prefer this lag length. The results of the Granger- causality tests are given in Table 8.

TABLE 8 GRANGER-CAUSALITY TEST RESULTS (LAGS=4)

Null Hypothesis F-Statistic Probability ∆UFSI does not Granger Cause ∆AFSI 1.881 0.114 ∆AFSI does not Granger Cause ∆UFSI 1.844 0.121

The results in Table 8 show that we fail to reject that null hypothesis at a 5% level of significance in both cases. Hence, there exists no causal relationship between the first differences of financial stress in the Australia and US markets. Moreover, past information of financial stress in the US provides no useful information for forecasting financial stress in Australia and vice versa.

CONCLUSION

As expected, given their relative economic sizes, the Granger-causality tests indicate that financial stress in Australia does not Granger-cause financial stress in the US. Given the relative importance of US trade to Australia, It is very interesting that the reverse is also true—that financial stress in the US does not granger-cause financial stress in Australia. These tests do not exclude the possibility that contagion from US financial stress, during the 2007 GFC, flowed into and caused financial stress in Australia. Nor does it indicate that either country is immune from financial crises episodes that occur in the other country. The inconclusive Granger-causality findings in this study suggest that Australian policy makers

Journal of Applied Business and Economics vol. 16(3) 2014 29 cannot be confident in using information from US financial markets to predict financial stress and risk of a crisis in Australian markets, nor should they use that information fine-tune their policies. Given the shifting trade focus in Australia, it was not entirely surprising that the results for financial stress in the US Granger-causing financial stress in Australia, were inconclusive—what is surprising is the apparent speed of financial institutions in Australia adjusting to the US losing trade-dominance over Australia. Given that trade dominance rises and falls over time, it may be interesting for financial and economic historians to evaluate the change in Australia’s contagion risks with the US over period of decades and the implications for the rising trade-dominance of China for many countries (e.g. Australia, Latin American countries, and African countries). Future research could explore inclusion of other financial stress indicators not considered in this research—e.g. other variables may prove useful as proxies for factors not available on a monthly basis. Alternatively other future research might explore the effects of using quarterly data instead of monthly data. Also, future research might explore the use of historical data within Australia, to forecast future episodes of financial stress. Further, the effect of Australia’s stringent banking rules on financial contagion should be examined.

ENDNOTES

1. As observed by Mr. Micawber (a Dickens’s character):"Annual income twenty pounds, annual expenditure nineteen pounds nineteen and six, result happiness. Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery" (Dickens’s (1850, Chapter 12) 2. Trade theory is discussed in this study to provide essential background for contagion, as it relates to financial stress. An extensive discussion of trade theories is out of scope of this study and is well covered in the literature. 3. Feenstra and Hanson (1996) found that outsourcing contributed to an increase in wages of non-production labour rates by about 31 percent in the 1980s. 4. See the House of Representatives Standing Committee on Infrastructure Transport Regional Development & Local Government, 2009. 5. Eviews7: User Guide (2009), Irvine: Quantitative Micro Software. 6. Patel and Sarkar (1998) designed CMAX to index the share price levels to their maximum value over a review period that is expressed in years. 7. The null hypothesis for all ADF tests is that has a unit root. In which case would be equal to zero. The null is rejected if the series does not contain a unit root or is stationary. 푡 8. The ARCH Lagrange Multiplier tests for all 푦series indicate that no ARCH left휌 in the standardized residuals. Hence there is no need to estimate a GARCH (1, 2) model.

REFERENCES

Allen, F. & Gale, D. (2000). Financial Contagion, Journal of Political Economy, 108, (1), 1-33. Australian Bureau of Statistics. (2012). 2012 Year Book Australia. Canberra: Australia Bureau of Statistics. Balakrishnan, R., Danninger, S., Elekdag, S., & Tytell, I. (2009). The Transmission of Financial Stress from Advanced to Emerging Economies. IMF Working Paper. International Monetary Fund. Ball, R. (2009). The Global Financial Crisis and the Efficient Market Hypothesis: What Have We Learned? Journal of Applied Corporate Finance, 21, (4), 8-16. Barberis, N., Schleifer, A., & Vishny, R. (1998). A Model of Investor Sentiment. Journal of Financial Economics, 49, (3), 307-343. Bollerslev, T. (1986). Generalized Autoregressive Conditional Heteroskedasicity. Journal of Econometrics, 31, (3), 307-327. Bray, J. R. (2001). Occasional Paper no. 4: Hardship in Australia An Analysis of Financial Stress indicators in the 1998–99 Australian Bureau of Statistics Household Expenditure Survey. Canberra: Commonwealth of Australia.

30 Journal of Applied Business and Economics vol. 16(3) 2014 Breunig, R., & Cobb-Clark, D. (2006). Understanding The Factors Associated With Financial Stress In Australian Households. Australian Social Policy, 2005, 13-64. Calvo, G. A., & Mendoza, E. (1998). Rational Herd Behavior and the Globalization of Securities Markets. College Park, Maryland: University of Maryland. Caramazza, F., Ricci, L., & Salgado, R. (2004). International financial contagion in currency crises. Journal of International Money and Finance, 23, (1), 51-70. Cardarelli, R., Elekdag, S., & Lall, S. (2009). Financial Stress, Downturns, and Recoveries. International Monetary Fund. IMF Working paper WP/09/100. Washington D.C. Chan, W. S. (2001). Stock Price Reaction to News and No-News: Drift and Reversal After Headlines. NBER Working Paper Series. Accessed on 9 August 2012 from: http://jfe.rochester.edu/02207.pdf. Commonwealth Department of Family and Community Services. (2003). Occasional Paper no. 9: Inquiry into poverty and financial hardship. Canberra: Commonwealth of Australia. Corsetti, G., Pesenti, P., & Roubini, N. (1999). What Caused the Asian Currency and Financial Crisis? Japan and the World Economy, 11, (3), 305-373. Crotty, J. (2009). Structural Causes of the Global Financial Crisis: A Critical Assessment of the ‘New Financial Architecture’. Cambridge Journal of Economics, 33(4), 563-580. Dabrowski, M. (2010). The Global Financial Crisis: Lessons for European Integration. Economic Systems, 34, (1), 38-54. doi: 10.1016/j.ecosys.2010.01.002 Dakurah, A. H., Davies, S. P., & Sampath, R. K. (2001). Defense Spending and Economic Growth in Developing Countries: A Causality Analysis. Journal of Policy Modeling, 23, (6), 651-658. Daniel, K., Hirshleifer, D., & Subrahmanyam, A. (1998). Investor Psychology and Security Market Under- and Overreactions. Journal of Finance, 53, (6), 1839-1886. Daniel, K, Titman, S., & Wei, J. (2001). Explaining The Cross-Section Of Stock Returns In Japan: Factors Or Characteristics? Journal of Finance 56, (2), 743-766. Daniels, J. D., & Radebaugh, L. H. (1995). International Business: Environments and Operations (7 Ed.). United States of America: Addison-Wesley Publishing Company. De Long, J., Shleifer, A., Summers, L. & Waldmann, R. (1990). Noise Trader Risk in Financial Markets, Journal of Political Economy, 98, (4), 703-738. Demirguc-Kunt, A., & Detragiache, E. (1998). The Determinants of Banking Crises in Developing and Developed Countries. IMF Staff Papers (a quarterly journal of the IMF), 45, (1), 81-109. Demirguc-Kunt, A., & Detragiache, E. (2005). Cross-Country Empirical Studies of Systemic Bank Distress: A Survey. IMF Working paper WP/05/96. International Monetary Fund. Washington D. C. DFAT. (2011). Australia's Trade in Goods and Services by Top Ten Partners, 2010-11. downloaded from: http://www.dfat.gov.au/publications/tgs/FY2011_goods_ services_%20top_10_partners.pdf Dickens, C. (1850).The Personal History of David Copperfield. Bradbury & Evans, London. Dickey, D. A., & Fuller, W. A. (1979). Distribution of the estimators for autoregressive time series with a unit root. Journal of American Statistical Association, 74, 427-431. Dickey, D. A., & Fuller, W. A. (1981). Likelihood Ratio Statistics for Autoregressive Time Series With a Unit Root. Econometrica: Journal of the Econometric Society, 1057-1072. Dremen, D., & Lufkin, E. (2000). Investor Overreaction: Evidence That its Basis is Psychological. Journal of Psychology and Financial Markets, 1, (1), 61-75. Dornbusch, R., Fischer, S., & Samuelson, P. A. (1977). Comparative Advantage, Trade and Payments in Ricardian Model with a Continuum of Goods. American Economic Review, 67, (5), 823-839. Duca, M. L., & Peltonen, T. (2011). Macro-financial vulnerabilities and future financial stress - Assessing systemic risks and predicting systemic events. BOFIT Discussion Papers. Bank of Finland Institute for Economies in Transition. Helsinki. Dun & Bradstreet Credibility Corp. (2012). D&B's Financial Stress Score: Business Definition. Retrieved 10/04/2012, from https://smallbusiness.dnb.com/glossaries/d-bs-financial-stress- score/11814010-1.html#1.1

Journal of Applied Business and Economics vol. 16(3) 2014 31 Ehrhardt, M. C., & Brigham, E. F. (2011). Financial Management: Theory and Practice (13 ed.). United States of America: South-Western Cengage Learning. Fama, E., & French, K. (1993). Common Risk Factors in the Returns on Stocks and Bonds. Journal of Financial Economics, 33, (1), 3-56. Fama, E. F., & French, K. R. (1996). Multifactor Explanations of Asset Pricing Anomalies. The journal of Finance, 51, (1), 55-84. Fama, E. F. (1965). Random Walks in Stock Market Prices. Financial Analysts Journal, 21, (5), 55-59. Fama, E. (1970). Efficient Capital Markets: A Review of Theory and Empirical Work, Journal of Finance, 25, (2), 383-417. Fama, E. (1991). Efficient Capital Markets: II. Journal of Finance, 46, (5), 1575-1617. Federal Reserve Bank of St Louis. (2010). Measuring Financial Market Stress. National Economic Trends. Retrieved from Economic Research website: http://research.stlouisfed.org/publications/net/20100101/net_20100107.pdf. Feenstra, R. C., & Hanson, G. H. (1996). Globalization, Outsourcing, and Wage Inequality. The American Economic Review, 86, (2), 240-245. Fox, J. (2010). Are Finance Professors and Their Theories to Blame for the Financial Crisis? CFA Institute Conference Proceedings Quarterly, 27, (2), 4. French, K. (1980). Stock Returns and the Weekend Effect. Journal of Financial Economics, 8, (1), 55-69. Gibson, H. D., Hall, S. G., & Tavlas, G. S. (2011). The Greek Financial Crisis: Growing Imbalances And Sovereign Spreads. Journal of International Money and Finance, 31(3), 498–516. doi: 10.1016/j.jimonfin.2011.10.009. Gramlich, D., & Oet, M. V. (2011). The Structural Fragility of Financial Systems: Analysis and Modeling Implications for Early Warning Systems. The Journal of Risk Finance, 12, (4), 270-290. Granger, C. W. (1969). Investigating Causal Relations by Econometric Models and Cross-spectral Methods. Econometrica, 37, (August), 424-438. Gujarati, D. (2011). Economics by example. New York: Palgrave Macmillan. Hakkio, C. S., & Keeton, W. R. (2009). Financial Stress: What is it, How Can it be Measured, and Why Does It Matter? Economic Review, 94, (2), 5-50. Hammoudeh, S., Chen, L.-H., & Yuan, Y. (2011). Asymmetric convergence and risk shift in the TED spreads. North American Journal of Economics and Finance, 22, (3), 277-297. Hardy, D. C., & Pazarbasioglu, C. (1999). Determinants and Leading Indicators of Banking Crises: Further Evidence. IMF Staff Papers, 46, (3), 247-258. Hettihewa, S. and Mallik, G. (2005). Policy and Investment Opportunities Arising from Recent Findings on Asian-Pacific Stock-Market Integration. Journal of Applied Business and Economics, 5, (2). 36 – 52. Hettihewa, S. and Wright, C. (2010). Financial gender and Continuity Issues in Regional Small Businesses: Management and Policy Implications Derived from Canada and New Zealand Survey Evidence, Journal of Management Policy and practice, 11, (4). 29-47. Holló, D., Kremer, M., & Duca, M. L. (2012). CISS – A Composite Indicator of Systemic Stress in the Financial System. (Working Paper Series1426). European Central Bank. Frankfurt am Main, Germany. House of Representatives Standing Committee on Infrastructure Transport Regional Development & Local Government. (2009). The Global Financial Crisis and regional Australia. Canberra: Commonwealth of Australia. Illing, M., & Liu, Y. (2003). An Index of Financial Stress for Canada. (Working Paper 2003-14). Bank of Canada. Ottawa, Ontario. Illing, M., & Liu, Y. (2006). Measuring Financial Stress In A Developed Country: An Application To Canada. Journal of Financial Stability, 2, (3), 243 – 265. International Monetary Fund. (2008). Financial Stress and Economic Downturns, World Economic Outlook, October 2008: Financial Stress, Downturns and Recoveries (pp. 129-158). Washington D.C.: International Monetary Fund.

32 Journal of Applied Business and Economics vol. 16(3) 2014 Jorion, P. (2000). Risk Management Lessons from Long-Term Financial Management. European financial management, 6, (3), 277-300. Kaminsky, G. L., & Reinhart, C. M. (2000). On Crises, Contagion, and Confusion. Journal of International Economics, 51, (1), 145-168. Kindleberger, C. P., & Aliber, R. Z. (2005). Manias, Panics, and Crashes: A History of Financial Crises (5 ed.). New Jersey: John Wiley & Sons, Inc. Kishore, R., (2006). Theory of Behavioural Finance and its Application to Property Market: A Change in Paradigm. 12th Annual Pacific Rim Real Estate Society Conference, January 22-25, Auckland, New Zealand. Keynes, J. M. (1936). The General Theory of Employment, Interest and Money, Cambridge: Cambridge University Press (1972). Krugman, P. (2009, 6 September). How Did Economists Get It So Wrong? The New York Times. Retrieved from http://www.nytimes.com/2009/09/06/magazine/06Economic- t.html?_r=1&pagewanted=print. Lakonishok, J., & Smidt, S. (1988). Are Seasonal Anomalies Real? A Ninety-year Perspective. Review of Financial Studies, 1, (4), 403-425. Lee, C.-f., Shrestha, K., & Welch, R. L. (2007). Relationship between Treasury bills and Eurodollars: Theoretical and Empirical Analyses. Review of Quantitative Finance and Accounting, 28, (2), 163-185. Louzis, D. P., & Vouldis, A. T. (2011). A Financial Systemic Stress Index for Greece. First Conference of the Macro-prudential Research (MaRs) network of the European System of Central Banks. Retrieved from European Central Bank SSRN: http://ssrn.com/abstract=1736570 or dx.doi.org/10.2139/ssrn.1736570 Markowitz, H. (1952). Portfolio Selection. The journal of finance, 7, (1), 77-91. Marks, G. (2007). Income Poverty, Subjective Poverty and Financial Stress. Australian Government Social Policy Research Paper, (29). Mazumder, M. I., & Ahmad, N. (2010). Greed, Financial Innovation Or Laxity Of Regulation?: A Close Look Into The 2007-2009 Financial Crisis And Stock Market Volatility. Studies in Economics and Finance, 27, (2), 110-134. Misina, M., & Tkacz, G. ( 2009). Credit, Asset Prices, and Financial Stress. International Journal of Central Banking, 5, (4), 95-122. Ng, S., & Perron, P. (2001). Lag Length Selection And The Construction Of Unit Root Tests With Good Size And Power. Econometrica, 69, (6), 1519–1554. Oxfam. (2002). Rigged Rules and Double Standards: Trade, Globalisation and The Fight Against Poverty. Oxford: Oxfam. Patel, S., & Sarkar, A. (1998). Crises in Developed and Emerging Stock Markets. Financial Analysts Journal, 54, (6), 50–61. Perlich, H. (2009). The Impact of the GFC on Australia as a 'Dual Economy’. Journal of Australian Political Economy, December 2009(64), 65-90. Pinto, B., & Ulatov, S. (2010). Financial Globalization and the Russian Crisis of 1998. Policy Research Working Paper 5312. The World Bank. Robert, F. E., & Bollerslev, T. (1986). Modelling the Persistence of Conditional Variances. Econometric Reviews, 5, (1), 1-50. Ross, S. A. (1976). The Arbitrage Theory of Capital Asset Pricing. Journal of Economic Theory, 13, (3), 341-360. Sharpe, W. (1964). Capital Asset Prices: A Theory of Market Equilibrium under Conditions of Risk. Journal of Finance, 19, (3), 425-442. Shleifer, A. (2000). Inefficient Markets: An Introduction to Behavioural Finance. Oxford: Oxford University Press. Shleifer, A. & Vishny, R. (1997). The Limits to Arbitrage. Journal of Finance, 52(1), 35-55. Siegel, J. J. (2009). Efficient Market Theory and The Crisis. The Wall Street Journal, A23.

Journal of Applied Business and Economics vol. 16(3) 2014 33 Stiglitz, J. E. (2010). Risk and Global Economic Architecture: Why Full Financial Integration May Be Undesirable. American Economic Review, 100, (2), 388–392. Stiglitz, J. E., & Charlton, A. (2007). Fair Trade for All: How Trade Can Promote Development. Oxford, NY: Oxford University Press Tversky. A. & Kahneman. (1979). Prospect Theory: An Analysis of Decision under Risk, Econometrica, 47, (2), pp. 263-291. Toda, H. Y., & Yamamoto, T. (1995). Statistical Inference In Vector Autoregressions With Possibly Integrated Processes. Journal of Econometrics, 66, (1), 225-250. United Nations Conference on Trade and Development. (2008). Globalization for Development: the international trade perspective. Geneva: United Nations. Wesley Mission. (2006). Financial Stress and its impact on the individual, family and the community. Retrieved from http://www.wesleymission.org.au/news/ publications/finstress/financial_stress_report.pdf Wolf, M. (2009). How Mistaken Ideas Helped to Bring the Economy Down. Financial Times. Worthington, A. C. (2006). Debt as A Source of Financial Stress in Australian Households. International Journal of Consumer Studies, 30(1), 2-15. Yang, H.-Y. (2000). A Note on The Causal Relationship Between Energy and GDP in Taiwan. Energy Economics, 22(3), 309-317.

34 Journal of Applied Business and Economics vol. 16(3) 2014

Impact of Hedge Funds on Traditional Investment Products

Kaouther Flifel Institut des Hautes Etudes Commerciales (IHEC-Carthage-Tunisia)

INTRODUCTION

Hedge funds are now one of the most popular in financial circles and in anything related to the economy in general. Bear markets are not strangers to this, because hedge funds have absolute return as target, that is to say that managers of such funds seek to provide a positive return, regardless of changes in the market as a whole. To achieve this, managers implement investment strategies different to the classic one. Most of them specialize in very specific niches such as convertible bonds, distressed companies or restructuring, macroeconomic movements or global emerging markets and they often use derivative products, they are usually paid on the basis of commission related to the performance. One feature of most hedge funds highlighted is their low correlation with traditional investment products such as stocks, bonds and raw materials. Several studies have been made to clarify this relationship, such as Ackermann et al. (1999), Agarwal and Naik (1999), Liang (1999), Edwards and Caglayan (2000), Amin and Kat (2001), Capocci and Hübner (2003) who put this forward. Except that their work is often limited to analyze the correlation between hedge funds, stocks and bonds, without considering other investment products such as corporate bonds, commodities or stocks, it's actually the first point deals with the study, in fact, we will analyze the correlation between hedge funds (taken together, by investment strategy and individually), actions, bonds and raw materials to determine the level of correlation, its evolution and its stability time. The low correlation would be one of the key benefits of hedge funds because if one refers to the modern portfolio theory of Markowitz (1952), the low correlation of an asset class with traditional media investment allows investors to achieve efficient frontier is more interesting to say that the investor gets a better return for the same risk or lower risk for a similar yield and this allows to build a diversified portfolio. To study the correlation recent research has shown that the characteristics of hedge fund returns are more complex than those of stocks and bonds. Thus studies have been conducted to propose a methodology that represents an alternative to the standard analysis. Indeed, this methodology constructs

Journal of Applied Business and Economics vol. 16(3) 2014 35 and rebalances portfolios composed of stocks, bonds and commodities. Grauer and Hakansson (1995) applied the dynamic investment model in discrete time to show the gain diversification through Real Estate (real estate). Thus, these authors were able to demonstrate a significant gain due to diversification through real estate. However, one question remains, no study has confirmed or denied: the properties of diversification offered by hedge fund strategies are they different from traditional assets? ... So the second purpose of this paper is to study the difference between diversification offered through a general index of hedge funds relative to diversification through a set of indices of hedge fund strategies. This article represents an extension of the literature on the correlation and diversification across hedge fund strategies. This study is evaluating the gain of diversification across hedge funds over a number of traditional assets using the dynamic investment model in discrete time of Grauer and Hakansson (1985, 1986, 1987, 1995, 2001) combined with the empirical probability assessment approach (EPAA). Indeed in this article there will be a further analysis of Hagelin and Promberger (2003) by examining the returns of a portfolio of different classes of shares and indices of hedge fund strategies. This paper is structured as follows: after a review of the literature, we conduct an analysis by grouping funds by investment strategy in composite indices and recitals individually. We also undertake an analysis of different sub-periods to check the consistency of previous results, then we will present data and the assumptions and methodology that is the model of discrete-time dynamic investment approach combined with evaluation of the empirical probability, the third part presents the results obtained from evaluating the performance of diversified portfolios containing equity indices, bonds and commodities with and without hedge funds to finally conclude.

LITERATURE REVIEW

Hedge funds are now the subject of numerous studies. It seems interesting to make an inventory of the main studies. Ackermann and Ravenscraft (1998) highlight the fact that the more stringent restrictions on mutual funds than for hedge funds affect their performance. Ackermann et al (1999) and Liang (2001), who compared the performance of hedge funds than mutual funds and indices, indicated that hedge funds have consistently achieved better performance than mutual funds but was not better than the indices used. Amenc and Martellini (2002) have shown that the inclusion of hedge funds in a portfolio can significantly improve the risk-return trade-off by their very nature limited correlation with other securities. This low correlation was also highlighted by Liang (1999) and Agarwal and Naik (2000a). According to Brown et al. (2001), hedge funds that offer good performance in the first part of the year reduce the volatility in the second half of the year. They provide the annual performance of sorts. Capocci and Hübner (2003) show the ability of some hedge funds to outperform the market over time. Point seen diversification of academic research has focused on the sources of hedge fund returns. Other studies have been done on the performance of hedge funds without taking into account the differentiation factors, styles and characteristics associated with abnormal returns of these funds. Brown, Goetzmann and Ibbotson (1999) studied a sample of offshore hedge funds between 1989 and 1995, and found a positive risk-adjusted. Their results did not support the hypothesis of the effect of the difference of talent managers, nor the persistence of the performance of some managers. Unlike Brown, Goetzmann and Ibbotson (1999), Ackermann, Mc Enally and Ravenscraft (1999) observed that, despite the fact that hedge funds have outperformed mutual funds in the period 1988-1995, they have not performed on the index of market standard in terms of yield risk over the same period, this is mostly due to poor performance of hedge funds in 1994 and 1995. They studied the commissions offered to managers, as a percentage of the gain beyond a break-even ("High Water Marks") and found a significant relationship between commissions and profitability. In their combined sample from the two databases, Ackermann, Mc Enally and Ravenscraft (1999) have included data on funds that have disappeared and whose omission could bias the study to eliminate survivorship bias "Survivorship bias ". Ackermann, Mc Enally and Ravenscraft (1999) support the idea that survivorship bias erroneous results as suggested by other studies. Indeed, they believe that another

36 Journal of Applied Business and Economics vol. 16(3) 2014 source of bias against balance survivorship bias caused by the hedge funds that perform well and which fail to disclose their results. This self-selection bias (self-selection bias) may be due to the desire of certain funds to remain discreet and maintain confidentiality of their strategy and their yields. Although the sample of Ackermann, Mc Enally and Ravenscraft (1999) is larger than that of Brown, Goetzmann and Ibbotson (1999), they did not take into account the data for the period before 1999 that experienced the Asian crisis in 1997-1998 and the collapse of the LTCM fund in 1998 was highly publicized. Their studies have focused on several questions about different periods and with a larger sample, and their results contradict most existing studies replies that hedge funds outperform traditional benchmarks. On the other hand, Brown, and Goetzma Ibbotson (1999) and Agarwal and Naik (1999) reach opposite conclusions with respect to the issue of performance persistence of hedge funds. Indeed, Brown, Goetzmann and Ibbotson (1999) observed a lack of persistence in performance, while Agarwal and Naik (1999) have shown the existence of a form of persistence of performance using data from the database HFR (Hedge Fund Research), they concluded that the persistence is less observed when using a multifactor model than when using a two-period model. They also showed that more the range of returns increases, the most the of persistence decreases. They attributed the persistence observed in the intervals and in the two-period model, not the losers that continue to be losers, only winners who continue to be winners. Hagelin and Promberger (2003) have taken the model used by Grauer and Hakansson (1985, 1986, 1987, 1995and 2001) to study a portfolio of stocks and bonds with and without hedge funds. For this they used as a proxy for hedge fund index "HFR Fund Weighted Composite index" of the database Hedge Fund Research (HFR), then an index fund of funds (HFR fund of funds index) of same basis. These authors showed an increase in the average profitability of the portfolio following the addition of the index HFR index without increasing the variance of the portfolio. Unlike the first result, diversification through an index fund of funds does not show a gain in terms of profitability, variance, the authors explain this result across the board double as support funds of funds and the problem of bias faced by hedge fund data. Harry M. Kat (2005) in her article, shows how investors can neutralize the unwanted skewness and kurtosis effects from investing in hedge funds by 1) purchasing out-of-the-money equity puts, 2) investing in managed futures, and/or by 3) overweighting equity market neutral and global macro and avoiding distressed securities and emerging market funds. The analysis suggests that all three alternatives are up to the job but also come with their own specific price tag. Bacmann, Jeanneret and Scholz (2008) measure the relationship between two time series from a static perspective. They introduce a methodology to measure the drivers of correlations between the returns of a dynamic strategy (hedge funds) and those of a static benchmark (traditional investments). Their approach explains why correlations between hedge funds and traditional investments are sometimes high and at other times low. Furthermore they highlight which elements of hedge funds strategies are responsible for making correlations vary over time. It constitutes an improvement in measuring the diversification potential of hedge funds to traditional investments. Mikael Haglund (2010), in her article, uses higher moment betas to examine the effects on portfolio volatility, skewness and kurtosis when hedge funds are added to an equity portfolio. The results show that hedge funds, in general, can lower the volatility, skewness and kurtosis of the portfolio but large variations are seen between different hedge funds strategies. Convertible Arbitrage, Equity Market Neutral, Fixed Income Arbitrage, Merger Arbitrage and Macro are identified as the most attractive strategies to include in an equity portfolio for investors who care about higher moment risks and want to limit downside risk. In this study, we deepen the analysis of Bacmann, Jeanneret and Scholz (2008) and Promberger and Hagelin (2003) by examining the returns of a portfolio of different classes of shares and indices of hedge fund strategies. Our contribution to the literature is twofold: First, we wish to highlight the existence of a gain diversification through a general index, but also through a set of indices of hedge fund strategies, in a second time, we evaluate the contribution of each hedge fund strategies portfolio of traditional assets. Three types of portfolios are studied: a portfolio of traditional assets (stocks and an index of real estate), a portfolio of traditional assets and hedge funds a general index and a portfolio of

Journal of Applied Business and Economics vol. 16(3) 2014 37 traditional assets and indices of hedge fund strategies. The data used in their articles are retrieved from the database “Hedge indexes”, these data have the advantage of allowing to overcome the survivorship bias (Survivorship bias) since they included data on age or missing in funds during the study period.

HEDGE FUNDS: PERFORMANCE UNCORRELATED SOURCE FOR INVESTORS AND DIVERSIFICATION TOOL

Data and Descriptive Statistics Prior to the analysis of correlations, we proceed to the descriptive analysis of our database. In our study, we use data on the strategies from the database “Hedge indexes”. We have net monthly returns of individual funds. The data cover the period from January 1994 to December 2010. We also have monthly returns of various equity indices, bonds and commodities. These indices are:

Indices actions : Indices obligations : Indices matières premières :  Le S&P 500  JPM chase L Goldman Sachs Commodity  Le MSCI  Le LH Lehman

The first table in appendix 1 reports the descriptive statistics of strategies1 and indices. The table shows that the best monthly performance returns to the LES strategy with a monthly average performance of 1.1% followed by D EV strategy with 1.09% then the GM strategy with 1.03%. Strategies and sub strategies that have the worst monthly returns are DSB with -0.04%, MF with 0.44% and FIA with a monthly average performance of 0.51%. When we consider also the risk according to the Sharpe ratio, the strategies which are distinguished are the EMN strategy with a monthly ratio of 2.161291, followed by RA EV strategy with 1.529584 and FIA strategy with 1.475149. The strategy MA Sharpe ratio lowest i.e. 0.02. Taking into account the t- statistics, the majority of monthly returns are significantly different from zero at the 5%, except for the strategy MF and DSB. The second part of the table reports the descriptive statistics of the indices. It indicates that the highest average performance is offered by the LH index. The other strategies are getting the lowest performance throughout the period. The t-statistics indicate that all indices (except JPM and GD) provide returns which are significantly different from zero at the 1%.

Analysis of Correlations Between Strategies and Indices We begin the analysis by studying the correlation between hedge funds (grouped by strategy) and the benchmarks used. Subsequently, we analyze the correlations between hedge funds; we will complete the study with an analysis of correlations between hedge funds and conventional indices over a rolling period. To calculate the correlation, we rely on a traditional measure which is the correlation coefficient. Recall that the value of the correlation coefficients varies between -1 and +1, and when the coefficient is near to 1, the underlying media follows the same trend, this coefficient is given by the following formula:

( , ) with ρ: correlation coefficient between p and q = σ (p, q) : covariance between p and q ×p q σ σ p and σ q : respectively the standard deviation of p and q ρ σp σq Correlation Between Hedge Funds Pooled by Strategies and Indices The results of the analysis of correlations between hedge funds (grouped by investment strategy) and the market indices used are shown in the table 2 of the first appendix: The idea behind this first analysis is to know if each strategy can be significantly correlated with one or more indices. For this, we gather in the same matrix, the correlation of different strategies and sub-

38 Journal of Applied Business and Economics vol. 16(3) 2014 strategies. First we analyze the figures in table horizontally (as in strategy and strategy) and then vertically (relation between indices and strategies). Strategies CA, EM, EMN, EV, DEV, MS EV, RA EV, GM, and MS are all significantly positively correlated with all of the equity indices. Moreover, they are not significantly correlated with the index LH. This shows that these strategies are essentially based on equities, products linked equities or high yield bonds. This is easily explained by event strategies investing in high yield bonds and stocks, but it is surprising that macro funds are not more strongly correlated to the bond market. In addition, these funds are generally diversified in some market and invest in different products, which justifies that all correlation coefficients are between -0.8 and 0.6. The strategy LES is significantly correlated with the equity indices at the 1%, reflecting the fact that managers combine long and short positions in equities generally without recourse to non-linear derivative products. The GM strategy is significantly correlated with all equity indices. This is due to its international aspect generally invested in other continents as North America. This strategy is also weakly correlated with the bond indices. The general indices CS was significantly and positively correlated with all indices except for the classic LH index where it is negatively correlated. The strategy DSB is significantly correlated with the LH index. EMN strategy is positively correlated with all indices. If we analyze the table vertically, we find that the correlations are very different from one strategy to another. If we do not consider the strategy DSB highly decorrelated, the strategies are all positively correlated with all indices (with some exceptions). There are minimal negative correlations especially in the case of the LH index. For the bond market, the correlation coefficients also vary greatly. They ranged from 0.440 for the strategy EV significant at 5% to -0.4203 but not significant for the strategy DSB. For the commodities market they range from -0.0766 to MS EV strategy, to -0.077 for the strategy DSB. The three strategies in D EV, MS EV and RA EV are correlated with the same ratings as the mother strategy. In this first analysis, we conclude that there are two types of strategies, those that can be grouped into families and others cannot be combined. The First Family (LES, RA EV, EV MS, D EV, EV, EM and CA) includes strategies all significantly related to equity indices, JPM, and GD. The second family (FIA, MF and MS) includes strategies significantly uncorrelated with all indices. The strategy DSB is itself due to its low correlation with all markets used. Other strategies are ultimately to be significantly correlated with virtually all indices (MNEs) or correlated with some indices close to the investment products they prefer.

Correlation Between the Strategies of Hedge Funds Table 3 in appendix 1 reports the correlation coefficients of alternative strategies. We notice a significant difference between the observed coefficients, ranging from a coefficient of 0.741 (significantly different from zero at the 1% level) between strategies DEV and EV, to a coefficient of -0.541 (significantly negative at the 1%) between the strategies LES and DSB. Note that the strategy DSB is negatively correlated with all other strategies.

Correlation Analysis Strategies with Clues Rolling Period Before concluding the analysis of correlation, it seems interesting to conduct a final analysis. We will examine the correlations of hedge fund strategies with benchmarks on various sub-periods in order to determine whether there is constant correlation coefficient over time. Then we will focus on various periods of crisis before determining if a pattern emerges from numbers. From our global analysis period of 192 months (January 1994-December 2010), it emerges eight rolling periods which are:

Journal of Applied Business and Economics vol. 16(3) 2014 39 - January 1994 to December 1995 - January 1996 to December 1997 - January 1998 to December 1999 - January 2000 to December 2001 - January 2002 to December 2003 - January 2004 to December 2005 - January 2006 to December 2007 - January 2008 to December 2010

Based on numbers in appendix 2 we constructed eight correlation matrices (one per period). These eight correlation matrices report specific outcomes by strategy and classic. They indicate that globally:  The correlation between a strategy and hedge fund and an index can vary significantly over time (ex the correlation between CA and the SP index decreased from 0.398, significantly different from zero on the first sub-period to 0.094 significant in the second before increasing during periods the other periods and finally be significantly different from zero);  The correlation of all indices with strategies typically progresses in the same direction over the different sub-periods analyzed (ex increase in overall correlations of the first to the second sub- period);  Correlations are less pronounced on the first sub-period as they are on the other seven periods analyzed; Note that the FIA strategy is not always significantly correlated with classic indices over the eight sub-periods, indicating a source of diversification for investors. The other strategies are significantly correlated with equity markets and then only in some cases.

Let's look to economic events occurring during these periods to see if we can make a comparison between the rolling periods where the coefficients are highest, and periods of particular market. Among the economic events during the analysis, we note the increase in American interest rates that occurred between February and April 1994. The performance of each strategies and bonds are summarized in the table 4 of appendex1. There is also the Asian crisis between August and October 1997, occurring during the second period, where, again, the coefficients were highest. The Russian crisis (July-September 1998) involved in the third period of analysis. During this period, the correlation between the indices and most of the hedge fund index is relatively high. In the following we will try to make the connection between having a certain degree of decorrelation between the different investment products traded and the advantage to be gained in the training and diversification of portfolios.

METHODOLOGY AND HYPOTHESIS OF THE DYNAMIC MODEL OF INVESTMENT

Presentation of Model The ability to diversify a standard portfolio is in fact the objective of this part where we build portfolios to assess the potential diversification of these strategies; this study will be made using the dynamic investment model in discrete time combined with the approach for evaluating the empirical probability. The dynamic model of investment used in this paper assumes that the investor has a utility function with iso-elastic form:

u(1 + r) = (1 + r) (1) 1 γ γ

40 Journal of Applied Business and Economics vol. 16(3) 2014 Where r is the return of a period and γ is a parameter of risk aversion. (γ = 1 is a neutral risk). This iso-elastic function assumes that the coefficient of relative risk aversion (RRA) is independent of wealth and that the coefficient of risk aversion γ must be less than 1 (γ <1) for that the coefficient of absolute risk aversion cannot increase with wealth. A simple calculation shows that the coefficient of relative aversion towards risk RRA is equal to the constant (1 - γ). This function is a special case when γ tends to 0; in fact, it is a logarithmic function of the form:

ln u (1 + r) = ln (1+r) (2)

The use of the power utility function or logarithmic model in a multi periodic involves solving a nonlinear optimization problem with constraint. For this the investor must choose a level of risk aversion through the parameter γ in equation (1). In our study γ takes the following values: -50; -30; -20; -15; -10; -5; -3; -1; -0.5; +0.5; and +1. These values cover respectively investors the most averse (γ = -50) to risk neutral (γ = 1). At the beginning of each period t, the investor chooses a portfolio Φt based on level of risk aversion and the set of constraints. This comes down to solve the optimization problem in each period t:

( ) ( ) max E (1 + r ) = max (1 + r ) (3) 1 γ 1 γ t t Φ �γ t Φt � Φ ∑ πts γ t Φt Φit ≥ 0, for all i (4) et Σ Φit = 1 (5)

with:  Rits (Φt): The ex-ante profitability of period t if state s occurs, with: r ( ) = r  γ: The parameter of risk aversion that remains fixed over time ts t i it its  Φit: The proportion of wealth invested in the risky asset i at time t Φ ∑ Φ  Φt: The weight vector  Rit: The anticipated total cost of asset i at time t  π : The probability of state s at the end of period t ts

The constraint (4) eliminates the possibility of short selling while the constraint (5) represents the budget constraint. The value of the inputs of the model is based on an estimation method explained below. Using the weights chosen for each asset class at the beginning, the profitability of the portfolio performed in month t is selected and recorded. This cycle is repeated for all months. The application of the model discussed above poses the problem of estimating the distribution of returns. We apply an estimation procedure used by Grauer and Hakansson (1985, 1986, 1987, 1995 and 2001) called the empirical probability assessment approach (EPAA). With this approach an estimation window of T sub-periods is formed. An estimation window of 24 periods is composed (T = 24), the vector of realized returns for 24 months is calculated, rt-j (j = 1, ..., 24). Each monthly return from this window of estimation has a weight of . Thus, the estimation of the profitability of the period T is: 1 T 1 E [r ] = r 24 24 t−1 t t−j �j=1 Therefore, with this approach for evaluating the empirical probability (EPAA), the estimated returns are obtained on a mobile base and used in a raw form without any adjustment. Another advantage of this approach is the use of the utility function that requires the specification of the entire return distribution used so that there is no loss of information, all periods and all correlations are taken into account.

Journal of Applied Business and Economics vol. 16(3) 2014 41 The returns are calculated for different asset classes independently of each other. This management program is applied to the period of January 1994 to December 2010. For each level of risk aversion, and for each portfolio, the optimal compositions are calculated every year. The out-of sample returns of the portfolios are then calculated by applying the weights obtained in the returns actually observed. For each portfolio and each level of risk aversion, we thus obtain a distribution of returns out-of sample. These distributions of returns allow us to compare the performance of different portfolios. In our study, we examine three portfolios with different compositions: 1) Portfolio of traditional assets (not including Hedge Fund assets) noted later P1. 2) Portfolio of traditional assets with hedge fund indices generally noted after P2. 3) Traditional asset portfolio with the ten indices of hedge fund strategies noted later P3.

For each level of risk aversion (γ), we calculated the optimal portfolio composition. Remember that the optimal portfolio is the portfolio that maximizes expected utility on the window of 24-month study as specified in the EPAA. The parameters of the different assets are re-estimated every month; our portfolio is reconstructed at the same frequency. We then applied the optimal composition of the period (January 1994-December 2010). We can analyze the performance of the optimal portfolio for each level of given risk aversion. We present, in the following paragraphs: the synthesis of the optimization program.

Composition of Optimal Portfolios The tables in appendix 3 represent the average proportions invested in each asset for our three portfolios depending on the level of risk aversion. They also show the variability in proportions of one month to another of the study period. The proportions shown represent the averages of optimal proportions that satisfy the objective of maximizing the expected utility of the investor on a sliding window of 24 months for the three portfolios. The variance presented in the three tables shows the variability of the proportions invested in each asset. These tables show that investors are more risk averse, the more the proportions of the capital invested in hedge funds are more important. This proportion is very important for high levels of risk aversion (up to 15% and 65% for P2 P3). We also notice that for a given level of risk aversion for portfolios P2 and P3, the proportion of capital that is invested in hedge funds is more important in the portfolio P3 than P2; this could be explained by the diversity statistics between different hedge fund strategies. For further clarification of results we will proceed in two stages: First, we study the returns and the Sharpe ratio of the optimal portfolio obtained, we can compare these ratios to see the efficiency of the portfolio from the perspective of conventional finance. In a second step, we will proceed to study the performance of optimal portfolios using the semi-variance and the ratio E (R) / SVM to see if the results are always the same.

Comparison of Returns and Sharpe Ratio of Optimal Portfolios Table 1 in appendix n°4 summarizes the average of returns in each optimal portfolio: P1, P2 and P3 for every level of risk aversion and volatility. The observation of the previous table shows that the returns are decreasing with the increase in risk aversion of the investor. Regarding the volatility of optimal portfolios, we observe that the standard deviation of portfolio that does not include hedge funds is more important than the other wallets for the same level of risk aversion. For the same level of risk aversion, the lowest variance is observed for the portfolio containing both traditional assets and hedge fund strategies. Comparing the Sharpe ratios of the three optimal portfolios shows that the optimal portfolio that contains traditional assets and the strategies of hedge funds outperforms the other two portfolios in high level of risk aversion (with some exceptions).

42 Journal of Applied Business and Economics vol. 16(3) 2014 Comparison of the Average / Semi-Variance Optimal Portfolios Table 2 in appendix n°4 presents a summary of the analysis of the performance of optimal portfolios through the semi-variance. The comparison of the semi-variance in the portfolios P1 and P2 indicates that P2 realizes for all levels of risk aversion, a lower semi-variance than P1. The diversified portfolio of assets in traditional indices and hedge fund strategies (P3) has the lowest semi-variance for each level of risk aversion, compared to portfolios P1 and P2. The semi-variance criterion does not allow us to conclude on the performance of portfolios or classify them. Indeed, for the same level of risk aversion, the portfolios, which have the lowest semi-variance, also hold the lowest average return. For this reason we present the ratio E(R)/SVM that takes into account these two dimensions namely the average of return and the semi- variance SVM. The observation of the ratio E (R) / SVM in the optimal portfolio containing the traditional assets and indices of hedge fund strategies shows that this ratio is consistently higher in P3 than for portfolios P1 and P2 at all levels of risk aversion. We can conclude after this result that the portfolio P3 is better than P1 and P2. So a diversified portfolio with both traditional assets and hedge fund strategies is more effective than a diversified portfolio only in the traditional asset or in classic assets with the general index of hedge fund. This finding confirms our hypothesis of the gain of diversification brought by the indices of hedge fund strategies to a portfolio of traditional assets.

CONCLUSION

Using our database we studied, initially, the correlation between the investment strategies used by hedge fund managers with various traditional benchmarks. We conducted an analysis by strategy, sub- strategy and considering the funds individually. This study was not only to analyze the correlation of hedge funds with equity indices, bonds and commodities as a whole, as had been able to do Edwards and Caglayan (2000) for example, but rather to see which indices each strategy could be linked. In addition, we analyzed the entire period for which data focused, then we looked at eight sub periods. The results allow the following conclusions: - The majority of sub-strategies are correlated with the same ratings as their strategies of belonging. - There is a significant difference between the strategies correlations with indices, and those of the individual funds with the same indices. The second is much lower; the investor may not be based on aggregate results for the hedge fund industry or in any strategies before making his investment; - The correlation of hedge funds with conventional indices tends to increase during periods of instability or crisis. The hedge funds then lose some of their advantage of decorrelation. In some cases they may even be highly correlated with classic indices.

The results show the complexity of the hedge fund industry. Hedge funds are not a homogeneous class of assets, which are thought usually to have low correlation with financial markets. In the second part of the study and to compare the contribution of hedge funds in a diversified portfolio, we analyzed the ex-post performance of managed portfolios that offer maximum expected utility of the 24 months preceding the choice portfolio, here based on historical asset returns. Indeed, the use of an iso-elastic utility function in the dynamic model of investment replaces the traditional portfolio optimization by the mean-variance, since recent studies show that hedge fund returns are not normally distributed. In addition, the EPAA enables us to incorporate all the information and the moments of the distribution of returns. Our study has shown the gain offered by the addition of hedge fund strategies to a portfolio of stocks and traditional assets, this gain is more significant when we use a set of indices of hedge fund strategies instead a general index of hedge funds. Indeed, the diversity indices of hedge fund strategies offer a better choice for investors.

Journal of Applied Business and Economics vol. 16(3) 2014 43 In this research, some issues have not been extensive, including the problem of Skewness and Kurtosis. Indeed, Kat, Lu and Davies (2006) point out that the use of a utility function represents an alternative approach to optimize a portfolio containing hedge funds and thus circumvent the problem of using the mean-variance optimization. According to these authors, this utility function has the three following properties: (1) non-satiation (2) risk aversion (3) risky assets are inferior goods. By cons, this function does not provide an exact order of preference of portfolio risk using the three first moments of the portfolio's return. Kat, Lu and Davies (2006) then propose to isolate the effect of every moment. An interesting extension of our study would take into account the work of Kat, Lu and Davies (2006) which proposes the use of an optimization model mean-variance-Skewness-Kurtosis.

ENDNOTES

1. CS: Credit Suisse/Tremont Hedge Fund Index CA: Convertible Arbitrage DSB: Dedicated Short Bias EM: Emerging Markets EMN: Equity Market Neutral EV: Event Driven D EV: Distressed MS EV: Multi-Strategy RA EV: Risk Arbitrage FIA: Fixed Income Arbitrage GM: Global Macro LES: Long/Short Equity MF: Managed Futures MS: Multi-Strategy

REFERENCES

Amenc and Martellini (2003) “The Brave New World of Hedge Funds Indices”, EDHEC Risk & Asset Management Research Centre, Working Paper. Amin G.S., and Kat H.M. (2003), “Welcome to the Dark Side: Hedge Fund Attrition and Survivorship Bias over the Period 1994-2001”, Journal of Alternative Investments, 6 (1), pp. 57-73. Bacmann, Jeanneret and Scholz (2008) “What correlation does not tell you about hedge funds: A factor approach to hedge fund correlations” Journal of Derivatives & Hedge Funds 14, 90–101. Brown S.J., Goetzmann W.N. and Ibbotson R.G. (1999), “Offshore Hedge Funds: Survival and Performance 1989-1995”, Journal of Business, 72 (1), pp. 91-117. Caglayan M.O., Edwards F.R. (2001), “Hedge Fund and Commodity Fund Investments in Bull and Bear Markets”, Journal of Portfolio Management, 27 (4), pp. 97-108. Capocci D.P.J., Corhay, A., and Hübner G. (2004), “Hedge Fund Performance and Persistence in Bull and Bear Markets”, Working Paper. Davies, R. J., H.M. Kat and S. Lu, (2004), “Fund of Hedge Funds Portfolio Selection: A Multiple- Objective Approach”, working paper, Cass Business School. Grauer, R.R. and N.H. Hakansson, (1985a), Returns on Levered, Actively Managed Long-Run Portfolios of Stocks, Bonds and Bills, 1934-1983, Financial Analysts Journal, 24-43. Grauer, R.R. and N.H. Hakansson, (1986b), “A half century of returns on levered and unlevered portfolios of stocks”, bonds and bills, 1934-1984, Journal of Business 59, 387-318. Grauer, R.R and N.G. Hakansson, (2001e), “Applying portfolio change and conditional performance measures: The case of industry rotation via the dynamic investment model”, Review of Quantitative Finance and Accounting. Hagelin, N. and B. Pramborg, (2003), “Evaluating gains from diversifying into hedge funds using dynamic investment strategies”, Barry Schachter (ed.), Intelligent Hedge Fund Investing, Risk Waters Group Ltd., London, 423–445. Harry M. Kat (2005), “Integrating Hedge Funds into the Traditional Portfolio” The Journal of Wealth Management, Vol. 7, No. 4: pp. 51-57. Kat, H M. and S. Lu, (2002), “An excursion into the statistical properties of hedge fund returns, working paper”, Cass Business School. Liang, B., (2000), “Hedge funds: The living and the dead, Journal of Financial and Quantitative Analysis” 35, 309-326.

44 Journal of Applied Business and Economics vol. 16(3) 2014 Liang N. (2003), “On the Performance of Alternative Investments: CTAs Hedge Funds and Funds of Funds”, University of Massachusetts at Amherst, Working Paper. Liew J. (2003), “Hedge Fund Index Investing Examined”, Journal of Portfolio Management, 29 (2), pp. 113- 123. Mikael Haglund (2010) “Higher moment diversification benefits of hedge fund strategy allocation” Journal of Derivatives & Hedge Funds 16, 53–69.

APPENDIX APPENDIX 1 TABLE 1 DESCRIPTIVE STATISTICS OF HEDGE FUNDS AND CONVENTIONAL INDICES

Standard Sharpe Mean average T-stat Median Min Max deviation ratio CS 0.009 5.002 0.022 0.008 -0.076 0.086 0.268 CA 0.007 6.741 0.013 0.010 -0.047 0.036 0.354 DSB -0.000 -0.095 0.049 -0.004 -0.087 0.227 0.202 EM 0.008 2.160 0.048 0.014 -0.230 0.164 2.161 EMN 0.008 11.852 0.008 0.008 -0.012 0.033 1.125 EV 0.009 7.210 0.016 0.010 -0.118 0.037 1.073

D EV 0.011 7.407 0.018 0.012 -0.125 0.041 1.188

Part 1 MS EV 0.009 6.232 0.017 0.008 -0.115 0.080 1.530 RA EV 0.006 5.954 0.012 0.006 -0.062 0.039 1.475 FIA 0.005 5.882 0.011 0.007 -0.070 0.021 0.472 GM 0.010 4.016 0.032 0.011 -0.116 0.106 0.617 LES 0.011 4.143 0.033 0.009 -0.114 0.210 0.582 MF 0.004 1.528 0.036 0.002 -0.147 0.100 1.176 MS 0.008 7.753 0.012 0.008 -0.048 0.036 0.430 SP 0.008 2.430 0.041 0.012 -0.146 0.097 0.430

MSCI 0.007 2.123 0.039 0.011 -0.135 0.089 0.466 JPM 0.007 0.854 0.104 0.009 -0.445 0.317 0.159

Part 2 LH 0.021 1.989 0.135 0.025 -0.566 0.364 0.127 GD 0.008 1.916 0.050 0.008 -0.146 0.142 0.636

TABLE 2 CORRELATION BETWEEN HEDGE FUNDS STRATEGIES AND CONVENTIONAL INDICES

SP MSCI JPM LH GD Min Max CS 0.477** 0.488** 0.272** -0.057 0.080 -0.057 0.488 CA 0.153* 0.133* 0.170* -0.065 0.010 -0.065 0.170 DSB -0.753** -0.745** -0.420** 0.066 -0.077 -0.753 0.066 EM 0.464** 0.493** 0.375** -0.040 0.081 -0.040 0.494 EMN 0.359** 0.329** 0.308** 0.004 0.055 0.004 0.359 EV 0.550** 0.578** 0.439** -0.112 0.141* -0.112 0.578 D EV 0.543** 0.556** 0.406** -0.088 0.093 -0.088 0.556 MS EV 0.431** 0.479** 0.394** -0.138* 0.188** -0.136 0.479 RA EV 0.405** 0.406** 0.374** -0.002 0.112 -0.002 0.406 FIA 0.033 -0.013 0.050 -0.052 0.028 -0.052 0.050 GM 0.194** 0.133* 0.125 -0.010 -0.064 -0.064 0.194 LES 0.533** 0.599** 0.259** -0.127 0.175* -0.127 0.599 MF -0.156* -0.121 -0.098 0.097 0.071 -0.156 0.097 MS 0.099 0.179* -0.015 -0.067 0.060 -0.067 0.179 Min -0.753 -0.745 -0.420 -0.136 -0.077 Max 0.550 0.599 0.439 0.097 0.188

Journal of Applied Business and Economics vol. 16(3) 2014 45 TABLE 3 THE CORRELATION COEFFICIENTS OF ALTERNATIVE STRATEGIES Corrélations (Tau-B de Kendall)

C CA DSB EM EMN EV D D EV MS EV RA GM LES MF MS CS 0.222* - 0.458** 0.199** 0.515** 0.419** 0.498** 0.310** 0.259* 0.518* 0.602** 0.174* 0.254** CA - 0.175** 0.302** 0.374** 0.311** 0.395** 0.280** 0.319* 0.148* 0.190** -0.049 0.248** DSB ------0.015 - EM 0.200** 0.478** 0.394** 0.483** 0.267** 0.197* 0.290* 0.426** 0.025 0.140** EMN 0.299** 0.266** 0.305** 0.214** 0.227* 0.153* 0.255** 0.054 0.193** EV 0.741** 0.668** 0.398** 0.317* 0.250* 0.551** 0.063 0.244** D EV 0.446** 0.311** 0.251* 0.200* 0.449** 0.072 0.149** MS 0.366** 0.323* 0.276* 0.483** 0.026 0.328** RA 0.172* 0.178* 0.379** 0.014 0.149** FIA 0.241* 0.198** 0.041 0.137** GM 0.226** 0.232* 0.069 LES 0.083 0.214** MF 0.047 MS ** Correlation is significant at the 0.01 level (one-tailed). * Correlation is significant at the 0.05 level (one-tailed).

TABLE 4 THE PERFORMANCE OF DIFFERENT HEDGE FUND INDICES AND STRATEGIES FOR EACH PERIOD ANALYZED

1/94-12/95 1/96-12/97 1/98-12/99 1/00-12/01 1/02-12/03 1/04-12/05 1/06-12/07 1/08-12/10 CS 0.660% 1.852% 0.917% 0.398% 0.730% 0.696% 1.42% 1.42% CA 0.300% 1.258% 0.450% 1.463% 0.681% 0.352% 1.13% 2.09% DSB 0.325% -0.114% -0.679% 0.634% -0.855% 0.298% 1.83% 2.31% EM -0.100% 2.328% -0.145% 0.076% 1.378% 1.224% 1.76% 1.78% EMN 0.355% 1.229% 1.124% 0.961% 0.586% 0.650% 1.52% 2.28% EV 0.745% 1.641% 0.679% 0.754% 0.780% 0.959% 1.62% 1.58% D EV 1.010% 1.757% 0.818% 0.852% 0.927% 1.058% 1.53% 1.96% MS EV 0.725% 1.650% 0.526% 0.747% 0.886% 0.757% 2.60% 3.10% RA EV 0.485% 0.939% 0.736% 0.811% 0.218% 0.411% 1.19% 1.40% FIA 0.270% 0.995% 0.281% 0.580% 0.557% 0.811% 0.67% 0.41% GM 0.810% 2.018% 0.183% 1.195% 1.272% 0.856% 1.22% 0.81% LES 1.360% 1.512% 2.399% -0.003% 0.567% 0.219% 1.18% 0.86% MF -0.395% 0.647% 0.634% 0.316% 1.337% 0.721% 0.46% 0.22% MS 0.365% 1.259% 0.695% 0.671% 0.886% 0.333% 1.15% 0.72% SP 1.200% 1.983% 1.868% -0.891% -0.003% 0.635% -0.75% -4.28% MSCI 0.896% 1.071% 1.857% -1.335% 0.257% 1.236% -0.97% -0.70% JPM 1.800% 2.801% 0.045% -2.115% 0.836% 1.070% 4.25% 0.49% LH 1.690% 5.231% 2.358% 1.225% 1.206% 2.270% 2.60% 1.95% GD 0.697% -0.180% 0.424% -0.496% 1.876% -0.180% 0.79% -1.93%

46 Journal of Applied Business and Economics vol. 16(3) 2014 APPENDIX 2 STUDY OF CORRELATIONS

January 1994 December1995 January 1996 December1997 SP MSCI JPM LH GD SP MSCI JPM LH GD CS 0.453* 0.364* 0.352* 0.176 -0.050 CS 0.758** 0.631* 0.611** 0.066 -0.067 CA 0.399* 0.125 0.393* 0.076 0.116 CA 0.094 0.190 0.542** -0.051 0.105 DSB -0.702** -0.516* -0.376* 0.088 0.199 DSB -0.646** -0.650* -0.467* 0.012 -0.358* EM 0.200 0.098 0.238 0.144 -0.135 EM 0.529** 0.554* 0.538** 0.103 -0.282 EMN 0.165 0.081 0.225 0.140 -0.069 EMN 0.686** 0.619* 0.681** 0.013 0.183 EV 0.587* 0.407* 0.420* -0.132 0.192 EV 0.711** 0.708* 0.652** -0.028 0.098 D EV 0.720** 0.575* 0.359* -0.159 0.038 D EV 0.684** 0.668* 0.621** 0.048 0.197 MS 0.197 0.176 0.255 -0.220 0.353* MS 0.629** 0.624* 0.579** -0.086 0.004 RA 0.032 -0.323 0.205 0.469* -0.378* RA 0.212 0.295 0.233 -0.166 -0.107 FIA 0.259 -0.138 0.491** 0.268 -0.047 FIA 0.302 0.383* 0.273 0.015 -0.018 GM 0.236 -0.100 0.455* 0.518** -0.057 GM 0.625** 0.484* 0.367* 0.137 -0.279 LES 0.503** 0.581* 0.185 -0.437* 0.099 LES 0.690** 0.654* 0.558** -0.102 0.382* MF -0.359* -0.263 -0.141 0.417* -0.143 MF 0.566** 0.552* 0.573** 0.010 -0.065 MS 0.073 0.259 -0.020 -0.334 0.248 MS -0.300 -0.298 -0.231 0.598** -0.039 January 1998 December1999 January 2000 December 2001 SP MSCI JPM LH GD SP MSCI JPM LH GD CS 0.542** 0.594** 0.199 -0.363* 0.052 CS 0.228 0.353* 0.151 0.111 0.430* CA 0.191 0.246 0.178 -0.240 0.078 CA 0.191 0.170 0.000 0.036 0.252 DSB -0.872** -0.871** -0.407* 0.279 -0.218 DSB -0.775** -0.795* -0.301 -0.213 -0.064 EM 0.640** 0.741** 0.457* -0.399* 0.357* EM 0.618** 0.671* 0.359* 0.347* 0.152 EMN 0.643** 0.681** 0.563** -0.143 0.204 EMN 0.510** 0.487* 0.354* 0.098 0.091 EV 0.685** 0.762** 0.522** -0.313 0.270 EV 0.381* 0.453* 0.190 0.179 0.453* D EV 0.670** 0.732** 0.436* -0.245 0.245 D EV 0.284 0.289 0.151 0.056 0.075 MS 0.654** 0.743** 0.573** -0.366* 0.270 MS 0.331 0.435* 0.157 0.246 0.590* RA 0.656** 0.697** 0.534** -0.207 0.349 RA 0.175 0.255 0.140 0.171 0.563* FIA 0.015 -0.029 -0.050 -0.263 -0.104 FIA 0.036 0.097 -0.072 0.143 0.227 GM 0.194 0.218 -0.066 -0.288 -0.194 GM -0.055 0.050 0.268 -0.046 0.179 LES 0.813** 0.838** 0.321 -0.291 0.169 LES 0.235 0.356* 0.089 0.127 0.429* MF -0.317 -0.351* -0.365* 0.126 -0.135 MF -0.453* -0.410* 0.022 0.066 -0.042 MS -0.210 -0.213 -0.327 -0.331 -0.212 MS 0.328 0.408* 0.118 0.102 0.423* January 2002 December 2003 January 2004 December 2005 SP MSCI JPM LH GD SP MSCI JPM LH GD CS 0.534** 0.595** 0.599** 0.144 -0.039 CS 0.567** 0.855** 0.153 -0.201 0.217 CA 0.170 0.149 0.330 0.140 -0.085 CA 0.278 0.453** 0.111 -0.170 0.012 DSB -0.850** -0.850** -0.778** 0.050 0.015 DSB -0.741** -0.749** -0.315* 0.212 0.155 EM 0.781** 0.793** 0.608** -0.040 -0.131 EM 0.446** 0.730** 0.386* -0.097 0.210 EMN -0.019** 0.019** 0.022** 0.042 -0.074 EMN 0.048 0.287* 0.484** -0.219 0.176 EV 0.590** 0.609** 0.638** 0.082 -0.132 EV 0.586** 0.815** -0.031 -0.169 0.143 D EV 0.553** 0.567** 0.622** 0.098 -0.140 D EV 0.466** 0.753** -0.012 -0.240 0.226 MS 0.579** 0.621** 0.571** 0.292 -0.163 MS 0.444** 0.750** 0.066 -0.233 0.234 RA 0.597 0.677 0.634 0.260 -0.035 RA 0.627** 0.789** 0.146 -0.230 0.092 FIA -0.228 -0.152 -0.009 -0.191 0.075 FIA 0.029 0.227 -0.106 -0.044 0.311* GM -0.075 -0.062 -0.073 -0.043 0.045 GM 0.263 0.559** 0.129 -0.188 0.338* LES 0.653** 0.735** 0.671** 0.197 -0.109 LES 0.653** 0.908** 0.241 -0.246 0.123 MF -0.418* -0.394* -0.328 0.066 0.442* MF 0.535** 0.643** 0.184 0.007 0.175 MS 0.579* 0.621* 0.571* 0.292 -0.163 MS 0.447** 0.731** 0.052 -0.264 0.249 January 2006 December 2007 January 2008 December 2010 SP MSCI JPM LH GD SP MSCI JPM LH GD CS -0.130 0.093 0.313 0.173 0.211 CS 0.144 0.144 -0.144 -0.144 0.144 CA 0.087 0.097 0.144 -0.097 0.016 CA 0.098 0.098 -0.098 -0.098 0.098 DSB -0.775** -0.789** -0.386 0.086 -0.079 DSB 0.345* 0.345* -0.345* -0.345* 0.345* EM 0.508** 0.580** 0.431* -0.127 0.189 EM 0.281 0.281 -0.281 -0.281 0.281 EMN 0.149 0.144 0.240 -0.003 0.017 EMN 0.501** 0.501** -0.501 -0.501 0.501** EV 0.450* 0.489* 0.421* -0.119 0.122 EV 0.203 0.203 -0.203 -0.203 0.203 D EV 0.456* 0.480* 0.401* -0.099 0.086 D EV 0.264 0.264 -0.264 -0.264 0.264 MS 0.305* 0.342* 0.403* -0.116 0.136 MS 0.075 0.075 -0.075 -0.075 0.075 RA 0.401* 0.422* 0.408* -0.034 0.132 RA 0.149 0.149 -0.149 -0.149 0.149 FIA -0.041 -0.014 -0.029 -0.140 0.038 FIA 0.007 0.007 -0.007 -0.007 0.007 GM 0.078 0.120 0.032 -0.171 -0.043 GM -0.088 -0.088 0.088 0.088 -0.088 LES 0.533** 0.605** 0.271 -0.079 0.176 LES -0.039 -0.039 0.039 0.039 -0.039 MF -0.219 -0.169 -0.140 0.065 0.108 MF 0.022 0.022 -0.022 -0.022 0.022 MS 0.126 0.175 0.021 -0.088 0.056 MS -0.178 -0.178 0.178 0.178 -0.178

Journal of Applied Business and Economics vol. 16(3) 2014 47 APPENDIX 3 CALCULATION OF THE AVERAGE COMPOSITION

Average composition of a portfolio of traditional portfolio without hedge funds

SP MSCI JPM LH GD Return Utility 1 Average 27.024% 19.754% 19.094% 14.977% 19.152% 0.904% 0.926 1 Variance 0.304% 0.442% 0.343% 0.115% 0.150% 0.009% 0.076 0.5 Average 27.001% 19.782% 19.089% 14.955% 19.174% 0.934% 1.843 0.5 Variance 0.304% 0.444% 0.347% 0.116% 0.152% 0.009% 0.305 -0.5 Average 27.020% 19.776% 19.083% 14.948% 19.174% 0.933% -1.824 -0.5 Variance 0.306% 0.445% 0.348% 0.116% 0.152% 0.009% 0.306 -1 Average 27.030% 19.773% 19.080% 14.945% 19.172% 0.933% -0.908 -1 Variance 0.306% 0.445% 0.348% 0.116% 0.153% 0.009% 0.077 -3 Average 27.065% 19.762% 19.067% 14.936% 19.171% 0.933% -0.297 -3 Variance 0.308% 0.447% 0.350% 0.116% 0.153% 0.009% 0.009 -5 Average 27.098% 19.751% 19.057% 14.926% 19.168% 0.932% -0.175 -5 Variance 0.310% 0.448% 0.352% 0.116% 0.154% 0.009% 0.003 -10 Average 27.172% 19.727% 19.033% 14.907% 19.160% 0.932% -0.083 -10 Variance 0.315% 0.452% 0.356% 0.116% 0.154% 0.009% 0.001 -15 Average 27.237% 19.705% 19.015% 14.889% 19.155% 0.932% -0.053 -15 Variance 0.318% 0.455% 0.359% 0.116% 0.155% 0.009% 0.000 -20 Average 27.292% 19.687% 18.999% 14.874% 19.149% 0.931% -0.038 -20 Variance 0.322% 0.457% 0.361% 0.116% 0.156% 0.009% 0.000 -30 Average 27.385% 19.654% 18.975% 14.850% 19.137% 0.931% -0.023 -30 Variance 0.327% 0.461% 0.365% 0.116% 0.157% 0.009% 0.000 -50 Average 27.518% 19.605% 18.943% 14.816% 19.118% 0.931% -0.012 -50 Variance 0.335% 0.467% 0.370% 0.116% 0.158% 0.009% 0.000 R: rentability of portfolio

Average composition of an optimal portfolio of traditional assets with a general index of hedge funds

CS SP MSCI JPM LH GD Return Utility 1 Average 14.838% 23.242% 17.683% 15.626% 12.537% 15.816% 0.925% 1.009 1 Variance 0.104% 0.000% 0.000% 0.000% 0.000% 0.000% 0.009% 0.000 0.5 Average 14.797% 23.238% 17.671% 15.625% 12.508% 15.833% 0.924% 2.009 0.5 Variance 0.111% 0.000% 0.000% 0.000% 0.000% 0.000% 0.009% 0.000 -0.5 Average 14.804% 23.266% 17.665% 15.616% 12.501% 15.831% 0.924% -1.991 -0.5 Variance 0.112% 0.000% 0.000% 0.000% 0.000% 0.000% 0.009% 0.000 -1 Average 14.812% 23.282% 17.663% 15.612% 12.496% 15.828% 0.924% -0.991 -1 Variance 0.112% 0.000% 0.000% 0.000% 0.000% 0.000% 0.009% 0.000 -3 Average 14.822% 23.334% 17.653% 15.595% 12.484% 15.824% 0.924% -0.32 -3 Variance 0.112% 0.000% 0.000% 0.000% 0.000% 0.000% 0.009% 0.000 -5 Average 14.920% 23.097% 16.638% 14.083% 13.658% 17.603% 0.884% -0.192 -5 Variance 0.106% 0.000% 0.000% 0.000% 0.000% 0.000% 0.009% 0.000 -10 Average 15.111% 22.498% 15.787% 13.952% 14.016% 18.925% 0.913% -0.092 -10 Variance 0.087% 0.000% 0.000% 0.000% 0.000% 0.000% 0.008% 0.000 -15 Average 15.120% 22.541% 15.767% 13.921% 14.001% 18.959% 0.913% -0.059 -15 Variance 0.087% 0.000% 0.000% 0.000% 0.000% 0.000% 0.008% 0.000 -20 Average 15.122% 22.572% 15.752% 13.893% 13.989% 18.990% 0.913% -0.042 -20 Variance 0.086% 0.000% 0.000% 0.000% 0.000% 0.000% 0.008% 0.000 -30 Average 15.125% 24.491% 15.598% 14.472% 13.543% 17.099% 0.947% -0.026 -30 Variance 0.086% 0.000% 0.000% 0.000% 0.000% 0.000% 0.008% 0.000 -50 Average 15.095% 24.300% 17.340% 15.469% 12.370% 15.682% 0.917% -0.014 -50 Variance 0.083% 0.000% 0.000% 0.000% 0.000% 0.000% 0.009% 0.000

48 Journal of Applied Business and Economics vol. 16(3) 2014 Average composition of an optimal portfolio of assets with traditional indices of hedge fund strategies (in %)

CA DSB EM EMN EV D EV MS EV RA EV FIA GM LES MF MS % 1 Averag 5.05 6.21 6.15 4.42 3.61 4.27 3.73 4.58 5.18 4.19 4.87 5.80 4.05 62.09 1 Varianc 0.01 0.02 0.02 0.01 0.01 0.01 0.01 0.02 0.01 0.02 0.04 0.02 0.00 0.5 Averag 5.12 6.18 6.26 4.48 3.57 4.29 3.74 4.63 5.71 4.28 4.93 5.87 4.11 63.17 0.5 Varianc 0.01 0.02 0.02 0.01 0.01 0.01 0.01 0.03 0.02 0.02 0.05 0.02 0.01 -0.5 Averag 5.12 6.09 6.26 4.47 3.53 4.26 3.73 4.66 5.41 4.29 4.93 5.83 4.12 62.72 -0.5 Varianc 0.01 0.01 0.02 0.01 0.01 0.01 0.01 0.03 0.01 0.02 0.06 0.02 0.01 -1 Averag 5.12 6.01 6.26 4.47 3.52 4.25 3.73 4.68 5.44 4.30 4.92 5.81 4.14 62.65 -1 Varianc 0.01 0.01 0.02 0.01 0.01 0.01 0.01 0.03 0.01 0.02 0.06 0.02 0.01 -3 Averag 5.14 5.91 6.24 4.50 3.55 4.26 3.77 4.73 5.50 4.33 4.93 5.76 4.19 62.79 -3 Varianc 0.01 0.01 0.02 0.01 0.01 0.01 0.01 0.03 0.01 0.02 0.06 0.02 0.01 -5 Averag 5.16 5.74 6.18 4.53 3.60 4.28 3.83 4.79 5.54 4.35 4.95 5.69 4.26 62.89 -5 Varianc 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.03 0.01 0.02 0.06 0.02 0.01 -10 Averag 5.19 5.64 6.16 4.57 3.65 4.31 3.89 4.83 5.58 4.39 4.98 5.67 4.32 63.18 -10 Varianc 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.03 0.01 0.02 0.06 0.01 0.01 -15 Averag 5.22 5.49 6.08 4.61 3.71 4.34 3.93 4.87 5.60 4.42 5.00 5.62 4.36 63.23 -15 Varianc 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.03 0.01 0.02 0.05 0.01 0.01 -20 Averag 5.33 5.54 6.17 4.72 3.82 4.46 4.05 4.96 5.70 4.53 5.08 5.70 4.47 64.52 -20 Varianc 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.03 0.01 0.02 0.05 0.01 0.01 -30 Averag 5.35 5.41 6.12 4.71 4.05 4.25 4.37 5.31 5.31 4.85 5.60 5.45 4.64 65.42 -30 Varianc 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.02 0.02 0.01 0.04 0.01 0.00 -50 Averag 5.34 5.35 6.09 4.72 3.90 4.49 4.09 5.31 5.67 4.73 5.32 5.68 4.52 65.20 -50 Varianc 0.01 0.01 0.02 0.01 0.01 0.01 0.00 0.02 0.01 0.02 0.04 0.02 0.01

Next: Average composition of an optimal portfolio of assets with traditional indices of hedge fund strategies

SP MSCI JPM LH GD R Utility 1 Average 6.133% 6.605% 7.602% 8.846% 8.721% 0.884% 1.009 1 Variance 0.033% 0.084% 0.104% 0.177% 0.302% 0.002% 0.000 0.5 Average 6.450% 6.040% 7.707% 9.031% 7.604% 0.984% 2.010 0.5 Variance 0.047% 0.061% 0.120% 0.218% 0.264% 0.003% 0.000 -0.5 Average 6.554% 6.138% 7.711% 9.094% 7.788% 0.962% -1.990 -0.5 Variance 0.057% 0.072% 0.125% 0.231% 0.254% 0.003% 0.000 -1 Average 6.590% 6.174% 7.704% 9.138% 7.740% 0.942% -0.991 -1 Variance 0.061% 0.077% 0.134% 0.242% 0.259% 0.004% 0.000 -3 Average 6.632% 6.226% 7.634% 9.096% 7.622% 0.939% -0.324 -3 Variance 0.064% 0.081% 0.142% 0.245% 0.261% 0.004% 0.000 -5 Average 6.649% 6.264% 7.593% 9.045% 7.564% 0.938% -0.191 -5 Variance 0.067% 0.085% 0.142% 0.240% 0.261% 0.004% 0.000 -10 Average 6.665% 6.292% 7.475% 8.922% 7.465% 0.936% -0.091 -10 Variance 0.069% 0.087% 0.136% 0.229% 0.261% 0.004% 0.000 -15 Average 6.632% 6.276% 7.498% 8.913% 7.449% 0.933% -0.058 -15 Variance 0.067% 0.084% 0.138% 0.226% 0.262% 0.004% 0.000 -20 Average 6.729% 6.382% 6.328% 8.652% 7.395% 0.929% -0.042 -20 Variance 0.080% 0.100% 0.080% 0.235% 0.259% 0.004% 0.000 -30 Average 5.917% 5.574% 6.576% 8.846% 7.668% 0.926% -0.026 -30 Variance 0.024% 0.020% 0.065% 0.188% 0.241% 0.004% 0.000 -50 Average 6.023% 5.595% 6.986% 8.242% 7.951% 0.904% -0.013 -50 Variance 0.014% 0.017% 0.084% 0.137% 0.220% 0.003% 0.000

Journal of Applied Business and Economics vol. 16(3) 2014 49 APPENDIX 4 TABLE 1 PROFITABILITY AND AVERAGE SHARPE RATIOS OF OPTIMAL PORTFOLIOS Stock portfolio without hedge funds (P1) Equity portfolio and general index of Hedge Fund (P2) Portfolio of stocks and indices of hedge fund strategies (P3)

1 0.5 -0.5 -1 -3 -5 -10 -15 -20 -30 -50 Average (%) 0.931 0.931 0.931 0.931 0.932 0.932 0.932 0.933 0.933 0.933 0.934

Standard deviation (%) P1 0.929 0.929 0.929 0.929 0.929 0.930 0.930 0.930 0.930 0.930 0.931 Ratio of Sharpe 0.772 0.771 0.771 0.771 0.771 0.771 0.771 0.771 0.770 0.770 0.770 Average (%) 0.917 0.947 0.913 0.913 0.913 0.884 0.924 0.924 0.924 0.924 0.925

P2 Standard deviation (%) 0.929 0.887 0.888 0.888 0.888 0.927 0.929 0.929 0.929 0.929 0.929 Ratio of Sharpe 0.763 0.763 0.763 0.763 0.763 0.721 0.786 0.786 0.785 0.824 0.756 Average (%) 0.904 0.926 0.929 0.933 0.936 0.938 0.939 0.942 0.962 0.984 0.884

Standard deviation (%)

P3 0.543 0.595 0.606 0.614 0.620 0.623 0.620 0.621 0.571 0.530 0.470 Ratio of Sharpe 1.422 1.451 1.307 1.170 1.166 1.160 1.162 1.169 1.177 1.195 1.268

TABLE 2 COMPARISON OF THE AVERAGE / SEMI-VARIANCE OPTIMAL PORTFOLIOS

1 0.5 -0.5 -1 -3 -5 -10 -15 -20 -30 -50 Average (%) 0.931 0.931 0.931 0.931 0.932 0.932 0.932 0.933 0.933 0.933 0.934

-2

P1 Semi variance (% 10 ) 59.65 59.57 59.57 59.57 59.56 59.56 59.55 59.55 59.56 59.56 59.58 E(R) / SV 156.04 156.29 156.34 156.38 156.43 156.50 156.55 156.59 156.59 156.61 156.71 Average (%) 0.917 0.947 0.913 0.913 0.913 0.884 0.924 0.924 0.924 0.924 0.925

-2

P2 Semi variance (% 10 ) 55.68 55.70 55.70 55.70 55.70 53.29 48.09 48.08 48.08 49.98 55.62 E(R) / SV 164.76 169.93 163.86 163.89 163.96 165.98 192.10 192.18 192.21 184.94 166.21 Average (%) 0.904 0.926 0.929 0.933 0.936 0.938 0.939 0.942 0.962 0.984 0.884

Semi variance (% 10-2) P3 11.74 13.00 16.63 21.75 21.72 21.80 21.44 20.95 20.37 19.42 17.00 E(R) / SV 770.03 712.33 558.28 429.05 430.96 430.44 437.80 449.63 472.39 506.82 520.02

50 Journal of Applied Business and Economics vol. 16(3) 2014

Technology Transfer from MNCs to Host Country Enterprises: An In-depth Analysis Based on Game Theories

Peihua Zhao Shantou University

Weiguo Zhang South China University of Technology

Robert Guang Tian Shantou University

INTRODUCTION

The problem of MNCs’ technology transfer is virtually the game problem between MNCs and host country enterprises. Some experts have realized this point and have done some research: Das(1987) set up a game model between MNCs and host country enterprises in the 1980s. Wang and Blomstrom(1992) also created their game models to discuss the game problem between MNCs and host country enterprises. However, the methods that these experts used was the traditional game theory, which paid attention to the technology spillover effect. Chinese experts such as Longying Hu(2000) used the model of a cooperative game to prove that, under certain restrictions of the policy system environment, both sides of MNCs and host country enterprises who were in information asymmetry can establish the cooperation mechanism; and Yixun Zhang(2002) also used the same method to prove that both the MNCs and host country enterprises can gain maximum profits with cooperation between them. Nevertheless, these experts hadn’t paid attention to a more microcosmic, specific and realistic problem: both MNCs and host country enterprises had easy success and earned maximum profits, as long as they made sure which kind of host country enterprises with whom they should cooperate ( viz: the MNCs should cooperate with which kind of host country enterprises, and host country enterprises should cooperate with which kind of MNCs).

Journal of Applied Business and Economics vol. 16(3) 2014 51 THE BASIC HYPOTHESES OF THE EVOLUTIONARY GAME MODEL

The MNCs have two strategies to select on technology transfer after they enter the host country One is that the MNCs cooperate with host country enterprises and transfer the technology to them; another is that the MNCs establish their own companies, do not cooperate with host country enterprises, and, of course, do not transfer the technology to them. At the same time, the host enterprises also have two strategies to select. One is that the host country enterprises cooperate with MNCs and receive the technology; another is that the host country enterprises do not cooperate with MNCs and develop by themselves. When the MNCs decide whether to transfer the technology to host country enterprises or not, and the host country enterprises decide whether to cooperate with MNCs or not, wing to the fact that information is not complete, and that both sides (MNCs and host country enterprises) are all bounded rationally, both are playing a game which is a learning process and a dynamic process that is adjusted continually. This paper analyzes the initiative and evolutionary process and the evolutionarily stable strategy (ESS) of the MNCs and host country enterprises when they cooperate with each other. To analyze this process, the paper makes the following hypotheses concerning the evolutionary model: Hypothesis 1. Pair wise game. Although one individual of two groups (MNCs and host enterprises) will face all other enterprises when he makes a decision, we can assume that the game is taking place between MNCs and host country enterprises. Hypothesis 2. Approximate eye. When one of the MNCs changes its strategy, it always takes the distribution of recent strategy as a known condition, then transforms to one type of the best strategies corresponding to the recent strategy. Despite that large numbers of MNCs change strategies, it will make the strategy space and payoff function different from the original ones . Hypothesis 3. Cooperative condition. In order to facilitate the following theoretical analysis, this paper assumes that the cooperation between host country enterprises and MNCs is based on technology transfer. That is, if the MNCs want to cooperate with host country enterprises, MNCs must transfer their technology to host country enterprises, otherwise, it is deemed to be uncooperative. Hypothesis 4. Always find the cooperator. If only one group (MNCs or host country enterprises) wants to, it can find his cooperator in another group (MNCs or host country enterprises). According to different aspects such as technology level and profitability of different enterprises (MNCs or host country enterprises), we assume that MNCs can find the cooperators only if they are willing to transfer their technologies; meanwhile, host country enterprises can find the cooperators only if they are willing to cooperate with MNCs and receive their technologies. Hypothesis 5. We assume that MNCs have not entered the host country and established companies, but is only ready to do, or, we assume that MNCs are not allowed to enter the host country using a sole proprietorship form of business organization.

THE EVOLUTIONARY GAME MODEL AND EVOLUTIONARILY STABLE STRATEGY ANALYSIS

When the MNCs and host country enterprises repeatedly play a dynamic game, the MNCs have two strategies to select ( namely N=2 ): Strategy 1 The MNCs transfer their technologies to the host country ( cooperate with host country enterprises ), Strategy 2 The MNCs do not transfer their technologies to the host country ( not cooperate with host country enterprises ). The host country enterprises also have two strategies to select ( namely N=2 ): Strategy 1 The host country enterprises cooperate with the MNCs, Strategy 2 The host country enterprises do not cooperate with the MNCs. We assume that the loss caused by technology spillover during technology transfer is L1, the cost needed by technology transfer is C1 ; the cost needed by cooperation between host country enterprises and MNCs is C2, the loss caused by decrease of market share in cooperation is L2. We assume that the gross profits gained by cooperation between host country enterprises and MNCs is R, in which, the profits gained by the MNCs is β R ( namely the proportion of both return on equity is β ：（1－ β ）), the profits gained by host country enterprises after

52 Journal of Applied Business and Economics vol. 16(3) 2014 cooperation with MNCs is （1－ β ）R. Moreover, the profits of the technology improvement caused by the host country enterprises’ cooperation is R. These elements always directly impact the game payoff function. When all of the MNCs have not entered the host countries but are ready to do, or the MNCs cannot enter the host countries using a sole proprietorship form of business organization, a situation is created which we will discuss accordingly.

The Host Country Enterprises Without Research and Development Capability When the industry that the MNCs want to set up is a new one for a host country, and the host country has no related basic technologies and is indifferent to carry on independent research and development, the game payoff matrix is shown in table 1.

TABLE 1 THE GAME PAYOFF MATRIX BETWEEN HOST COUNTRY ENTERPRISES AND MNCS WITHOUT ENTRY( ENTERPRISES WITH NO R＆D CAPABILITY)

Host country enterprises 2 1 Cooperation non- cooperation

，， transfer β −− CLR 11 β )1( −+− CrR 2 β −− CLR 11 0 MNCs ， Non-transfer 0 β )1( −+− CrR 2 0，0

This paper assumes that P denotes the proportion of strategy 1 (transfer), and is adopted by MNCs, and q denotes strategy 1 (cooperation), and is adopted by host country enterprises. Therefore, one point 1 1 2 2 (p, q) in the area of S = [ 0，1 ] × [ 0 ，1] is used to describe the state s = { ( s1 ， s2 )，( s1 ， s2 ) } = { 1 2 1 2 1 (p， 1 –p )，(q， 1 - q) }, in which, s1 = p ， s1 = q ，thus, s2 =1 - p ， s2 =1 – q. r = (1，0) denotes the strategy 1 (transfer) that MNCs select with probability 1, r 2 = (0，1) denotes the strategy 2 (transfer) that MNCs select with probability 2. For MNCs, we can know from table 1: The strategy 1 (transfer) is adopted by MNCs, which fitness is: 1 （ 1 ，）（）（）（） f r s = q β −− CLR 11 + 1− q β −− CLR 11 The fitness of adopting strategy 2 (non-transfer) is: 21 qqsrf =×−+×= 00)1(0),( Its average fitness is: 1 11 21 = −+ ([),()1(),(),( β 11 )(1() β −−−+−−= CLRqCLRqpsrfpsrpfspf 11 )] Similarly, for the host country enterprises, the strategy 1 (cooperation) is adopted by host country enterprises, whose fitness is: 12 β +−= 2 β −+−−+− CrRpCrRpsrf 2 ])1)[(1(])1[(),( The fitness of adopting strategy 2 (non-transfer) is: 22 ppsrf =×−+×= 00)1(0),( Its average fitness is：

Journal of Applied Business and Economics vol. 16(3) 2014 53 2 12 221 = −+ β 2 β )1)[(1(])1[({),()(),(),( −+−−+−+−= CrRpCrRpqsrfqsrqfsqf 2 ]} In evolutionary game theory, duplicators of populations are dynamically assumed to be : the growth rate of one strategy depends on its fitness, and the strategies that produce higher profits have higher growth rate. Therefore, according to Malthusian equation, the strategy 1 (transfer) is adopted by the ⋅ MNCs, whose fitness f 1 （ r1 ， s ） minus the average fitness f 1 （ p ， s ）equals its growth rate p / p , that is, ⋅ p / p = f 1 （ r1 ， s ）– f 1 （ p ， s ） ⋅ thus， )(1( β −−−= CLRppp 11 ) （1） Similarly, we can know, the strategy 1 (transfer) is adopted by host country enterprises, whose growth rate is: ⋅ 12 −= 2 sqfsrfqq ),(),(/ ⋅ （） That is, β −+−−= CrRqqq 2 ])1)[(1( 2 A two-dimensional, dynamic and autonomous (not containing time t) system is made up of (1) and (2).

According to the theory of differential equations, if（ p0 ， q0 ）makes the right side of (1) and (2) be zero, so then we have equations （）（） p0 1– p0 β −− CLR 11 =0 （） − qq 00 )1( β )1( −+− CrR 2 =0

Then（ p0 ， q0 ）is the equilibrium point or singular point. Therefore, this autonomous system has four equilibrium points (singular point):

E1（0，0）、E2（0，1）、E3（1，0）、E4（1，1） According to the method proposed by Friedman, if there is a population dynamic systematically described by the differential equation, its stability at the equilibrium point is obtained through a local stability analysis of the Jacobian matrix. The system is composed by equation (1) and (2), whose Jacobian matrix is: ⋅ ⋅    )(21( β −−− CLRp ) 0   // ∂∂∂∂ qppp  11 J= ⋅ ⋅ =     0 β −+−− CrRq ))1)[(21(  // ∂∂∂∂ qqpq   2  The determinant of the Jacobian matrix is:

detJ= )(21( β −−− CLRp 11 ) β −+−− CrRq 2 ])1)[(21( The trace of the Jacobian matrix is:

trJ= )(21( β −−− CLRp 11 ) + β −+−− CrRq 2 ])1)[(21(

TABLE 2 THE LOCAL STABILITY ANALYSIS OF THE EQUILIBRIUM POINT

equilibrium The determinant of Jacobian matrix: The trace of Jacobian matrix: trJ point detJ （，）（）（） E1 0 0 β −− CLR 11 [ β )1( −+− CrR 2 ] β −− CLR 11 +[ β )1( −+− CrR 2 ] （，）－（）（）－ E2 0 1 β −− CLR 11 β )1( −+− CrR 2 ] β −− CLR 11 [ β )1( −+− CrR 2 ]

（，）－（）－（） E3 1 0 β −− CLR 11 [ β )1( −+− CrR 2 ] β −− CLR 11 +[ β )1( −+− CrR 2 ] （，）（）－ E4 1 1 β −− CLR 11 [ β )1( −+− CrR 2 ] [ β −− CLR 11 + β )1( −+− CrR 2 ]

54 Journal of Applied Business and Economics vol. 16(3) 2014 ，（－） Proposition 1 When β R< + CL 11 1 β R+r< C2 , E1（0，0）is a locally and asymptotically stable point, its evolutionary stable strategy (ESS) is (non-transfer, non-cooperation). ，（－），（） Proof: When β R< + CL 11 1 β R+r< C2 β −− CLR 11 [ β )1( −+− CrR 2 ，（）， ]>0 and β −− CLR 11 +[ β )1( −+− CrR 2 ]<0 that is, we can know from table 2 that detJ>0，trJ<0，in the equilibrium point E1（0，0），now, the E1（0，0）is the evolutionary stable ，（－） strategy (ESS) of the system. And, when β R< + CL 11 1 β R+r< C2 , we can know from the table 2 that E4（1，1）is an unstable point of the system, and E2（0，1）、E3（1，0）are the saddle points of the system (table 3), thus, E1（0，0）is the only ESS of system. (Completion of proof)

TABLE 3 THE LOCAL STABILITY ANALYSIS’ RESULT SHOWING THE EQUILIBRIUM POINT OF PROPOSITION 1

equilibrium point detJ’s sign trJ’ sign local stability

E1（0，0） + － ESS （，） E2 0 1 － ± saddle point （，） E3 1 0 － ± saddle point （，） E4 1 1 + + unstable point

The proposition shows that: When all the profits that the MNCs and host country enterprises gain （，（－） through cooperation with each other do not reach a certain value β R< + CL 11 1 β R+r< C2 ）, the MNCs and host country enterprises all trend to cooperate: the MNCs do not transfer the technologies to host country enterprises, and the host country enterprises do not cooperate with them. ，（－），（，） Proposition 2 When β R< + CL 11 1 β R+r> C2 E2 0 1 is locally and asymptotically stable point, so the evolutionary stable strategy (ESS) of the system is (transfer, non-transfer). ，（－），－（） Proof: When β R< + CL 11 1 β R+r> C2 β −− CLR 11 [ β )1( −+− CrR 2 ，（）－， ]>0 and β −− CLR 11 [ β )1( −+− CrR 2 ]<0 that is, we can know from table 2 that in equilibrium point E2（0，1），detJ>0，trJ<0， through analyzing the stability of the equilibrium point, now, E2（0，1）is the evolutionary stable strategy (ESS) of the system. In addition, we can know from Table 2 that E3（1，0）is an unstable point of system, and E1（0，0）, E4（1，1）are the saddle points of system (table 4), so E2（0，1）is the only ESS of system. (Completion of proof)

Journal of Applied Business and Economics vol. 16(3) 2014 55 TABLE 4 THE LOCAL STABILITY ANALYSIS’S RESULT OF THE EQUILIBRIUM POINT OF PROPOSITION 2

equilibrium point detJ’s sign trJ’ sign local stability

E1（0，0）－ ± saddle point （，） E2 0 1 + － ESS （，） E3 1 0 + + unstable point （，） E4 1 1 － ± saddle point

The proposition shows that: When the profits that the MNCs gain through technology transfer with （） host country enterprises is less than a certain value β R< + CL 11 , the MNCs tend to not transfer the technologies; when the profits that the host country enterprises gain through cooperation with MNCs （（－）） exceed the cost 1 β R+r> C2 , the host country enterprises tend to cooperate with MNCs. ，（－） Proposition 3 when β R> + CL 11 1 β R+r< C2 , E3（1，0）is the locally and asymptotically stable point, the evolutionary stable strategy (ESS) of system is (transfer, non- cooperation). ，（－）－（） Proof: When β R> + CL 11 1 β R+r< C2 , β −− CLR 11 [ β )1( −+− CrR 2 ，－（）， ]>0 and β −− CLR 11 +[ β )1( −+− CrR 2 ]<0 that is, through stable analysis to the equilibrium point, we can know from table 2 that detJ>0，trJ<0 in the equilibrium point E3（1，0）. At this time, E3（1，0）is the evolutionary stable strategy (ESS) of system. And this time, we can know from table 2 that E2（0，1）is the unstable point of system, and E1（0，0）、E4（1，1）are the saddle points (table 5), so E3（1，0）is the only evolutionary stable strategy (ESS) of the system. (Completion of proof)

TABLE 5 THE LOCAL STABILITY ANALYSIS’ RESULT SHOWING THE EQUILIBRIUM POINT OF PROPOSITION 3

equilibrium point detJ’s sign trJ’ sign local stability

E1（0，0）－ ± saddle point （，） E2 0 1 + + unstable point （，） E3 1 0 + － ESS （，） E4 1 1 － ± saddle point

The proposition shows that: When the profits that the MNCs gain through technology transfer with （） host country enterprises exceed a certain value β R< + CL 11 , the MNCs tend to transfer the technologies; when the profits that the host country enterprises gain through cooperation with MNCs （（－）） cannot reach the cost 1 β R+r> C2 , the host country enterprises tend to not cooperate with MNCs.

56 Journal of Applied Business and Economics vol. 16(3) 2014 ，（－） Proposition 4: when β R> + CL 11 1 β R+r< C2 , that is, when R> + CL 11 + C2 －，（）（－），（，） r + CL 11 /R< β < R+r C2 /R E4 1 1 is locally and asymptotically stable, the evolutionary stable strategy (ESS) of system is (transfer, cooperation). ，（－）（） Proof: When β R> + CL 11 1 β R+r< C2 , β −− CLR 11 [ β )1( −+− CrR 2 ，－（）， ]>0 and β −− CLR 11 +[ β )1( −+− CrR 2 ]<0 that is, through stable analysis at the equilibrium point, we can know from table 2 that detJ>0，trJ<0 in the equilibrium point E3（1，0）. At this time, E4（1，1）is the evolutionary stable strategy (ESS) of the system. And this time, we can know from Table 2 that E1（0，0）is the unstable point of the system, and E2（0，1）、E3（1，0）are the saddle points (table 6), so E4（1，1）is the only the evolutionary stable strategy (ESS) of the system. (Completion of proof)

TABLE 6 THE LOCAL STABILITY ANALYSIS’ RESULT OF THE EQUILIBRIUM POINT OF PROPOSITION 4

equilibrium point detJ’s sign trJ’ sign local stability

E1（0，0） + unstable point （，） E2 0 1 － ± saddle point （，） E3 1 0 － ± saddle point （，） E4 1 1 + － ESS

The proposition shows that: When the profits that the MNCs gain through technology transfer with （） host country enterprises exceed a certain value β R< + CL 11 , the MNCs tend to transfer the technologies; when the profits that the host country enterprises gain through cooperation with MNCs （（－）） exceed the cost 1 β R+r> C2 , the host country enterprises tend to cooperate with MNCs. Then they cooperate successfully; that is, they come to comprehensive co-operation.

The Host Country Enterprises with Research Capability We now consider the situation when the industry that the MNCs want to set up is not a new one for the host country, and the host country has the related basic technologies, research and development capability and is ready or has been ready to carry on the independent research and development. We assume that profits are gained by host country enterprises that have research and development capability, and who do not cooperate with MNCs but whose activities are shown by R02. The loss caused by decrease of market share in cooperation is L2.The host country enterprises cooperate with the MNCs that affect their independent research and development and lead to loss, and we assume this loss to be L21 , so the game payoff matrix is shown in table 7.

Journal of Applied Business and Economics vol. 16(3) 2014 57 TABLE 7 THE GAME PAYOFF MATRIX OF HOST COUNTRY ENTERPRISES AND MNCS WHO DO NOT ENTER

Host country enterprises 2 1 Cooperation non- cooperation

，， β −− CLR 11 β −− CLR 11

transfer R02 β )1( −−−+−+ CLLrR 2212 − LR 202 0， 0， non- MNCs transfer R02 β )1( −−−+−+ CLLrR 2212 R02

At this time, the same to 2.1, we can know from the Malthusian’s system of equations: ⋅ （）（） p = − pp )1( β −− CLR 11 3 ⋅ q = − qq )1( { 2 pL β )1[( −−−+−+ CLLrR 2212 ]} （4）

So there are four equilibrium points (singular point) in an automatic system: E1（0，0）、E2（0，1）、E3（1，0）、E4（1，1）

And the corresponding Jacobian matrix is:  ⋅ ⋅  // ∂∂∂∂ qppp J=    ⋅ ⋅   // ∂∂∂∂ qqpq 

 )(21( β −−− CLRp 11 ) 0  =  − − β −−−+−+   )1( Lqq 2 { 2 pLq [ )1()21( CLLrR 2212 ] }

The determinant of the Jacobian matrix：

detJ= )(21( β −−− CLRp 11 ) 2 β )1()[21( −−−+−+− CLLrRpLq 2212 ]

The trace of the Jacobian matrix： β −−− β −−−+−+− trJ= )(21( CLRp 11 ) + 2 )1()[21( CLLrRpLq 2212 ]

58 Journal of Applied Business and Economics vol. 16(3) 2014 TABLE 8 THE LOCAL STABILITY ANALYSIS OF THE EQUILIBRIUM POINT equilibriu determinant of Jacobian matrix:detJ trace of Jacobian matrix: trJ m points （， E1 0 0 (β CLR 11 β )1() −−−+−×−− CLLrR 2212 (β CLR 11 β )1() −−−+−+−− CLLrR 2212 ） E （0，1 β β −−−+−−−− 2 (β CLR 11 β )1() −−−+−×−−− CLLrR 2212 ( CLR 11 )1() CLLrR 2212 ） E （1，0 β β −−+−+−−− 3 (β CLR 11 β )1() −−+−×−−− CLrR 221 ( CLR 11 )1() CLrR 221 ） E （1，1 β β −−+−×−−− 4 (β CLR 11 β )1() −−+−×−− CLrR 221 {( CLR 11 )1() CLrR 221 } ）

，（－）（，） Proposition 5: When β R> + CL 11 1 β R+r< C2 , E1 0 0 is locally and asymptotically stable, the evolutionary stable strategy (ESS) of the system is (non-transfer, non-cooperation). Proof: ，（－）（） When β R< + CL 11 1 β R+r< + CL 221 , β −− CLR 11 [ β )1( −−−+− CLLrR 2212 （），， ]>0, and β −− CLR 11 + [ β )1( −−−+− CLLrR 2212 ]<0 that is, detJ>0 trJ<0 in the equilibrium point E1（0，0）. At this time, through stable analysis to the equilibrium point, we can know from table 8 （，），（－ that E1 0 0 is the evolutionary stable strategy (ESS) of the system. And when β R< + CL 11 1 ）（，） β R+r< + CL 221 , we can know from table 8 that E4 1 1 is the unstable point of the system, and E2（0，1）、E3（1，0）are the saddle points (table 9), so E1（0，0）is the only the evolutionary stable strategy (ESS) of system. (Completion of proof)

TABLE 9 THE LOCAL STABILITY ANALYSIS RESULT SHOWING THE EQUILIBRIUM POINT OF PROPOSITION 5

equilibrium detJ’s sign trJ’ sign local stability points E （0，0） + － 1 ESS （，）－ ± E2 0 1 saddle point E （1，0）－ ± 3 saddle point E （1，1） + + 4 unstable point

The proposition shows that: When the profits that the MNCs gain through technology transfer with host country enterprises cannot reach a certain value（ β R< + CL 11 , （1－ β ）R+r> ++ CLL 2212 , no matter whether it is the MNCs or host country enterprises, they both tend to not cooperate: the MNCs do not transfer their technologies, and the host country enterprises do not cooperate with them.

Journal of Applied Business and Economics vol. 16(3) 2014 59 ，（－）（，） Proposition 6 When β R< + CL 11 1 β R+r> ++ CLL 2212 , E2 0 1 is locally and asymptotically stable, the evolutionary stable strategy (ESS) of system is (non-transfer, non-cooperation). ，（－）－（ Proof: When β R< + CL 11 1 β R+r > ++ CLL 2212 , β −− CLR 11 ）（）， [ β )1( −−−+− CLLrR 2212 ]>0, and β −− CLR 11 [ β )1( −−−+− CLLrR 2212 ]<0 that is, detJ>0，trJ<0 in the equilibrium point E2（0，1）. At this time, through stable analysis to the equilibrium point, we can know from table 8 that E2（0，1）is the evolutionary stable strategy (ESS) of the system. ，（－） And when β R< + CL 11 1 β R+r< + CL 221 , we can know from table 8 that E3（1，0）is the unstable point of the system, and E1（0，0）、E4（1，1）are the saddle points (table 10), so

E2（0，1）is the only evolutionary stable strategy (ESS) of the system. (Completion of proof)

TABLE 10 THE LOCAL STABILITY ANALYSIS RESULT OF THE EQUILIBRIUM POINT

equilibrium detJ’s sign trJ’ sign local stability points E （0，0）－ ± 1 saddle point （，） + － E2 0 1 ESS E （1，0） + + 3 unstable point E （1，1）－ ± 4 saddle point

The proposition shows that: When the profits that the MNCs gain through cooperation with host country enterprises is less than a certain value β R< + CL 11 , they tend to not transfer their technologies; when the profits that the host country enterprises gain through cooperation with MNCs exceed a certain （－） value 1 β R+r> ++ CLL 2212 , the host country enterprises tend to cooperate with the MNCs. ，（－）（，） Proposition 7 when β R< + CL 11 1 β R+r> ++ CLL 2212 , E3 1 0 is locally and asymptotically stable, the evolutionary stable strategy (ESS) of the system is (transfer, non-cooperation).

Proof:When β 11 β )1(, +<+−−> CLrRCLR 221 , (β CLR 11 )×−−− β )1[( CLrR 221 >−−+− 0] ，－（），， and β −− CLR 11 +[ β )1( −−+− CLrR 221 ]<0 that is, detJ>0 trJ<0 in the equilibrium point E3（1，0）. At this time, through stable analysis to the equilibrium point, we can know from table （，），（－ 8 that E3 1 0 is the evolutionary stable strategy (ESS) of the system. And when β R< + CL 11 1 ） β R+r< + CL 221 , we can know from table 8 that E2（0，1）is the unstable point of the system, and

E1（0，0）、E4（1，1）are the saddle points (table 11), so E3（1，0）is the only the evolutionary stable strategy (ESS) of the system. (Completion of proof)

60 Journal of Applied Business and Economics vol. 16(3) 2014 TABLE 11 THE LOCAL STABILITY ANALYSIS’ RESULT SHOWING THE EQUILIBRIUM POINT OF PROPOSITION 7

equilibrium detJ’s sign trJ’ sign local stability points E （0，0）－ ± 1 saddle point （，） + + E2 0 1 unstable point E （1，0） + － 3 ESS E （1，1）－ ± 4 saddle point

The proposition shows that: When the profits that the MNCs gain through technology transfer with host country enterprises exceed a certain value β R< + CL 11 , they tend to transfer their technologies; when the profits that the host country enterprises gain through cooperation with the MNCs is less than a （－） certain value 1 β R+r> ++ CLL 2212 , host country enterprises tend to cooperate with the MNCs. ，（－） Proposition 8 when β R< + CL 11 1 β R+r> ++ CLL 2212 , that is, when ，（，） 21211 −+++> rCCLLR 11 β (/)( −−+<<+ 221 /) RCLrRRCL , E4 1 1 is locally and asymptotically stable, the evolutionary stable strategy (ESS) of the system is (transfer, cooperation). ，（－），（ Proof: (1) When β R> + CL 11 + CL 221 < 1 β R+r< ++ CLL 2212 β −− CLR 11 ），）， [ β )1( −−+− CLrR 221 ]>0 and {(β −−− CLR 11 +[ β )1( −−+− CLrR 221 ]}<0 that is, detJ>0，trJ<0 in the equilibrium point E4（1，1）. At this time, through stable analysis to the equilibrium point, we can know from table 8 that E4（1，1）is the evolutionary stable strategy (ESS) of the system. And we can know from table 8 that E2（0，1）is the unstable point of the system, and

E1（0，0）、E3（1，0）are the saddle points (table 12), so E4（1，1）is the only the evolutionary stable strategy (ESS) of system.

TABLE 12 THE LOCAL STABILITY ANALYSIS’S RESULT OF THE EQUILIBRIUM POINT OF PROPOSITION 8(1)

equilibrium detJ’s sign trJ’ sign local stability points E （0，0）－ ± 1 saddle point （，） + + E2 0 1 unstable point E （1，0）－ ± 3 saddle point E （1，1） + － 4 ESS

Journal of Applied Business and Economics vol. 16(3) 2014 61 ，（－）－（ (2) when β R< + CL 11 1 β R+r> ++ CLL 2212 , β −− CLR 11 ），（） [ β )1( −−−+− CLLrR 2212 ]>0 and β −− CLR 11 [ β )1( −−−+− CLLrR 2212 ]<0, that is, detJ>0，trJ<0 in the equilibrium point E4（1，1）. At this time, we can know from table 8 that

E4（1，1） is the evolutionary stable strategy (ESS) of the system. And when β R< + CL 11 ，（1－ β

）R+r> + CL 221 , we can know from table 8 that E4（1，1）is the unstable point of the system, and

E2（0，1）、E3（1，0）are the saddle points (table 13), so E4（1，1）is the only the evolutionary stable strategy (ESS) of the system. (Completion of proof)

TABLE 13 THE LOCAL STABILITY ANALYSIS’ RESULT CONCERNING THE EQUILIBRIUM POINT OF PROPOSITION 8(2)

equilibrium detJ’s sign trJ’ sign local stability points E （0，0） + + 1 unstable point saddle （，）－ ± E2 0 1 point E （1，0）－ ± 3 saddle point E （1，1） + － 4 ESS

The proposition 8 was proved through summing up of (1) and (2). The proposition shows that: When the profits that the MNCs gain through technology transfer with host country enterprises exceed a certain value β R< + CL 11 , they tend to transfer their technologies; when the profits that the host country enterprises gain through cooperation with the MNCs exceed a （－） certain value 1 β R+r> ++ CLL 2212 , the host country enterprises tend to cooperate with the MNCs. Then the MNCs cooperate with the host country enterprises successfully, namely, they come to Comprehensive cooperation.

THE DYNAMIC AND EVOLUTIONARY DIAGRAM OF GAMES BETWEEN MNCS AND HOST COUNTRY ENTERPRISES

The following is to further confirm the equilibrium point which has been confirmed and to show the evolutionary track from the different initial value point to the equilibrium point with the method of numerical simulation. This paper uses the MATLAB 7.0 software to make a numerical simulation analysis. The initial value is taken from [0.2，0.8]，[0.4，0.6]，[0.3，0.3]，[0.7，0.4]and [0.9，0.2]，time quantum is [0,100], Lateral Axis and longitudinal axis separately represents p and q, and in the space of [0，1]×[0，1], the dynamic evolutionary process is described from five different initial points to each equilibrium point.

62 Journal of Applied Business and Economics vol. 16(3) 2014 FIGURE 1 PROPOSITION 1 DYNAMIC EVOLUTIONARY FIGURE OF DIFFERENT INITIAL VALUE

=0.55； =2.1； =0.3； =2； =1； =1.5 Hypothesize parameter β R r L1 C1 C2

The figure shows that, given the proportion p of technology transfer strategy that is adopted by the MNCs and given the proportion q of cooperation strategy that is adopted by the host country enterprises, their different initial value point[p，q] trends toward the path of equilibrium point(0，0): when the p of the initial value point[p，q] is too large, its evolutionary path is that, first, decreasing the value p quickly ( that is, the proportion of technology transfer strategy that is adopted by MNCs decreases quickly), then decreasing the collective value q (that is, the proportion of cooperation strategy that is adopted by the host country enterprises decreases), so it tends to the equilibrium point(0，0).

FIGURE 2 PROPOSITION 2 DYNAMIC EVOLUTIONARY FIGURE OF DIFFERENT INITIAL VALUE

=0.55； =2.1； =0.3； =2； =1； =1.5 Hypothesize parameter β R r L1 C1 C2

The figure shows that, different initial value points[p，q] trend toward the path of the equilibrium point(1，0): when the p of the initial value point[p，q] is too large, its evolutionary path is that, first, decreasing the value p quickly ( that is, the proportion of technology transfer strategy that is adopted by MNCs decreases quickly), then increasing the collective value q (that is, the proportion of cooperation

Journal of Applied Business and Economics vol. 16(3) 2014 63 strategy that is adopted by the host country enterprises increases), so it tends to the equilibrium point(0，1).

FIGURE 3 PROPOSITION 3 DYNAMIC EVOLUTIONARY FIGURE OF DIFFERENT INITIAL VALUE

=0.7； =5； =0.3； =2； =1； =2 Hypothesize parameter β R r L1 C1 C2

The figure shows that, different initial value points[p，q] trend toward the path of equilibrium point(1，0): when the p of the initial value point[p，q] is too small, its evolutionary path is that, first, increasing the value p quickly ( that is, the proportion of technology transfer strategy that is adopted by MNCs increases quickly), then decreasing the collective value q (that is, the proportion of cooperation strategy that is adopted by the host country enterprises decreases), so it tends to the equilibrium point(1，0).

FIGURE 4 PROPOSITION 4 DYNAMIC EVOLUTIONARY FIGURE OF DIFFERENT INITIAL VALUE

=0.7； =4.5； =0.3； =2； =1； =1.5 Hypothesize parameter β R r L1 C1 C2

The figure shows that, the different initial value point[p，q] trends toward the path of the equilibrium point(1，1): If the initial value point[p，q] is close to the equilibrium point(1，1), it will approach the

64 Journal of Applied Business and Economics vol. 16(3) 2014 value of the equilibrium point quickly(such as p or q is 0.7, 0.8, 0.9, so the value will be increased up to 1 quickly) ; Contrarily, if it is far from the corresponding value’s point of equilibrium point(1，1), it become slow to approach the value of equilibrium point(such as p or q is 0.2, 0.3, so the value will be increased up to 1 slowly).

FIGURE 5 PROPOSITION 5 DYNAMIC EVOLUTIONARY FIGURE OF DIFFERENT INITIAL VALUE

=0.55； =3； =0.3； =2； =0.5； =0.8； =1； =1.5 Hypothesize parameter β R r L1 L21 L2 C1 C2

The figure shows that, different initial value points [p，q] all trend toward the path of equilibrium point(0，0) slowly.

FIGURE 6 PROPOSITION 6 DYNAMIC EVOLUTIONARY FIGURE OF DIFFERENT INITIAL VALUE

=0.5； =5.5； =0.3； =2； =0.5； =0.8； =1； =1.5 Hypothesize parameter β R r L1 L21 L2 C1 C2

The figure shows that, the different initial value point[p，q] trends toward the path of equilibrium point(0，1): when the q of initial value point[p，q] is too small, its evolutionary path is that, first, increasing the value q quickly ( that is, the proportion of cooperation strategy that is adopted by host country enterprises increases quickly), then decreasing the collective value p (that is, the proportion of

Journal of Applied Business and Economics vol. 16(3) 2014 65 technology transfer’s strategy that is adopted by the MNCs decreases), so it tends to the equilibrium point(0，1).

FIGURE 7 PROPOSITION 7 DYNAMIC EVOLUTIONARY FIGURE OF DIFFERENT INITIAL VALUE

=0.7； =5.5； =0.3； =2； =0.5； =0.8； =1； =1.5 Hypothesize parameter β R r L1 L21 L2 C1 C2

The figure shows that, different initial value point[p，q] trends toward the path of equilibrium point(1，0): when the p of initial value point[p，q] is too small, its evolutionary path is that, first, increasing the value p quickly ( that is, the proportion of strategy that is adopted by MNCs increases quickly), then decreasing the collective value q (that is, the proportion of cooperation strategy that is adopted by the host country enterprises decreases), so it tends to the equilibrium point(1，0).

FIGURE 8 PROPOSITION 8 DYNAMIC EVOLUTIONARY FIGURE OF DIFFERENT INITIAL VALUE

=0.6； =5.5； =0.3； =2； =0.5； =0.8； =1； =1.5 Hypothesize parameter β R r L1 L21 L2 C1 C2

The figure shows that, different initial value points [p，q] all trend toward the path of the equilibrium point(1，1) slowly.

66 Journal of Applied Business and Economics vol. 16(3) 2014 COMPARATIVE ANALYSES AND CONCLUSION

On the issue of technology transfer between MNCs and host country enterprises, this paper has established the evolutionary game model and has discussed separate issues based on two situations: the host country enterprises had research and development capacity or they had not. We make a comparative analysis to the equilibrium result of the two situations, and as a result, we drew useful conclusions and implications, as follows: 1. The host country enterprises have research and development capacity. Their unwillingness to cooperate is stronger than before when they did not this capacity, and the conditions are better than before. Such MNCs have not entered the host country. If the host country enterprises had no research and development capacity, their willingness to cooperate was poor; if the host country enterprises had this capacity, their willingness to cooperate would be strengthened. 2. Before MNCs and host country enterprises cooperate, their profitability is not related directly to the later cooperation between the two sides and their successful cooperation. 3. Because the last evolutionary stable strategy that this paper discusses is based on cooperation between the two sides, as long as one side of the MNCs or host country enterprises select not to cooperate the end result is that the cooperation cannot be realized. Therefore, the cooperation can be realized only when the host country enterprises are lack of research and development capacity, and at same time the condition of proposition 4 is satisfied, namely, the profits that MNCs gain through technology transfer with host country enterprises exceed the total costs. It is necessary to indicate that the cost of technology transfer and the loss caused by technology spillover should be included in the total costs. When the profits that host country enterprises gain through cooperation with MNCs exceed the needed cost, both MNCs and host country enterprises select the cooperation strategy, then they come to cooperate. If the host country enterprises have research and development capacity, the condition of proposition 8 should be satisfied: the profits that MNCs gain through technology transfer with host country enterprises (cooperation) exceed the total costs, which contain the cost of technology transfer and the loss caused by technology spillover. In addition, when the profits that host country enterprises gain through cooperation with MNCs exceed the total costs, which contain the needed cost and loss caused by their cooperation but affecting independent innovation, both MNCs and host country enterprises select the cooperation strategy, and then successful cooperation comes true. 4. Comparing proposition 4 with proposition 8, we discover that, given comprehensive cooperation between MNCs and host country enterprises, if the host country enterprises have research and development capacity, the conditions for cooperation will be improved. We can state clearly, according to numerical modeling, that if the host country enterprises have research and development capacity, a successful cooperation should meet the following conditions: 1) After cooperation, the total profits are more than before, namely, the cooperative scale becomes bigger or the technology becomes more advanced; 2) through cooperation with MNCs, the host country enterprises’ profits caused by improvement of technology capacity become more; 3) (And this is a very important point) There will be a change of equity structure. The host country enterprises have research and development capacity, so the proportion of host country enterprises’ return on equity will be decreased, the corresponding proportion of host country enterprises’ return on equity will be increased.

REFERENCES

Das, S. (1987). Externalities and technology transfer through multinational corporations. Journal of International Economics, 90：1142-1165. Magnus, B. & Jianye Wang (1992). Foreign investment and technology transfer：a simple model. European Economic Review, 36(1)：137-155. Longying Hu, Zhixin Tang (2000). Research on cooperative game model of international technology transfer’ behaviour. Journal of Harbin Institute of Technology, 32(2): 70-73.

Journal of Applied Business and Economics vol. 16(3) 2014 67 Yiyun Zhang (2002). Analysis on both sides’ cooperative game of the MNCs’ technology transfer. Journal of Northwestern Polytechnic University (Social Sciences), 22(2): 22-25. Webull, J. (1995). Evolutionary game theory. Princeton: Princeton Press. Friedman, D.(1998). On Economic Applications of Evolutionary Game Theory. Journal of Evolutionary Economics ,8,15-43.

This research is sponsored by Guangdong Province Social Science Scheme Program (GD12XGL25) , Shantou University Innovation Team Program (ITC12003) and Shantou University Cultivating Program for National Fund (NFC13006).

68 Journal of Applied Business and Economics vol. 16(3) 2014

Factor Decomposition of the Gender–Job Satisfaction Paradox: Evidence from Japan

Shiho Yukawa Keio Advanced Research Centers, Keio University

Yuki Arita Bank of Tokyo-Mitsubishi UFJ

INTRODUCTION

In this paper, we examine the gender differences in job satisfaction by using a data set that combines personnel data from an anonymous company and data from a survey of the company’s employees. In addition, we examine whether the “gender–job satisfaction paradox” exists in Japan and reveal the main factors contributing to this paradox. Job satisfaction has received significant attention from economists in recent years. Many previous studies focused on specific factors such as wage, race, education, country, union membership, training and firm size1. Gender difference, especially, is an important factor in job satisfaction. Sloane and Williams (2000) compared job satisfaction between men and women and showed that the factors affecting job satisfaction differ for men and women even when they have the same work environment. In addition, with respect to gender differences in job satisfaction, previous studies found that although women have disadvantages in terms of wage and working conditions in labor markets, they derive more satisfaction from their work than do men. This is called the gender–job satisfaction paradox, which is observed in Western countries such as the United States, the United Kingdom and others. Previous studies identified several factors that lead to the gender–job satisfaction paradox including gender differences in preferences and work–life balance policies. Gender difference in job satisfaction is an important research area in the recent literature in job satisfaction and women’s labor supply. In Japan, the supply of female labor supply has been increasing in recent years; however, compared to the United States and the United Kingdom, the work environment for women in Japan has not improved enough. If women indeed have significant disadvantages in terms of wage and working conditions than do men in labor markets, then the mere observation of higher job satisfaction in women than in men does not necessarily indicate the existence of the gender–job satisfaction paradox. Thus, for this reason, it is still unclear whether the paradox exists in Japan. In this

Journal of Applied Business and Economics vol. 16(3) 2014 69 paper, we therefore examine whether the relationship between job satisfaction and its factors is structurally different for men and women in Japan and whether the gender–job satisfaction paradox exists in Japan. There is currently no detailed study that uses the Oaxaca–Ransom decomposition technique to examine the gender difference in job satisfaction in Japan. The results obtained from our study may therefore be useful for policies on personnel and women’s labor supply. This paper is organized as follows. Section 2 provides a survey of the related literature. Section 3 provides the information from the anonymous company that cooperated in this study and a detailed description of the data. Section 4 provides the empirical results, while Section 5 provides the Oaxaca– Ransom decomposition results. Section 6 presents the conclusion of this study.

LITERATURE REVIEW

There have been many studies on the gender–job satisfaction paradox in other countries. For instance, Sousa-Poza and Sousa-Poza (2000) used cross-country data from Western countries, Israel and Japan to examine gender differences in job satisfaction. They have found that the gender–job satisfaction paradox exists in the United States, the United Kingdom and Switzerland, but not in Japan. Kaiser (2007) used data from 14 European Union countries and found that the paradox exists in these countries. In addition to these researches, some previous studies examined the factors affecting the gender–job satisfaction paradox. Clark (1997) used data from the British Household Panel Survey and found that women have lower job expectations, which leads to the gender–job satisfaction paradox. In addition, Clark (1997) pointed out that increasing the women’s labor supply would remove this paradox. Sousa-Poza and Sousa- Poza (2003) used pooled data for 10 years for the United Kingdom and showed that the gender–job satisfaction paradox decreased every year owing to the decreasing level of women’s job satisfaction. Bender et al. (2005) found that women in workplaces occupied by women express higher job satisfaction because such workplaces provide more job flexibility. Sloan and Williams (2000) suggested that women are more likely to choose jobs that provide high satisfaction than do men, which leads to the gender–job satisfaction paradox. In addition, they found that the metrology of choosing jobs is different for men and women, but that both men and women choose jobs that maximize their utility. In addition, they were unable to examine the gender differences in job satisfaction. In this study, we use factor analysis and make a continuous index of job satisfaction, enabling us to use ordinary least squares (OLS) estimation and the Oaxaca–Ransom decomposition technique. By using the Oaxaca–Ransom decomposition technique, we can examine the mechanism that leads to the gender–job satisfaction paradox. We also decompose job satisfaction into explained components and unexplained components using the Oaxaca– Ransom decomposition technique, enabling us to find the gender–job satisfaction paradox that has not been observed in the previous studies in Japan.

DATA

Personnel Data and Employee Survey Data We use a unique data set composed of employee survey data and personnel data provided by the anonymous company that cooperated in this study. We call this anonymous company “Firm Z.” Firm Z is a long-established consumer products manufacturing firm with about 300 employees, and is a representative firm of the area. In recent years, Firm Z saw a rapid increase in sales, and gained national reputation as a brand. Because of this growth, the number of Firm Z employees has been increasing in recent years; in particular, it has recruited young workers in their 20s and now they comprises 40% of the total employees. The employee survey used in this study was conducted in 2008. The survey was administered to all employees of Firm Z; it consists of items on issues such as job satisfaction, promotion and payment, their awareness of how the workplace is run, and the role of the workplace in job satisfaction. In addition, the survey asked 134 items related to productivity, especially with regard to the employees’ direct departments. The survey’s collection rate is 100%; this high collection rate enabled the accumulation of abundant information related to the workplace. Therefore, by using this survey, we can

70 Journal of Applied Business and Economics vol. 16(3) 2014 analyze the factors that affect the gender–job satisfaction paradox in more detail than in other studies. In addition, this survey was synced with the employees’ ID codes, thus enabling us to combine their personnel data and their answers. Clark (1997) and Sousa-Poza and Sousa-Poza (2007) pointed out a problem in having a sample selection of women. According to these researchers, since women are more likely to leave their jobs than men are, only women with higher job satisfaction remain in the workplace. However, the turnover rate at Firm Z is very low; thus, this sample selection problem may not be a significant issue in this study.

Overall Job Satisfaction and Treatment Job Satisfaction We categorize job satisfaction into overall job satisfaction and treatment job satisfaction. Overall job satisfaction is a comprehensive measure of job satisfaction. The survey contains nine statements about overall job satisfaction to which the employees answered “Absolutely no,” “No,” “Neither,” ”Yes,” or “Absolutely yes.” On the other hand, treatment job satisfaction is job satisfaction associated with job responsibilities, wage, and job position. The survey contains seven items about treatment job satisfaction to which the employees answered “Very dissatisfied,” “Dissatisfied,” “Neither,” “Satisfied,” or “Very satisfied.” Table 1 shows each of the statements and items, the average answers by gender, and the results of the independent t-test for men and women. With respect to overall job satisfaction, the men expressed higher job satisfaction than did the women except for the statement “If I work hard in this company, my effort will be rewarded.” In particular, the men expressed higher job satisfaction in the following three statements: “My job responsibilities are of use to customers and the society,” “I feel that my job responsibilities are interesting,” and “Compared to the previous year, my job-related abilities have progressed.” For these three statements, we found a significant mean difference in overall job satisfaction between men and women. In contrast, for the items on treatment job satisfaction, the women expressed higher job satisfaction than did the men except for the item “opportunity for training.” For four items, namely, “appropriate amount of job responsibilities,” “work hours,” “wage,” and “job grade,” we found a significant mean difference between men and women. Therefore, the results show that on average, men express higher overall job satisfaction for worthwhile and fruitful work than do women, but women express higher treatment job satisfaction than do men. These results suggest that factors such as work fulfillment should be separated from factors regarding wage and working conditions when comparing job satisfaction between men and women. In addition, they indicate that overall job satisfaction and treatment job satisfaction stem from the different preferences of men and women.

Journal of Applied Business and Economics vol. 16(3) 2014 71 TABLE 1 SURVEY ITEMS AND AVERAGE ANSWERS TO EACH ITEM

Women Men Mean difference Overall job satisfaction I feel more wanted in this company. 3.344 3.467 -0.123 I would recommend this company to my friends and relatives. 2.857 2.878 -0.0207 My job responsibilities are of use to customers and the society. 3.516 3.773 -0.257** I feel that my job responsibilities are interesting. 2.984 3.358 -0.374** I realize my own growth through my job responsibilities. 3.516 3.681 -0.165 I take pleasure in going to work. 2.969 2.996 -0.0269 If I work hard in this company, my effort will be rewarded. 2.594 2.530 0.0633 Compared to the previous year, my job-related abilities have progressed. 3.328 3.596 -0.268** I am satisfied working for this company. 3.234 3.448 -0.213 Treatment job satisfaction Description of job responsibilities 3.397 3.346 0.0503 Appropriate amount of job responsibilities 3.313 3.00437 0.308** Opportunity for training 2.902 2.903 -0.00145 Work hours 3.641 3.215 0.426** Wage 3.094 2.515 0.578*** Job position 3.210 2.987 0.223 Job grade 3.081 2.812 0.268* Significance levels: * denotes 10%; **, 5%; ***, 1%

Dependent and Independent Variables In this paper, we extract the factors that significantly affect overall job satisfaction and treatment job satisfaction from the employees’ answers to the nine overall job satisfaction statements and seven treatment job satisfaction items. We use these extracted factors for each index of job satisfaction. Specifically, we assume that both overall job satisfaction and treatment job satisfaction consist of several basic concept linear sums; we then extract the basic concept or common factor by using factor analysis. The specific observed variables are the nine statements for overall job satisfaction and the seven items for treatment job satisfaction. The factor loading is the value obtained by the promax method. We use a scree plot to determine the criteria for the number of factors. Figure 1 shows the scree plot for each type of job satisfaction; the plot shows that on the number of factors, the line gradually starts decreases between two and three. The vertical axis corresponds to the Eigen value and the horizontal axis corresponds to the number of factors. Therefore, we use factor numbers two and three, respectively. Table 2 shows the results of the factor analysis; the results show that for both overall job satisfaction and treatment job satisfaction, the variance contribution ratio decreases to factor one from factor two; thus, in this paper, we use factor one as an dependent variable. Table 2 also shows that with respect to factor one of overall job satisfaction, the statements “I would recommend this company to my friends and relatives” (statement 2), “If I work hard in this company, my effort will be rewarded” (statement 7), and “I am satisfied working for this company” (statement 9) have high loading factors. These items express the employees’ overall satisfaction with the company. In addition, with respect to treatment job satisfaction, the items “job position” (item 6) and “job grade” (item 7) have high loading factors, and we extracted these factors as related to the degree of promotion. Since these factors are continuous variables, we can use the Oaxaca–Ransom decomposition method.

72 Journal of Applied Business and Economics vol. 16(3) 2014 FIGURE 1A OVERALL JOB SATISFACTION 4 3

2 1

Eigenvalue 0 -1 1 2 3 4 5 6 7 8 9 Number

FIGURE 1B TREATMENT JOB SATISFACTION 3

1 Eigenvalue 0

-1 1 2 3 4 5 6 7 Number

Journal of Applied Business and Economics vol. 16(3) 2014 73 TABLE 2 RESULTS OF FACTOR ANALYSIS

Number Item F1 F2 F3 Overall job satisfaction 1 I feel more wanted in this company. 0.140 0.518 2 I would recommend this company to my friends and relatives. 0.676 0.0366 3 My job responsibilities are of use to customers and the society. 0.00810 0.552 4 I feel that my job responsibilities are interesting. 0.364 0.527 5 I realize my own growth through my job responsibilities. 0.201 0.540 6 I take pleasure in going to work. 0.500 0.294 7 If I work hard in this company, my effort will be rewarded. 0.640 -0.0270 8 Compared to the previous year, my job-related abilities have progressed. -0.0198 0.551 9 I am satisfied working for this company. 0.644 0.195 Characteristic root 3.616 0.423 Variance contribution 1.00360 0.118 Treatment job 1 Description of job responsibilities 0.0913 -0.0249 0.632 2 Appropriate amount of job responsibilities -0.00100 0.0445 0.620 3 Opportunity for training 0.346 0.140 0.00330 4 Work hours -0.175 0.419 0.195 5 Wage 0.393 0.402 -0.0106 6 Job position 0.827 -0.0156 0.0736 7 Job grade 0.852 -0.0231 -0.0279 Characteristic root 2.342 0.709 0.121 Variance contribution 0.892 0.270 0.046 F1 indicates factor1.F2 indicates factor2 .Factor3 indicates factor3.

In this paper, we use the dependent variable obtained by factor analysis and the independent variable for the objective variables such as employee wage and overtime hours. In addition, we use the subjective variable derived from the employee survey items related to working condition and job responsibility equality. Specifically, we use the log wage and log relative wage. We use the relative wage made by same gender, age, and education. The other objective variables we used as independent variables are the education dummy variables and overtime hours. We also use the employees’ health condition obtained from the results of the employee survey as independent variables. In addition to these independent variables, we use the results of the employee survey items related to job satisfaction as independent variables. In overall job satisfaction, the responses to the working condition and workplace relationship items are used as independent variables. More precisely, we use the individual factor variables such as “Are you strongly aware of your achievement in workplace?” (recognition of achievement) and “How many hours in a year do you devote to off-the-job training to improve your skill?” (off-the-job training), and workplace relation variables such as “Do employees in your workplace communicate with each other?” (communication in the workplace), “Does your supervisor usually observe how you work and your ability, and provide you with appropriate and timely advice?” (supervisor advice), and “Do you know someone working close to you who was forced by the supervisor to do more than what is necessary in job responsibilities?” (supervisor harassment). In treatment job satisfaction, we use the results of the employee survey on “Are you satisfied with your job responsibilities?” (satisfaction with job responsibilities), “Does your workplace allocate job responsibilities fairly according to each person’s ability?” (fairness of job responsibility allocation), and “How many hours in a year do you devote to off-the-job training to improve your skill? (off-the-job training)” as independent variables. With respect to off-the-job training, we eliminate the samples whose mean value for both men and women is more than triple the standard deviation.

74 Journal of Applied Business and Economics vol. 16(3) 2014 Descriptive Statistics Tables 3 and 4 show the descriptive statistics of overall job satisfaction and treatment job satisfaction, respectively. In treatment job satisfaction, the women expressed higher job satisfaction than did the men, a result that is statistically significant at the 10% level (Table4). In contrast, the men expressed high job satisfaction than did the women with respect to overall job satisfaction, but this result is not statistically significant. With respect to wage, the mean of the men’s wage is higher than that of the women’s wage, a difference that is statistically significant at the 1% level for both overall job satisfaction and treatment job satisfaction. In addition, the mean value of the men’s off-the-job training and overtime hours are higher than those for the women, differences that are statistically significant for both overall job satisfaction and treatment job satisfaction. For the subjective variables, namely health condition, recognition of achievement, supervisor harassment and fairness of job responsibility allocation, we observed a statistically significant mean difference between men and women. On average, the men expressed more health anxiety than did the women. In addition, on average, the men were more aware of their achievements than were the women, and felt more harassment from their supervisors than did the women.

TABLE 3 DESCRIPTIVE STATISTICS OF OVERALL JOB SATISFACTION

Total Men Women Difference of mean Dependent variable Overall job satisfaction -0.027 0.007 -0.156 -0.1632 (0.878) (0.874) (0.890) (0.1299) Independent variables Objective data Wage 12.485 12.526 12.336 -0.190*** (0.291) (0.301) (0.189) (0.0416) Relative wage -0.009 -0.012 0.004 0.0161 (0.132) (0.142) (0.084) (0.0196) Educational status Junior high school and high 0.526 0.523 0.534 0.0111 school (0.500) (0.501) (0.503) (0.0742) Junior college 0.040 0.028 0.086 0.0582** (0.197) (0.165) (0.283) (0.0291) University 0.331 0.355 0.241 -0.1138 (0.471) (0.480) (0.432) (0.0696) Graduate school 0.103 0.093 0.138 0.0445 (0.304) (0.292) (0.348) (0.0451) Off-the-job training 1.634 1.745 1.224 -0.521** (1.575) (1.601) (1.412) (0.2314) Overtime 0.939 1.088 0.388 -0.700*** (1.159) (1.196) (0.808) (0.1666) Standard deviation in parentheses. Significance levels: * denotes 10%; **, 5%; ***, 1%

Journal of Applied Business and Economics vol. 16(3) 2014 75 TABLE 3 DESCRIPTIVE STATISTICS OF OVERALL JOB SATISFACTION (CONTINUED)

Total Men Women Difference of mean Questionnaire data Supervisor advice Answer 1 0.143 0.140 0.155 0.0150 (0.351) (0.348) (0.365) (0.0521) Answer 2 0.404 0.407 0.397 -0.0100 (0.492) (0.492) (0.493) (0.0729) Answer 3 0.099 0.107 0.069 -0.0385 (0.300) (0.310) (0.256) (0.0444) Answer 4 0.279 0.266 0.328 0.0612 (0.450) (0.443) (0.473) (0.0666) Answer 5 0.074 0.079 0.052 -0.0277 (0.261) (0.271) (0.223) (0.0387) Communication_in the workplace Answer 1 0.088 0.089 0.086 -0.0026 (0.284) (0.285) (0.283) (0.0421) Answer 2 0.338 0.322 0.397 0.0741 (0.474) (0.469) (0.493) (0.0702) Answer 3 0.088 0.093 0.069 -0.0245 (0.284) (0.292) (0.256) (0.0421) Answer 4 0.368 0.369 0.362 -0.0071 (0.483) (0.484) (0.485) (0.0716) Answer 5 0.118 0.126 0.086 -0.0400 (0.323) (0.333) (0.283) (0.0478) Standard deviation in parentheses. Significance levels: * denotes 10%; **, 5%; ***, 1% Answer1 indicates “Absolutely no”. Answer2 indicates “No”. Answer3 indicates “Neither”. Answer4 indicates ”Yes”. Answer5 indicates “Absolutely yes”.

76 Journal of Applied Business and Economics vol. 16(3) 2014 TABLE 3 DESCRIPTIVE STATISTICS OF OVERALL JOB SATISFACTION (CONTINUED)

Total Men Women Difference of mean Health condition Answer 1 0.125 0.117 0.155 0.0383 (0.331) (0.322) (0.365) (0.0491) Answer 2 0.272 0.248 0.362 0.114* (0.446) (0.433) (0.485) (0.0658) Answer 3 0.063 0.061 0.069 0.0082 (0.243) (0.239) (0.256) (0.0360) Answer 4 0.338 0.336 0.345 0.0084 (0.474) (0.474) (0.479) (0.0703) Answer 5 0.202 0.238 0.069 -0.169*** (0.402) (0.427) (0.256) (0.0588) Supervisor harassment Answer 1 0.129 0.107 0.207 0.0994** (0.335) (0.310) (0.409) (0.0494) Answer 2 0.324 0.299 0.414 0.115* (0.469) (0.459) (0.497) (0.0692) Answer 3 0.096 0.107 0.052 -0.0558 (0.295) (0.310) (0.223) (0.0436) Answer 4 0.320 0.332 0.276 -0.0559 (0.467) (0.472) (0.451) (0.0692) Answer 5 0.132 0.154 0.052 -0.102** Standard deviation in parentheses. Significance levels: * denotes 10%; **, 5%; ***, 1% Answer1 indicates “Absolutely no”. Answer2 indicates “No”. Answer3 indicates “Neither”. Answer4 indicates ”Yes” . Answer5 indicates “Absolutely yes”.

Journal of Applied Business and Economics vol. 16(3) 2014 77 TABLE 3 DESCRIPTIVE STATISTICS OF OVERALL JOB SATISFACTION (CONTINUED)

Total Men Women Difference of mean

Recognition of

achievement Answer 1 & 2 0.188 0.168 0.259 0.0904 (0.391) (0.375) (0.442) (0.0577) Answer 3 0.107 0.098 0.138 0.0398 (0.309) (0.298) (0.348) (0.0458) Answer 4 0.540 0.537 0.552 0.0143 (0.499) (0.500) (0.502) (0.0740) Answer 5 0.165 0.196 0.052 -0.145*** (0.372) (0.398) (0.223) (0.0545)

Observations 272 214 58 Standard deviation in parentheses. Significance levels: * denotes 10%; **, 5%; ***, 1% Answer1 indicates “Absolutely no”. Answer2 indicates “No”. Answer3 indicates “Neither”. Answer4 indicates ”Yes” . Answer5 indicates “Absolutely yes”.

78 Journal of Applied Business and Economics vol. 16(3) 2014 TABLE 4 DESCRIPTIVE STATISTICS OF TREATMENT JOB SATISFACTION

Total Men Women Difference of mean Dependent variable Treatment job satisfaction -0.033 -0.089 0.171 0.260* (0.894) (0.918) (0.772) (0.1339) Independent variables Objective data Wage 12.480 12.520 12.334 -0.186*** (0.288) (0.298) (0.187) (0.0419) Relative wage -0.010 -0.013 0.002 0.01489 (0.132) (0.142) (0.084) (0.0199) Educational status Junior high school and high 0.525 0.527 0.518 -0.00871 school (0.500) (0.501) (0.504) (0.0755) Junior college 0.042 0.029 0.089 0.0603** (0.201) (0.168) (0.288) (0.0300) University 0.327 0.348 0.250 -0.09783 (0.470) (0.477) (0.437) (0.0707) Graduate school 0.106 0.097 0.143 0.04624 (0.309) (0.296) (0.353) (0.0465) Off-the-job training 1.623 1.727 1.239 -0.488** (1.576) (1.601) (1.427) (0.2359) Overtime 0.962 1.113 0.402 -0.711*** (1.164) (1.199) (0.819) (0.1701) Standard deviation in parentheses. Significance levels: * denotes 10%; **, 5%; ***, 1%

Journal of Applied Business and Economics vol. 16(3) 2014 79 TABLE 4 DESCRIPTIVE STATISTICS OF TREATMENT JOB SATISFACTION (CONTINUED)

Total Men Women Difference of mean Questionnaire data Fairness of responsibility allocation Answer 1 0.103 0.101 0.107 0.00569 (0.304) (0.303) (0.312) (0.0459) Answer 2 0.395 0.411 0.339 -0.07134 (0.490) (0.493) (0.478) (0.0738) Answer 3 0.141 0.116 0.232 0.116** (0.348) (0.321) (0.426) (0.0521) Answer 4 0.308 0.314 0.286 -0.02830 (0.463) (0.465) (0.456) (0.0698) Answer 5 0.053 0.058 0.036 -0.02226 (0.225) (0.234) (0.187) (0.0339) Satisfaction with job responsibilities Answer 1 0.046 0.043 0.054 0.01009 (0.209) (0.204) (0.227) (0.0315) Answer 2 0.186 0.179 0.214 0.03554 (0.390) (0.384) (0.414) (0.0588) Answer 3 0.163 0.169 0.143 -0.02622 (0.371) (0.376) (0.353) (0.0559) Answer 4 0.479 0.469 0.518 0.04926 (0.501) (0.500) (0.504) (0.0755) Answer 5 0.125 0.140 0.071 -0.06867 (0.332) (0.348) (0.260) (0.0499) Observations 263 207 56 Standard deviation in parentheses. Significance levels: * denotes 10%; **, 5%; ***, 1% Answer1 indicates “Absolutely no”. Answer2 indicates “No”. Answer3 indicates “Neither”. Answer4 indicates ”Yes” . Answer5 indicates “Absolutely yes”.

RESULTS

Results of Overall Job Satisfaction Table 5 shows the results for overall job satisfaction. The results show that although the men expressed higher overall job satisfaction than did the women, no statistically significant mean difference was found between men and women (Table3). In addition, the female dummy variable is not statistically significant. These results imply that there is no gender–job satisfaction paradox with regard to overall job satisfaction. The effects of the other independent variables on overall job satisfaction are as follows. Based on the results from the entire sample, having significant anxiety owing to one’s health condition negatively affects job satisfaction, a statistically significant result at the 1% level. This finding parallels

80 Journal of Applied Business and Economics vol. 16(3) 2014 those of previous studies that show bad health negatively affects job satisfaction. The results also show that the relationship with one’s supervisor is also important for overall job satisfaction. Supervisor advice significantly and positively affects overall job satisfaction, while supervisor harassment significantly and negatively affects overall job satisfaction. Having a sense of one’s achievement (recognition of achievement) and good communication in the workplace (communication in the workplace) significantly and positively affect overall job satisfaction. While wage positively affects overall job satisfaction for both men and women, the results also show other differences between men and women. For instance, for men, bad health negatively affects overall job satisfaction, while good communication in the workplace positively affects overall job satisfaction. In addition, having a sense of achievement is higher for men for overall job satisfaction. The relationship with one’s supervisor is another important factor for overall job satisfaction for men. Specifically, receiving advice from one’s supervisor positively affects overall job satisfaction, while harassment from one’s supervisor negative affects overall job satisfaction.

TABLE 5 ESTIMATION RESULTS FOR OVERALL JOB SATISFACTION

Independent Total Men Women variables Women -0.0415 (0.126) Wage 0.584*** 0.539** 1.463** (0.190) (0.209) (0.688) Relative wage -0.155 -0.0193 -1.974 (0.442) (0.512) (1.366) Overtime -0.0127 0.00173 -0.0875 (0.0448) (0.0487) (0.162) Standard errors in parentheses. Significance levels: * denotes 10%; **, 5%; ***, 1%

Journal of Applied Business and Economics vol. 16(3) 2014 81 TABLE 5 ESTIMATION RESULTS FOR OVERALL JOB SATISFACTION (CONTINUED)

Independent variables Total Men Women Education status2 Junior college -0.320 -0.0309 -0.720 (0.243) (0.308) (0.538) University -0.0797 -0.223* 0.702** (0.108) (0.120) (0.288) Graduate School -0.471*** -0.554** 0.0444 (0.176) (0.216) (0.326) Health condition Answer 2 0.00766 0.0816 -0.156 (0.147) (0.177) (0.299) Answer 3 -0.286 -0.344 -0.0156 (0.217) (0.257) (0.519) Answer 4 -0.166 -0.0139 -0.235 (0.153) (0.180) (0.270) Answer 5 -0.461*** -0.378* -0.387 (0.174) (0.198) (0.454) Recognition of achievement Answer 3 -0.0635 0.223 -0.681** (0.187) (0.203) (0.325) Answer 4 0.237* 0.326** 0.230 (0.127) (0.151) (0.253) Answer 5 0.616*** 0.702*** 0.247 (0.163) (0.175) (0.572) Off-the-job training 0.0577 0.0624 -0.0809 (0.0351) (0.0383) (0.0740) Standard errors in parentheses. Significance levels: * denotes 10%; **, 5%; ***, 1% The reference group of the recognition of achievement is “Absolutely no” and “No.” The reference group of the other dummy variables is “Absolutely no.” Answer 1 indicates “Absolutely no”; Answer 2 indicates “No”; Answer 3 indicates “Neither”; Answer 4 indicates ”Yes”; and Answer 5 indicates “Absolutely yes.” Age is controlled.

82 Journal of Applied Business and Economics vol. 16(3) 2014 TABLE 5 ESTIMATION RESULTS FOR OVERALL JOB SATISFACTION (CONTINUED)

Independent variables Total Men Women Communication in the workplace Answer 2 0.182 0.218 0.462 (0.201) (0.223) (0.699) Answer 3 0.395* 0.462* 0.367 (0.231) (0.246) (0.827) Answer 4 0.392* 0.482** 0.628 (0.200) (0.221) (0.667) Answer 5 0.612** 0.634** 1.126 (0.240) (0.257) (0.740) Supervisor advice Answer 2 0.320* 0.220 0.583 (0.165) (0.181) (0.558) Answer 3 0.499** 0.436* 0.294 (0.226) (0.254) (0.640) Answer 4 0.574*** 0.447** 0.715 (0.188) (0.212) (0.543) Answer 5 0.657*** 0.595** 0.468 (0.249) (0.258) (0.808) Supervisor harassment Answer 2 -0.174 -0.267 -0.135 (0.144) (0.178) (0.290) Answer 3 -0.237 -0.407* 1.219 (0.213) (0.232) (0.737) Answer 4 -0.342** -0.423** -0.475 (0.154) (0.184) (0.361) Answer 5 -0.710*** -0.771*** -0.941 (0.190) (0.218) (0.578) Constant -7.782*** -7.254*** -18.92** Observations 272 214 58 R-squared 0.399 0.419 0.666 Standard errors in parentheses. Significance levels: * denotes 10%; **, 5%; ***, 1%.The reference group of the dummy variables is “Absolutely no.” Answer 1 indicates “Absolutely no”; Answer 2 indicates “No”; Answer 3 indicates “Neither”; Answer 4 indicates ”Yes”; and Answer 5 indicates “Absolutely yes.” Age is controlled.

Journal of Applied Business and Economics vol. 16(3) 2014 83 Results of Treatment Job Satisfaction The results for treatment job satisfaction are shown in Table 6. For treatment job satisfaction, the women expressed higher treatment job satisfaction than did the men and the mean difference between the men and the women is statistically significant at the 10% level (Table4). In addition, the female dummy variable is statistically significant at the 5% level. These results indicate that even if we control for the other independent variables, women still express higher treatment job satisfaction than do men. Therefore, there is a gender–job satisfaction paradox in treatment job satisfaction. The effect of the other independent variables on treatment job satisfaction is as follows. First, we consider the results for the entire sample. Wages, satisfaction with job responsibilities, and fairness of job responsibility allocation significantly and positively affect treatment job satisfaction. In contrast, overtime hours significantly and negatively affect treatment job satisfaction. There are also gender differences in treatment job satisfaction. While wage significantly and positively affects treatment job satisfaction only for men, relative wage significantly and positively affects treatment job satisfaction only for women. This suggests that men place emphasis on the absolute wage, while women place emphasis on the relative wage. Finally, although satisfaction with job responsibilities and fairness of job responsibility allocation positively affect treatment job satisfaction for both men and women, these effects are stronger for women than for men.

TABLE 6 ESTIMATION RESULTS FOR TREATMENT JOB SATISFACTION

Independent variables Full Men Women Women 0.295** (0.124) Wage 0.466** 0.551** -0.231 (0.235) (0.267) (0.469) Relative wage 0.653 0.659 1.734* (0.460) (0.505) (0.881) Overtime -0.0841* -0.0690 -0.219 (0.0489) (0.0548) (0.136) Off-the-job training - 0.0149 -0.143 0.000200 (0.0377) (0.0421) (0.0888) Standard errors in parentheses. Significance levels: * denotes 10%; **, 5%; ***, 1%

84 Journal of Applied Business and Economics vol. 16(3) 2014 TABLE 6 ESTIMATION RESULTS FOR TREATMENT JOB SATISFACTION (CONTINUED)

Independent variables Full Male Female Educational status2 Junior college 0.0472 0.189 0.117 (0.227) (0.339) (0.310) University 0.262** 0.246* 0.364 (0.119) (0.136) (0.240) Graduate School 0.217 0.137 0.965*** (0.149) (0.181) (0.271) Satisfaction with job responsibilities Answer 2 0.332 0.336 0.557 (0.218) (0.276) (0.339) Answer 3 0.307 0.362 0.120 (0.215) (0.281) (0.287) Answer 4 0.497** 0.452* 1.029*** (0.211) (0.268) (0.299) Answer 5 0.698*** 0.740** 0.811* (0.259) (0.309) (0.415) Fairness of responsibility allocation Answer 2 0.468*** 0.435** 0.228 (0.168) (0.195) (0.338) Answer 3 0.762*** 0.779*** 0.818** (0.191) (0.223) (0.365) Answer 4 0.796*** 0.668*** 1.188*** (0.182) (0.209) (0.331) Answer 5 0.845*** 0.723** 0.939** (0.303) (0.353) (0.389) Constant -6.960** -7.989** 1.674 (2.952) (3.347) (5.718) Observations 263 207 56 R-squared 0.280 0.256 0.640 Standard errors in parentheses. Significance levels: * denotes 10%; **, 5%; ***, 1% The reference group of the dummy variables is “Absolutely no.” Answer 1 indicates “Absolutely no”; Answer 2 indicates “No”; Answer 3 indicates “Neither”; Answer 4 indicates ”Yes”; and Answer 5 indicates “Absolutely yes.” Age is controlled.

FACTOR DECOMPOSITION ANALYSIS

Oaxaca–Ransom Decomposition We use the results of the OLS estimation and decompose them into explained and unexplained differences using the Oaxaca–Ransom decomposition technique. Explained difference is the mean

Journal of Applied Business and Economics vol. 16(3) 2014 85 difference in each independent variable between men and women; this difference exists even if the coefficient is the same for men and women. Unexplained difference is the coefficient difference of each independent variable; this difference exists even if the mean is the same for men and women. The decomposition equation is as follows:

= ( ) + ( ) + ( ) ∗ ∗ ∗ The first푦� 퐹 term− 푦�푀 indicates훽 푥̅퐹 −the 푥̅ 푀explained훽 − 훽difference,푀 푥̅푀 훽while퐹 − 훽 the푥̅퐹 second and third terms indicate the unexplained difference. is the mean value of the overall job satisfaction or treatment job satisfaction. is the mean value of the independent variables, while is the coefficient that assumes there is no gender difference. Although there푦� are several methods to determine∗ the criteria parameter , we use the 푥̅, which is the OLS estimator of the entire sample, which훽 has been used in many previous ∗studies. By using∗ the Oaxaca–Ransom decomposition technique, we can examine the mechanism that leads훽 to the gender훽 – job satisfaction paradox found in this study.

Decomposition of Overall Job Satisfaction Table7 shows the decomposition results for overall job satisfaction. The difference in overall job satisfaction between men and women is composed mainly of the explained component. In addition, both the explained component and the unexplained component are not statistically significant. The results also show the detailed decomposition for overall job satisfaction. Wage is statistically significant in both the negative explained component and the positive unexplained component. In addition, wage is mainly composed of the positive unexplained component. These results suggest that an increase in wage contributes to the gender–job satisfaction paradox. In addition, having a sense of one’s achievement is mainly composed of the positive unexplained component, and contributes to the paradox. The explained component of health condition is positive and statistically significant, which shows that on average, men have more health anxieties than do women.

86 Journal of Applied Business and Economics vol. 16(3) 2014 TABLE 7 DECOMPOSITION RESULTS FOR OVERALL JOB SATISFACTION

Explained Unexplained Total -0.122 -0.0415 (0.101) (0.117) Wage -0.111*** 11.40* (0.0390) (6.675) Relative wage -0.00249 -0.00524 (0.00715) (0.0206) Overtime 0.00890 -0.0448 (0.0298) (0.0597) Educational status -0.0305 0.0674 (0.0296) (0.112) Health condition 0.0752** -0.117 (0.0302) (0.0719) Recognition of achievement -0.0882*** 0.163* (0.0332) (0.0859) Communication in the -0.0234 0.0278 workplace (0.0255) (0.0768) Supervisor advice -0.00547 0.141 (0.0313) (0.0876) Supervisor harassment 0.0852** -0.182 (0.0353) (0.118) Off-the-job training -0.0301 -0.178** (0.0213) (0.0877) Constant -11.31* (6.659) Observations 272 272 Standard errors in parentheses. Significance levels: * denotes 10%; **, 5%; ***, 1%

Decomposition of Treatment Job Satisfaction Table 8 shows the decomposition results for treatment job satisfaction. The explained component is negative, while the unexplained component is positive and statistically significant at the 1% level. The gender difference in treatment job satisfaction is mainly composed of the positive unexplained component. This result shows that the unexplained component contributes to the gender–job satisfaction paradox. The primary factor for the paradox is the constant term. Satisfaction with job responsibilities is another important factor for this paradox, a statistically significant result at the 10% level. Although relative wage also contributes to the gender–job satisfaction paradox, this result is not statistically significant.

Journal of Applied Business and Economics vol. 16(3) 2014 87 TABLE 8 DECOMPOSITION RESULTS FOR TREATMENT JOB SATISFACTION

Explained Unexplained Total -0.0814 0.643*** (0.178) (0.205) Wage -0.0885 -14.72 (0.0716) (10.25) Relative wage 0.0203 0.00432 (0.0232) (0.0342) Overtime 0.105* -0.110 (0.0580) (0.0850) Educational status -0.0416 -0.0900 (0.0378) (0.172) Satisfaction with job -0.0735 0.295* responsibilities (0.0809) (0.158) Fairness of responsibility 0.00994 -0.0802 allocation (0.0882) (0.167) Off-the-job training -0.0129 -0.329 (0.0300) (0.200) Constant 15.68 (10.22) Observations 263 263 Standard errors in parentheses. Significance levels: * denotes 10%; **, 5%; ***, 1%

CONCLUSION

In this paper, we examined whether the gender–job satisfaction paradox exists in Japan by using a combined data set composed of company personnel data and employee survey data. In addition, we use the Oaxaca–Ransom technique to reveal the main factors contributing to this paradox. We categorized job satisfaction into overall job satisfaction and treatment job satisfaction. For overall job satisfaction, we found that there is no gender–job satisfaction paradox. On the other hand, for treatment job satisfaction, there is a paradox. Meanwhile, the results from the Oaxaca–Ransom decomposition analysis revealed that with respect to treatment job satisfaction, the unexplained component leads to the gender–job satisfaction paradox. Moreover, the main factor contributing to this paradox is the constant term and the satisfaction with job responsibilities. Although this study yielded useful insights on gender differences, there are still questions regarding this issue that remain. First, because our survey did not include items about work–life balance, we were unable to examine the effect of the work–life balance policy on job satisfaction. According to Bender et al. (2005), the work–life balance variable should be considered in examining job satisfaction. Second, because our results were based on data from only one firm, the results may not necessarily be generalizable to all of Japan. Research on gender difference in job satisfaction should therefore be also conducted using a macro–level framework.

88 Journal of Applied Business and Economics vol. 16(3) 2014 ENDNOTES

2. Hamermesh (1977), Freeman (1978) and Clark and Oswald (1996) analyzed the linkage between wage and job satisfaction. Kim et al. (1996) and Clark (1997) took into account the effect of race, education level, gender, and country of residence. Miller (1990) and Bender and Sloane (1998) focused on the impact of participation in a union. Idson (1990) analyzed the impact of the size of a firm. Jones et al. (2009) analyzed the relationship between training and job satisfaction. 3. The reference group of education is junior high school and high school.

REFERENCES

Bender, K.A., Donohue, S.M. & Heywood, J.S. (2005). Job satisfaction and gender Segregation. Oxford Economic Papers, 57, 479–96. Bender, K.A. & Sloane, P.J. (1998). Job Satisfaction, Trade Unions, and Exit-Voice Revisited. Industrial and Labor Relations Review, 51, 222–241. Clark, A.E. (1997). Job Satisfaction and Gender: Why Are Women So Happy at Work? Labour Economics, 4,341–372. Clark, A.E. & Oswald, A.J. (1996). Satisfaction and Comparison Income. Journal of Public Economics, 61,359–381. Freeman, R.B. (1978) Job satisfaction as an Economic Variable. American Economic Review, 68,135– 141. Hamermesh, D.S. (1977). Economic aspects of job satisfaction. O. Ashenfelter & W. Oates (eds.), Essays in Labor Market Analysis, New York: Wiley, Inc. Idson, T.L. (1990). Establishment Size, Job Satisfaction and the Structure of Work. Applied Economics, 22, 1007–1018. Jones M.K., Jones R.J., Latreille P.L. & Sloane P.J. (2009). Training, Job Satisfaction, and Workplace Performance in Britain: Evidence from WERS 2004. Labour: Review of Labour Economics and Industrial Relations, 23,139–175. Kaiser, L.C. (2007). Gender-job Satisfaction Differences across Europe: An Indicator for Labour Market Modernization. International Journal of Manpower, 28, 75–94. Kim S.W., Price J.L., Mueller C.W. & Watson T.W. (1996). The Determinants of Career Intent among Physicians at a U.S. Air Force Hospital. Human Relations, 49, 947–976. Miller, P. (1990). Trade Unions and Job Satisfaction. Australian Economic Papers, 29,226–248. Oaxaca, R.L. & Ransom, M.R. (1994). On Discrimination and the Decomposition of Wage Differentials. Journal of Econometrics, 61, 5–21. Sloane, P.J. & Williams, H. (2000). Job Satisfaction, Comparison Earnings and Gender. Labour: Review of Labour Economics and Industrial Relations, 14,473–502. Sousa-Poza, A. & Sousa-Poza, A.A. (2000). Taking Another Look at the Gender/Job Satisfaction Paradox. Kyklos, 53,135–52. Sousa-Poza, A. & Sousa-Poza, A.A. (2003). Gender Differences in Job Satisfaction in Great Britain, 1991–2000: Permanent or Transitory? Applied Economics Letters, 10,691–694. Sousa-Poza A. & Sousa-Poza A.A. (2007). The Effect of Job Satisfaction on Labor Turnover by Gender: An Analysis for Switzerland. Journal of Socio-Economics, 36,895–913.

ACKNOWLEDGEMENT

We are deeply grateful to the various department personnel of the anonymous company that cooperated in this study: Tetsuo Nakashima, Hisakazu Matushige, Tsunehiro Otsuki, Masaru Sasaki and Yasuhiro Yamada. We acknowledge that all remaining errors in this study are ours. This work was supported by a Grant-in-Aid for the Japan Society for the Promotion of Science (JSPS) Fellows.

Journal of Applied Business and Economics vol. 16(3) 2014 89

Inter-Firm Relationships and the Creation of Social Capital

Cameron Gordon University of Canberra, Australia

Shaun Cheah University of Canberra, Australia

INTRODUCTION

Understanding the development of social capital has become an area of significant interest among economists and social scientists (Leana and Van Buren 1999; Adler and Kwon 2002). Cutting across diverse disciplines (Coleman 1988; Putnam 1995; Burt 1997), social capital is a model of the creation of a valuable asset stemming from access to resources through an actor’s social relationships (Granovetter 1985). The closer the pre-existing relationship, the greater the value of social capital and the likelihood that the relationship in question will be maintained (Wathne, Biong et al. 2001). Essentially, the premise behind the notion of social capital is that it is ‘investment in social relations with expected returns’. There has been several definitions established but an overriding consensus is that relationships play a central role in social capital formation, accumulation and generation of returns. Most social capital literature, and the notion of social capital itself, focuses on public externalities and public goods. In other words, economic transactions require a certain amount of trust and relationship ties. These fundamentals cannot be completely captured and priced through a market mechanism and in many ways have to be built prior to a market transaction. Thus, the biggest issues around this process arise in areas which are quasi-market at best, e.g. social cohesion and civic responsibility. Private markets function better where prerequisite social harmony and respect for ‘rules of the game’ are high and here social capital building is seen as essential. This article, however, maintains that social capital is equally critical for, what seems to the outside observer, purely market transactions as well. The focus here will be in the realm of the social network of inter-firm or business-to-business (B2B) relations. It will be argued that adopting a social capital approach is implicitly critical to gaining commercial success over the long term, where businesses must

90 Journal of Applied Business and Economics vol. 16(3) 2014 rely on each other with transactions based on a core sharing of trust, not directly mediated through the market or the price system or, in many cases, even the formal legal system. Commercial success generates long term profits, and long-term profits are essentially based upon resources which are deeply embedded in the social capital built between collaborators (and in some cases competitors) in the B2B context. In essence, firms invest in social capital shared between them so as to be able to focus on activating a relational function (as opposed to a transactional function) which allows them to synergistically exploit existing external resources and skills, and creation of shared productive inputs. A dearth of research exists regarding social capital within the inter-firm or B2B context. (Kelley and Davis 1994; Lee, Sandler et al. 1997; McDonald and Milne 1997; Cousens, Babiak et al. 2001; Farrelly and Quester 2003; Cousens, Babiak et al. 2006). Yet understanding the different types of social capital is important, given social capital’s influence on a firm’s ability to enhance inter-firm value delivery (Griffith and Harvey 2004). To fill that research gap, this paper reviews pertinent literature to arrive at new insights into the joint actions/drivers of inter-firm social capital. In the first part of the article, the conceptual advantages and constructs of social capital will be identified. This will be followed by the identification of inter-firm (B2B) relationship types and a review of the reasons for building these relationships. Thereafter, the link between social capital to inter-firm (B2B) relationships is established. Finally, a framework will be constructed and this will be used to conceptually model the joint actions/drivers of social capital accumulation in an inter-firm or B2B context. This model represents a novel insight into the social capital formation process in the inter-firm relationship and as such provides an understanding of the joint actions/drivers necessary for social capital building. Application of this framework also fills in the gap in managerial knowledge for use in practice, thus improving processes and planning.

Theories of Social Capital: Why Does It Exist? A primary question arising with the social capital construct is: why does it exist at all? With physical and human capital the economic productivity payoffs are inherently obvious. But the need for some kind of embodiment of social relationships in the form of a persistent capital stock is not immediately clear. Thus most social capital begins with this existential question. Social network researchers have taken the lead in formalizing and empirically testing theories related to social capital. They regard relationships, or ties, as the basic data for analysis. A network is defined as the pattern of ties linking a defined set of persons or social actors. Each person is described in terms of their ties with people in the network. The focal person in such an analysis is referred to as ‘ego’, and those they are associated with are ‘alters’ (Knoke and Kuklinski 1982). There are three main theories. The first is Burt's (1997) ‘structural holes approach’. This approach focuses not on the characteristics of ego's direct ties, but on the pattern of relations among the ‘alters’ in ego's social network. A ‘structural hole’ exists between two alters who are not connected to each other. According to this theory, it is advantageous for the ego to be associated to many alters who are themselves unconnected to the other alters in ego's network. Networks rich in ‘structural holes’ provide an individual with three primary benefits: (1) more unique and timely access to information; (2) greater bargaining power and thus control over resources and outcomes, and (3) greater visibility and career opportunities throughout the social system. Initial empirical evidence supports this theory but it has also provided a number of boundary conditions limiting the range of the theory's application (Burt 1997; Sparrowe, Liden et al. 2001). To date, the role of the proposed explanatory processes-access to information, bargaining control, and referral have not been empirically examined. The second theory is ‘weak tie theory’. Here, Granovetter (1985), focuses on the strength of the social tie and argues that ties among members of a social clique are likely to be strong. Thus, the information attained by any one member of the clique is likely to be either shared (or redundant) with the information attained by other members. However, ties that reach outside of one's social clique are likely to be weak. The weak ties are often a bridge between densely interconnected social cliques, eventually providing a source of unique information and resources (Granovetter 1973).

Journal of Applied Business and Economics vol. 16(3) 2014 91 The third theoretical approach is ‘social resources theory’ (Lin, Ensel et al. 1981a), which focuses on the nature of the resources within an embedded network. Here, the authors argued that it is not the weakness of an association that conveys advantage, but rather the fact that such associations are more likely to reach someone with the type of resource required for ‘ego’ to fulfill their objectives. Concurrently, an ‘alter’ who possesses characteristics, or controls resources useful for the attainment of the ego's goals, can be considered a social resource.

Theories of Social Capital: How Does It Generate Returns? As the above discussion reveals, a certain amount of controversy exists regarding the proper conceptualization of social capital causes. ‘Weak tie theory’ focuses on the nature of underlying social ties; ‘structural holes theory’ focuses on the pattern of social ties; whilst the focal point of ‘social resource theory’ is on the characteristics of the specific ties chosen. All three concepts no doubt apply and thus a fruitful integration of the differing conceptualizations of social capital is possible. The key elements of this integration are the structural properties of networks and the nature of the social resources embedded in networks (Lin 1999). ‘Weak tie theory’ and ‘structural hole theory’ each centre on the structure of a network, whilst ‘social resources theory’ focuses on the content of a network. These theories are not mutually exclusive but can function together because the focal point is on different areas in the process of accumulating social capital. Thus, the overarching social capital construct is best thought of as both the different network structures that facilitate (or impede) access to social resources and the nature of the social resources embedded in the network. With these fundamentals in place, one can now closely consider how social capital develops and what its impacts might be. To date, leveraging and understanding the development of social capital has become an area of significant interest among economists and social scientists (Williamson 1979; Burt 1997; Leana and Van Buren 1999; Adler and Kwon 2002; Moran 2005). Researchers have defined social capital in diverse ways (Coleman 1988; Nahapiet and Ghoshal 1998; Knoke 1999; Burt 2000; Adler and Kwon 2002; Hitt, Lee et al. 2002; Moran 2005). The fundamental definition of social capital is that it is investment in social relations with expected returns. This general definition is consistent with various renditions by Bourdieu (1985), Burt (1997), Coleman (1988), Putnam (1995), Lin (1999) and Portes (1988). Lin (1999) identified two perspectives at which return or profit is conceived – one at the individual level and the other at the group level. The individual’s perspective focuses on how individuals invest in social relations and how embedded resources are captured in relations to generate return. The group perspective revolves around how certain groups develop and maintain social capital as a collective asset and how such a collective asset enhances group members’ prospects and returns. Whether social capital is seen from the societal-group level or the relational level, scholars remain committed to the view that it is the interacting members who make the maintenance and reproduction of this social asset possible. An overriding agreement is that relationships play a central role in social capital (Adler and Kwon 2002; Moran 2005). The relevant definitions which illustrate this are outlined by (Hunt 2000; Bolino, Turnley et al. 2002; Kostova and Roth 2003) which characterizes social capital as ‘an asset that is engendered via social relations and that can be employed to facilitate action’. Moreover, a summary of the definitions outlined by Putnam (1995), Durlauf (2002), Adler & Kwon (2002) defines social capital as the ‘relationships between individuals and organizations that facilitate action and yields opportunities to the members of the social network or structure. It is characterized by a sense of trust and mutual interconnectedness, enhanced over time though positive interaction’. Other authors (Coleman 1988; Putnam 1995; Burt 1997) posit that social capital results in benefits, such as re-patronage behaviour resulting in loyalty over time. The closer the pre-existing relationship, the greater the prior investment in social capital and the likelihood that the relationship in question will be maintained (Wathne, Biong et al. 2001). Additionally, the literature posits that social capital facilitates informal contract enforcement (Kostova and Roth 2003). Returning to the basic notion of social capital as investment in social relations

92 Journal of Applied Business and Economics vol. 16(3) 2014 with expected returns, one can see that this general definition is consistent with various renditions by scholars who have contributed to the discussion (Bourdieu 1985; Coleman 1988; Putnam 1995; Burt 1997). These returns are potentially far-ranging and diverse. This also leads Putnam (1995) to reiterate the common idea that social capital is enhanced over time through positive interaction and collaboration in shared interests which can take place in a range of informal and formal meeting places. Is social capital ‘public’ or ‘private’ then? One group of social network researchers upholds the notion of social capital as being a private good that primarily benefits the actors who possess such capital (Granovetter 1985; Burt 1997). Private social capital varies and mainly facilitates the pursuit of individual goals. While other actors might also benefit from such a private good, access is controlled by those who create social capital (Leana and Van Buren 1999). Other researchers view social capital as a collective good and thus underscore its collective benefits. The bulk of the literature lies here, and in this view, trust, reciprocity, and strong social norms facilitate integration (and co-operation), and effectively regulate cooperative social behavior (Fukuyama 1995; Putnam 1995). Collective social capital also has spillover benefits and this relates both to actors that create this capital and their respective network members (Coleman 1988). Here, social capital assists in the pursuit of collective goals by allowing network actors to tap into resources without necessarily having participated in their creation (Kostova and Roth 2003). These views are, of course, not inconsistent. They all focus on the need for cooperation across otherwise unaligned or possibly competing agents to take some sort of joint effort based on mutual trust and connection. This also can be seen in Griffith and Harvey’s (2004) view that ‘the advantage in building social capital in the firm and between firms accrues from both the context and relationship- relevant consequences of interpersonal interaction’. Context’ refers to the manager’s ability to have relevant behaviour or knowledge in interactions, generally referred to as social interaction (Kelly 1984). Thus, managers should possess intimate social insights into how to effectively interact, communicate, and relate to individuals, both internally and externally to develop social capital. ‘Relationship-relevant’ consequences of interpersonal interaction are also required for the development of social capital. Thus, managers have to elicit a high level of associability with individuals via consistent and continued mutual interaction (Kelly 1984). These two dimensions of context-specific orientation facilitates interpersonal trust, which may grow over time due to repeated interaction between individuals in firms (Griffith and Harvey 2004). While business transactions are the frame for this analysis, it can be viewed that these dynamics apply in wider arenas as well. Social capital can also either be a substitute for, or complement to, other resources. As a substitute, actors can compensate for a lack of financial or human capital by superior ‘connections’. More often, though, social capital complements other forms of capital. For example, social capital can improve the efficiency of human and physical capital by reducing transaction costs (Lazerson 1995). Social capital, similar to physical and human capital, also requires maintenance as social bonds have to be periodically renewed or they lose efficacy. Also, like human capital, social capital does not have a predictable rate of depreciation, as it does not depreciate with use. For example, trust demonstrated today is typically reciprocated and augmented tomorrow. While social capital is sometimes rendered obsolete by contextual changes, the rate at which this happens is difficult to predict as even conservative accounting principles cannot estimate a meaningful depreciation rate.

Social Capital and Inter-Firm (B2B) Relationships Theoretically, social capital has been conceptualized at multiple levels; a national level (Fukuyama 1995), an individual level (Belliveau, O'Reilly et al. 1996; Ahuja 2000; Seibert, Kraimer et al. 2001; Kostova and Roth 2003; Perry-Smith and Shalley 2003), the industry level (Baker 1990; Gulati 1995; Walker, Kogut et al. 1997), the group level (Krackhardt 1990; Sparrowe, Liden et al. 2001; Reagans, Zuckerman et al. 2004), the organization/firm level (Leana and Van Buren 1999; Burt 2000; Florin, Lubatkin et al. 2003) and the inter-organizational/firm level (Hunt 2000).

Journal of Applied Business and Economics vol. 16(3) 2014 93 Various reasons have been examined for the establishment of inter-firm or business-to-business B2B relationships. Buttle (2008) points out five reasons for creating and sustaining the relationship. They are product complexity, product strategic significance, service requirements, financial risk and reciprocity. Heeda and Ritter (2005) also delineate that the development of B2B relationships has had different objectives. They are product/competence, product/offering, marketing orientation/solution, customer orientation/problem and networks. Biggemann and Buttle (2004) also outlines that a leading reason for firms to build relationships is the business value that relationships create. In addition, Eisingerich and Bell (2008) point toward three reasons which compel B2B service providers to make relationships with other firms, even competing ones. Firstly, long-term exchanges between firms are central in a services marketing context. Here, business customers may find it difficult to evaluate service quality as services are intangible. As relational exchanges become apparent over time, exchange partners may benefit from reduced uncertainty, exchange efficiency and effective collaboration. Secondly, most service providers face intense competition and incur substantial costs in their development of new services. The relationships with other actors that offer specialised activities can facilitate profitable de-integration of value chains and improve innovation by facilitating greater specialisation of both inputs and outputs. In other words, this ‘flexible specialisation’ may lead to improved efficiency, reduced input prices and greater speed to market. Third, in a B2B service context, networks can be noteworthy as strong business linkages between firms can result in complementarities with respect to resources, which assist in the provision of integrated solutions. Moreover, openness to new and diverse exchange partners facilitates access to new technologies and service know-how. A growing literature deals with the dynamics of inter-firm co-operation and relationship. Here, inter- firm cooperation is defined as the ‘presence of deliberate relations between otherwise autonomous firms for the joint accomplishment of individual operating goals’. The parent literature on inter-firm analysis though, has been criticized for theoretical insufficiency and a lack of empirical research (Litwak and Hylton 1962). Melcher & Adamek (1971) further asserts a lack of contribution and lamented that there was "little attempt to relate the approaches of the different contributors as scholars (and researchers alike) failed to expand or build upon each other's concepts”. In particular several questions persist. Firstly, the variables which lead to successful B2B relationships (i.e. buyer-seller, manufacturer-dealer, distributor-supplier) are still unclear. Secondly, there are limited studies examining these variables and its correlation with economic results. Thirdly, relationships evolve, and thus it would be prudent to understand the variables that lead to relationship success during the beginning or ending stages of the relationship. Answers to these questions would facilitate the management of B2B relationships in a manner which would be mutually beneficial to all B2B participants. The social capital construct appears to be particularly useful in this regard. Although many studies have been carried out in order to understand social capital from various perspectives limited research has focused on understanding how social capital influences relationships from the inter-firm or business-to- business (B2B) level. Griffith & Harvey (2004) asserts that business partner social capital (or B2B social capital) is the area which warrants further empirical investigation. Business partner social capital is defined as ‘an asset that an organization has developed within its infrastructure of global business relations (i.e., buyers and suppliers) that can be mobilized to facilitate action and enhance value delivery’ (Peng and Luo 2000). In essence, B2B social capital facilitates the functioning of the relationship and the enhancement of a firm’s ability to deliver customer value in the global marketplace. This sort of ‘business partner capital’, where the ‘partnership’ is between firms, is inherently based on the individual social capital developed via marketing managers. But this is, in a sense, the building block. Peng and Luo (2000) posit business networks as a being a set of interconnected organizations or alternatively a set of connected relationships. This perspective view firms as dependent upon their network partners for inputs necessary for effective operation. As firms can be seen as purposive social actors, it is inevitable that researchers have extended the logic of social capital to the inter-firm level (Burt 1997; Tsai and Ghoshal 1998)

94 Journal of Applied Business and Economics vol. 16(3) 2014 Inter-firm or B2B relationships, as conduits and control of key information, create entrepreneurial opportunities (Burt 1997). The interactions between firms also establish a pattern of expectations based on norms of reciprocity and equity. If these two patterns persist, then the sum of resources that accrue to a firm (by virtue of possessing a durable network of inter-firm relationships) transpires and a social capital base is built. Thus, social capital is greater than inter-firm relationships in isolation (Bourdieu 1985). Social capital thus provides a way to characterize a firm’s complete set of relationships but with a focus on ongoing access to a flow of resources (i.e. knowledge, information, and other capital) to the firm through its alliances. Here, understanding the nature of social capital is necessary because it is a key element of a firm’s competitive advantage. What does this advantage consist of? Social capital, based on inter-firm (B2B) relationships, provides access to capabilities and resources that may otherwise be unavailable (Koka and Prescottt 2002). The premise of this argument lies in the view that firms are heterogeneous entities enriched with capabilities and resources (Wernerfelt 1984). These firms normally resort to alliances to access the means necessary for competitive advantage as interactions between firms establish a pattern of expectations that are based on norms of reciprocity and equity. Furthermore, Ford et al (1998) highlight that management of relationships among all business stakeholders has become critical to a firm’s very existence. As such, the literature has seen a shift of emphasis from firms engaging in discrete transactions toward longer-term mutually beneficial exchange relationships (Claycomb and Martin 2001). In the course of their business activities, firms establish a variety of business-to-business (B2B) ties which are well documented (Burt 1997; Tsai and Ghoshal 1998). Such ties include buyer-supplier relationships, strategic alliances, and joint memberships in industry associations. These ties enable firms to exchange a variety of information, knowledge, and other forms of capital. This results in increased productivity and as such, firms have been focusing on long-term engagements with their business partners. A critical dimension is the buyer and seller relationship between firms (Wilson 1995). The fact that a relationship exist between buyers and sellers is nothing new. Over time, relationships are naturally developed as buyers and sellers develop trust and friendships. These relationships become “strategic” when the process of relationship development is accelerated and concretised to achieve mutually beneficial goals (Wilson 1995). Godson (2009) asserts that any business relationship ultimately comes down to individuals and reiterates that a successful outcome for the buyer-supplier relationship relies on the strength and nature of this relationship. The literature also posits that some desired outcomes in inter- firm networks are individual in nature (i.e. the manager of one firm networking with manager of another firm) as alliance partners are connected by interplay of cooperation and competition (Oliver 1990; Grandori and Giuseppe 1995). Inter-firm relationships, however, have been explored with a focus on different relationship characteristics that go beyond links between individuals (Oliver 1990; Ring and Van de Ven 1992; Grandori and Giuseppe 1995). For researchers, the formal mechanisms of contractual and procedural coordination have become a key focus for the governance of such relationships. Contractual coordination refers to “the mutual exchange of rights between parties involved in a relationship in order to govern the combination of agents or functions towards the production of results” (Portes 1988). These rights define the establishment of operating procedures to govern the exchange and thus, the distribution of the rights is a key determinant of how coordination occurs. When entering a relationship, each partner gives up some of their rights whilst gaining others, through either explicit or implicit contracts. Other causes of successful B2B relationships that directly link to social capital formation and accumulation have been comprehensively investigated by Anderson and Hakansson (1994), Ganesan (1994), Morgan & Hunt (1994), Mohr et (1996), Wright (2004), Cheng (2006) and Power & Reagan (2007). Power and Reagan (2007) point out the importance of B2B relationship success for a firms’ reputation, performance satisfaction, possibilities of alternatives, mutual goals, technology, non- retrievable investments, adaptation and structural bonds. Lehtonen (2004) also asserts that successful B2B relationships require the establishment and execution of clearly defined goals. In addition, Cheng (2006) reaffirms that longevity, frequency of

Journal of Applied Business and Economics vol. 16(3) 2014 95 contact, efficient service and skills, and personalities of firm representatives as success of relationships. These various relationships may be characterized by motives that could involve mutually compatible and incompatible goals. The key element though, of successful relationship and social capital building, is the shared recognition and joint commitment to work towards a mutually beneficial relationship. One other critical element to B2B social capital is governance structure. Governance structure to minimize the sum of production costs for a given transaction is the core issue investigated by transaction costs economics (Williamson 1979). Although governance structure is used to organise and guide economic behaviour, the theoretical definitions (Williamson 1979) and empirical operationalisations (Sparrowe, Liden et al. 2001; Bowler and Brass 2006) focus on the informal and formal contractual dimensions. In establishing governance structures, partners have to ‘choose between either prescribing and enforcing specific actions or using means to create a general commitment between the partners from which desirable actions evolve’ (Williamson 1979). This process is associated to procedural coordination and is in accordance to the mutual exchange of information or functions towards the production of results. These exchange opportunities may be structurally identified by the form of contractual mechanisms chosen and may also be governed by ‘internal’ or ‘psychological’ contracts. Contractual coordination mechanisms also allow institutions to achieve alignment of partner incentives. However, it is unfeasible to ascertain how they are employed to coordinate the activities of the partners in the relationship. Doz, Hamel and Prahalad (2002) argue that the actual coordination is not achieved through contractual mechanisms but is realised by the day-to-day communication of the participants involved in the relationship activities. Top management establishes strategic alliances whilst setting legal parameters for exchange. ‘What actually gets traded is determined by day-to-day interactions of engineers, marketers, and product developers' (Settoon and Mossholder 2002). These 'day-to-day interactions' are at the core of the construct 'procedural coordination'. Procedural coordination does not refer to institutions that may be in place to govern the relationship but asks how these institutions are used. In this instance, social capital can be seen as a very useful construct for building understanding.

Towards a Model of B2B Social Capital Building and Returns There are many theoretical perspectives on inter-firm relationship and value creation such as resource dependence theory (Pfeffer and Salancik 1978), marketing channel theory (Frazier 1983); transaction cost economics (Williamson 1979), transactional value analysis (Zajac and Olsen 1993; Dyer and Singh 1998), and resource-based theory (Wernerfelt 1984). Social capital theory in the B2B context is quite limited (Granovetter 1985) and thus will be treated as a variable evoked by certain antecedents, and thus the focus is on the joint drivers/actions of social capital. This is critical to the understanding of the inter- firm relationships as the social context in which the joint drivers/actions are embedded have largely been ignored relative to social capital. This also responds to Palmatier and colleagues’ (Palmatier, Dant et al. 2006) call for more research synthesizing different theoretical perspectives for understanding inter-firm joint actions/drivers and exchanges. One could conceptualise B2B social capital in the following way. Assume for simplicity there are two firms, 1 and 2. Assuming atomistic and unrelated operations at the start, one could posit the following firm production functions: Firm 1 Output = f (K1, L1) Firm 2 Output = f (K2, L2)

Where K refers to physical capital owned and used be each firm, L refers to labour owned and used by each firm and the superscripts indicating the specific stocks that each respective firm has under its control. These two firms would be unrelated to each other but perhaps they may individually determine that they could be better off if they engaged in some sort of relationship. The relationship between the firms in the market could be quite varied, ranging from suppliers at different points in the supply chain to direct competitors of the same product in the same market. One way to conceive of this possibility (and there are, of course, many such conceptions), is to imagine a jointly produced and used third input, ‘S’

96 Journal of Applied Business and Economics vol. 16(3) 2014 which refers to Social Capital and which is available to both firms. We would then have the following situation:

Firm 1 Output = f (K1, L1, S1+2) Firm 2 Output = f (K2, L2, S1+2)

Now both firms now have an extra input with which to produce their output. The natural extension of this model is a production function for production of S1+2. Generically this looks like the following: S1+2 = f (inputs….)

This generic model tells us nothing, of course, but does suggest a framework for arranging the various strands of literature discussed above and generating testable hypotheses. The two key questions which arise from this are: (1) What does ‘f’ consist of? A production function model posits a ‘technology’ that transforms inputs into outputs and we might usefully as what the nature of that transformational process is in a B2B context. (2) What are the relevant ‘inputs’ that go into making the output S? Trust, shared experience, and joint knowledge are three of the numerous possibilities suggested by the literature. These can be made more specific depending upon the specific B2B situation and the nature of those jointly producing S. For example, are the input different for, or, if the same, look differently for direct competitors versus interdependent suppliers. If so, how?

We could ask a third question as well which pertains to the ‘publicness’ of S. As modelled above it is clear that the two firms share the joint input S1+2. But is it possible that there are times when others not directly investing and participating may nonetheless be able to use it in their production functions? In other words might we have a third firm, outside the network, which has a production function looking like this? Firm 3 Output = f (K2, L2, S1+2)

One can see that this simple framework allows for a variety of other possible ‘spillover’ scenarios or, on the other side, barriers to sharing of social capital. For example there might be this possibility within the network: Firm 1 Output = f (K2, L2, S1+2) Firm 2 Output = f (K2, L2) S1+2 = f (inputs from firm 1, inputs from firm 2)

In this case we are modelling an outcome where both firms have determined it to be worthwhile to generate S1+2 but only one firm is actually benefitting from it (in this case Firm 1).

Conclusions and Suggestions for Further Research This paper has reviewed the literature on social capital and the B2B relationship literature. It has argued that social capital can be a very useful concept for modelling B2B relationships but that relatively little conceptual crossover has yet occurred. A generic modelling structure has been suggested for accomplishing such a crossover. The agenda for future researchers would be for the conceptual model to be empirically tested in specific business or marketing context (e.g. sports sponsorship, retail or finance). The findings from this would then provide a general framework which can be applied to general businesses. It will also facilitate the management of B2B relationships in a manner which would be mutually beneficial to all B2B participants. Moreover, the significance of the correlation of joint social capital to outcomes also sheds new light on how firms can leverage social capital to achieve desired outcomes. Overall, this will represent a major

Journal of Applied Business and Economics vol. 16(3) 2014 97 advance in expanding knowledge of social capital theory and utility for both researchers and business managers alike.

REFERENCES

Adler, P. S. and S. W. Kwon (2002). Social capital: Prospects for new concept. Academy of Management Review 27(1): 17-40. Ahuja, G. (2000). The duality of collaboration: inducements and opportunities in the formation of inter- firm linkages. Strategy Management Journal 21: 317-343. Anderson, J. C., H. Hakansson, et al. (1994). Dyadic Business Relationships within a Business Network Context. Journal of Marketing (October): 1-15. Baker, W. E. (1990). Market networks and corporate behaviour. American Journal of Sociology 96(3): 589-625. Belliveau, M. A., C. A. O'Reilly, et al. (1996). Social capital at the top: effects of social similarity and status on CEO compensation. Academy of Management Journal 39(6): 1568–1593. Biggemann, S. and F. Buttle (2004). Conceptualising Business to Business Relationship Value. Rotterdam, Erasmus University. Bolino, M., W. Turnley, et al. (2002). Citizenship behaviour and the creation of social capital in organisations. Academy of Management Review 27(4): 505-522. Bourdieu, P. (1985). The forms of capital. New York: Greenwood. Bowler, W. M. and D. Brass (2006). Relational correlates of interpersonal citizenship behavior: A social network perspective. Journal of Applied Psychology 91: 70-82. Burt, R. S. (1997). The contingent value of social capital. Administrative Science Quarterly 42(2): 339- 365. Burt, R. S. (2000). The network structure of social capital. Research in organisational behaviour 22: 345- 423. Buttle, F. (2008). Customer Relationship Management. Burlingthon, MA, Butterworth-Heinemann. Cheng, C. S. (2006). Financial Services Marketing: Empirical evidence on the impact and effectiveness of marketing events: dissertation. Claycomb, C. and C. L. Martin (2001). Building customer relationships: an inventory of service providers' objectives and practices. Marketing Intelligence & Planning 19(6): 385-399. Coleman, J. (1988). Social capital in the creation of human capital. American Journal of Sociology 94: 95-120. Cousens, L., K. Babiak, et al. (2006). Beyond sponsorship: re-framing corporate-sport relationships. Sport Management Review: 1-23. Cousens, L., K. Babiak, et al. (2001). Adopting a relationship marketing paradigm: the case of the national basketball association. Sports Marketing and Sponsorship (December/January): 331-55. Durlauf, S. (2002). On the empirics of social capital. Economic Journal 112(483): 459−479. Dyer, J. H. and H. Singh (1998). The relational view: cooperative strategy and sources of inter- organisational competitive advantage. Academy of Management Review 23: 660-679. Eisingerich, A. B. and S. J. Bell (2008). Managing networks of inter-organisational linkages and sustainable firm performance in business-to-business service contexts. Journal of Services Marketing 22(7): 494-504. Farrelly, F. and G. P. Quester (2003). What Drives Renewal of Sponsorship Principal/ Agent Relationships? Journal of Advertising Research: 353-359. Florin, J., M. Lubatkin, et al. (2003). A social capital model of high-growth ventures. Academy of Management Journal 46(3): 374-384. Ford, D., L. Gadde, et al. (1998). Managing business relationships. Chichester. England, John Wiley & Sons. Frazier, G. L. (1983). Inter-organizational exchange behaviour in marketing channels: a broadened perspective. Journal of Marketing 47(4): 52-67.

98 Journal of Applied Business and Economics vol. 16(3) 2014 Fukuyama, F. (1995). Trust: The social virtues and the creation of prosperity. New York, Free Press. Ganesan, S. (1994). Determination of Long-Term Orientation in Buyer Seller Relationships. Journal of Marketing 58(April): 1-19. Godson, M. (2009). Relationship Marketing. New York Oxford University Press. Grandori, A. and S. Giuseppe (1995). Inter-firm networks: Antecedents, mechanisms and forms. Organization Studies 16/2: 183-214. Granovetter, M. (1973). The strength of weak ties. American Journal of Sociology 78: 1360-1379. Granovetter, M. (1985). Economic Action and Social Structure: The Problem of Embeddedness American Journal of Sociology 91(Nov): 481. Griffith, D. and M. Harvey (2004). The influence of individual and firm level social capital of marketing managers in a firm's global network. Journal of World Business 39: 244-254. Gulati, R. (1995). Social structure and alliance formation patterns: a longitudinal study. Administration and Society Quarterly 40(4): 619-652. Heeda, L. and T. Ritter (2005). Business relationships on different waves: Paradigm shift and marketing orientation revisited. Industrial Marketing Management 34: 714-721. Hitt, M. A., H. Lee, et al. (2002). The importance of social capital to the management of multinational enterprises: Relational networks among Asian and Western firms. Asia Pacific Journal of Management 19(2-3): 353−372. Hunt, S. D. (2000). A general theory of competition. Thousands Oaks, CA, Sage. Kelley, S. and M. Davis (1994). Antecedents to Customer Expectations for Service Recovery. Journal of the Academy of Marketing Science 22(Winter): 52-61. Kelly, H. (1984). The transitional descriptions of interdependence by means of transition lists. Journal of Personality and Social Psychology 47: 956–982. Knoke, D. (1999). Organizational networks and corporate social capital (Eds.). Boston, Kluwer. Knoke, D. and J. H. Kuklinski (1982). Network analysis. Beverly Hills, CA, Sage. Koka, B. R. and J. Prescottt (2002). Strategic alliances as social capital: a multi-dimensional view. Strategic Management Journal 23: 795-816. Kostova, T. and K. Roth (2003). Social capital in multi-national corporations and a micro-macro model of its formulation. Academy of Management Review 28(2): 297-313. Krackhardt, D. (1990). Assessing the political landscape: structure, cognition and power in organizations. Administrative Science 35: 342–369. Lazerson, M. (1995). A new phoenix: Modern putting-out in the Modena knitwear industry." Administrative Science Quarterly 40: 34-59. Leana, C. R. and H. J. Van Buren (1999). Organisational social capital and employment practices. Academy of Management Review 24(3): 538-555. Lee, M., D. Sandler, et al. (1997). Attitudinal constructs towards sponsorship. International Marketing Review Vol. 14(3): 159-169. Lehtonen, T. (2004). Attributes and success factors of partnering relations -a theoretical framework for facility services. Nordic Journal of Surveying and Real Estate Research Special Series (2): 31- 46. Lin, N. (1999). Building a network theory of social capital. Connections 22(1): 28-51. Lin, N., W. M. Ensel, et al. (1981a). Social resources and occupational status attainment. Social Forces 59: 1163-1181. Litwak, E. and L. Hylton (1962). Inte-rorganizational Analysis: A Hypothesis on Co-ordinating Agencies. Administrative Science Quarterly 6: 395-420. McDonald, M. A. and G. R. Milne (1997). A conceptual framework for evaluating marketing relationships in professional sport franchises. Sport Marketing Quarterly 6(2): 27-32. Melcher, A. and R. Adamek (1971). Inter-organizational Models: A Critical Evaluation. Kent, Ohio, The Comparative Administration Research Institute. Mohr, J., R. Fisher, et al. (1996). Collaborative communication in inter-firm relationships: Moderating effects of integration and control. Journal of Marketing 60(3): 103- 117.

Journal of Applied Business and Economics vol. 16(3) 2014 99 Moran, P. (2005). Structural vs. relational embeddedness: Social capital and managerial performance. Strategic Management Journal: 1129−1151. Morgan, R. M. and S. Hunt (1994). The Commitment- Trust Theory of Relationship Marketing. Journal of Marketing 58 (July): 20-38. Nahapiet, J. and S. Ghoshal (1998). Social capital, intellectual capital, and the organisational advantage. Academy of Management Review 23: pp. 242-266. Oliver, C. (1990). Determinants of inter-organizational relationships: Integration and future directions. Academy of Management 15/2: 241-265. Palmatier, R., R. Dant, et al. (2006). Factors Influencing the Effectiveness of Relationship Marketing: A Meta-Analysis. Journal of Marketing 70: 136-153. Peng, M. W. and Y. Luo (2000). Managerial ties and organisation performance in a transition economy: The nature of micro-macro link. Academy of Management Journal 43(3): 486-501. Perry-Smith, J. E. and C. E. Shalley (2003). The social side of creativity: a static and dynamic social network perspective. Academy of Management Review 28(1): 89-106. Pfeffer, J. and G. Salancik (1978). The External Control of Organizations: A Resource Dependence Perspective New York, NY., Harper & Row. Portes, A. (1988). Social capital: Origins and applications. Annual review of sociology. I. Hagan and E. Cook. Palo Alto. CA, Annual Reviews. vol. 24: 1-24. Power, T. L. and W. R. Reagan (2007). Factors in Influencing successful buyer-seller relationship. Journal of business Research 60: 1234-1242. Putnam, R. (1995). Bowling Alone: America's Declining Social Capital. Journal of Democracy 6(1): 65- 78. Reagans, R., E. Zuckerman, et al. (2004). How to make the team: social networks vs. demography as criteria for designing effective teams. Administration and Society Quarterly 49: 101–33. Ring, P. and A. H. Van de Ven (1992). Structuring Cooperative Relationships Between Organizations. Strategic Management Journal 13: 883-498. Seibert, S. E., M. L. Kraimer, et al. (2001). A social capital theory of career success. Academy Management Journal 44(2): 219–237. Settoon, R. P. and K. W. Mossholder (2002). Relationship quality and relationship context as antecedents of person- and task-focused interpersonal citizenship behavior. Journal of Applied Psychology(87): pp. 255-267. Sparrowe, R. T., R. C. Liden, et al. (2001). Social networks and the performance of individuals and groups. Academy of Management Journal (44): 316-325. Tsai, W. and S. Ghoshal (1998). Social capital and value creation: The role of intra-firm networks. Academy of Management Journal 41: pp. 464−479 Walker, G., B. Kogut, et al. (1997). Social capital, structural holes, and the formation of an industry network. Organisational Science 8(2): 109-125. Wathne, K., H. Biong, et al. (2001). Choice of supplier in embedded markets: relationship and marketing program effects. Journal of Marketing 65 (April): pp 54-56. Wernerfelt, B. (1984). A Resource-Based View of the Firm Strategic Management. Strategic Management Journal of Academy of Marketing Sciences 5(2): pp. 171-180. Williamson, O. E. (1979). Transaction cost economics: the governance of contractual relations.Journal of Law and Economics 22(October): 223-261. Wilson, D. T. (1995). An integrated model of buyer-seller relationships. University Park, Pennsylvania State University. Wright, R. (2004). Business to business marketing: a step by step guide. Glasgow. Zajac, E. J. and C. P. Olsen (1993). From Transaction Cost to Transaction Value Analysis: Implications for the Study of Interorganizational Strategies. Journal of Management Studies 30(January): pp. 131-145.

100 Journal of Applied Business and Economics vol. 16(3) 2014

Tax Morale in Socio-Political Interactions: Insiders and Outsiders

Savaş Çevik Selcuk University

INTRODUCTION

Since the pioneering work of Allingham and Sandmo that was based on deterrence factors (1972), the literature of tax compliance has expanded with discussions on the importance and explanatory powers of non-economic factors such as social norms, moral appeals, social cohesion, political institutions and attitudes toward tax system/tax administration. Kirchler (2007), Torgler (2007) and Torgler 2008) present an extensive discussion of new approaches. Compliance level differences among countries can be explained by differences in these social and institutional structures (Torgler, 2003a; Cullis et al., 2012). Numerous studies demonstrate that tax morale that indicates intrinsic motivation to pay taxes has an important impact on individuals’ participation in collective cooperation in the form of paying tax (Feld and Frey, 2005; Torgler et al., 2010; Cummings et al., 2009; Alm and Torgler, 2011). From the perspective of ‘psychological tax contract’ of Feld and Frey (2010), taxation is a quasi- voluntary exchange, and the complex interaction between taxpayers and government/tax authorities can be seen as an implicit contractual relationship that involves emotional ties and loyalties as well as duties and rights of the parties involved. Neglecting the obligations of either party would undermine the psychological sanctions of the contract for the other party. This contract includes not only fiscal exchange and reciprocity in relationship in related to public services-tax prices but also elements of positive treatment, respect, and participation in political decision-making at the procedural level. Therefore, taxpayer’s tax morale is determined by government policies, public services, tax authorities’ treatment of taxpayers and political context (Frey, 2003; Feld and Frey, 2007; Feld and Frey, 2010). According to Feld and Frey (2007 and 2010), intrinsic motivation that is shaped by the interaction between the government and the taxpayer responds negatively to external pecuniary motives such as punishments and rewards. Deterrence factors crowd out intrinsic norm-guided behavior when the

Journal of Applied Business and Economics vol. 16(3) 2014 101 taxpayer perceives them to be oppressive. On the other hand, Akerlof and Kranton (2008) argue that firms’ management policies shape workers’ identities. Similarly, one can expect that public administration will shape taxpaying motivations through tax administration as well as public services. In fact, with recent developments in behavioral and psychological economics contradicting the basic understanding of ‘homo economicus’, economists recognize that non-pecuniary motives such as personal tastes, identities and social norms should be included in individual utility functions for exact understanding of individual motivations. Social norms and identities are accepted as “powerful sources of motivation. Norms affect fine-grain decision[s] of the moment . . . [and] drive life-changing decisions as well: …quit[ting] school, whether and who to marry, and whether to work, save, invest, retire, and fight wars” (Akerlof and Kranton, 2010; p. 6). In their examination of identity economics, Akerlof and Kranton (2010) emphasize that some workers who identify with their organization intrinsically contributes high effort and gain identity utility, while some others who do not identify with the firm gain identity utility by contributing low effort. Akerlof and Kranton (2010) call the high effort contributors as ‘insider’ and low effort contributors as an ‘outsider’ with reference to the efficient wages models. The issue of tax morale can be approached with this insider/outsider metaphor if the ‘identity economics’ of Akerlof and Kranton (2010) and the ‘psychological tax contract’ of Feld and Frey (2007 and 2010) are incorporated in analyzing tax compliance of socio-political context. Taxpayers who have high motivation to pay taxes can be called ‘insiders’, and taxpayers who are not intrinsically motivated to pay taxes can be called ‘outsiders’. Why do some individuals have high tax morale while others have low tax morale? Which channels and factors make individuals an ‘insider’ or an ‘outsider’ in their taxpaying motivations in the context of socio-political interactions? Intrinsic motivation is shaped by dynamic interactions with the government and ‘others’ in the society as well as by personal values. The first channel is related to government and society as a whole unity, while the second channel relates to personal identities and norms acquired in a social context. Therefore, in addition to a ‘psychological contract’ formed by an interaction between the taxpayer and the government, taxation can also be seen as a ‘social act’ with others in context of social interaction (Frey and Torgler 2007). A taxpayer’s motivation is also shaped by social context, and an individual taxpayer’s behavior is influenced by ‘others’ who live in the same community. Of course, even in this framework, enforcement actions (audits and penalties) are needed to deter reluctant taxpayers and to ensure that willing taxpayers are not exploited by free riders. The rest of this paper is structured as follows: Section 2 attempts to incorporate political interaction, social context, and personal orientations to explain intrinsic taxpaying motivation. It reviews and synthesizes tax compliance and social interaction literature. Section 3 empirically analyzes using logistic regression with data obtained from the World Values Survey (WVS) to test hypothetical expectations in Section 2. This study places greater emphasis on personal value orientations and personal norms than other papers that deal with social context on tax morale and that use WVS data. Section 4 concludes.

TAX COMPLIANCE IN SOCIETAL AND POLITICAL INTERACTIONS: A FRAMEWORK

When tax compliance is considered a collective action matter, the reasons people choose to cooperate or not to cooperate with authority and society is an important issue. Standard economic theory of the expected utility approach and the Allingham-Sandmo model as it reflects on the tax compliance puzzleissue consider behaviors to be motivated by pecuniary self-interest because of rational-actor assumption. However, experimental evidences especially and unambiguously demonstrate that people cooperate and contribute more than theoretically predicted by standard economic theory (Gächter 2007). Attempting to explain this voluntary cooperation reveals three possible explanations. First, scholarship generally considers people to have pro-social, purely altruistic preferences and an aversion to inequality in distribution. According to Fehr and Scmidt (1999), “people have altruistic feelings toward others and want to increase their material payoffs when those payoffs fall below some benchmark, and they feel envy toward others and want to decrease their payoffs when those payoffs exceed some benchmark.” Frey (1997) and Orvinsa and Hudson (2002) argue that in taxpaying behavior, people are

102 Journal of Applied Business and Economics vol. 16(3) 2014 motivated not only by maximizing their own well-being but also by a sense of the civic duty toward society and its political organization. The second explanation can be to revise the standard utility function. In other words, people have utility functions that contain not only egoistic preferences but also altruistic preferences which create ‘warm-glow’ feelings (Andreoni, 1990). Tax compliance literature holds some important studies that attempt to revise the Allingham-Sandmo model by adding psychological costs caused by the disutility from behaving dishonestly into taxpayers’ utility function. According to these models (Spicer, 1986; Gordon, 1989; and Erard and Feinstein, 1994), tax evasion creates non-pecuniary costs of guilt, shame, and a reduction in self-image in addition to pecuniary costs of legal sanctions. Finally, it can be assumed that voluntary cooperation behavior is not independent of others’ behaviors. Individuals’ decisions and behaviors are systematically affected by the behavior of other members of the group. Thus, while individuals will make decisions about whether or not they will cooperate, they seek cooperative from others. If they believe others cooperate, they cooperate as well (Gächter, 2007:21). There is numerous evidence supporting this reciprocal tendency and conditional cooperation. Fischbacher et al. (2001) and Fischbacher and Gächter (2010) found important evidence for conditional cooperation from results of experimental studies. Subjects are sensitive to the level of others’ contributions, and their beliefs about others’ contributions shape contribution decisions. Apparently, individuals have a mostly intrinsic reciprocity desire and are willing to sacrifice in favor of others in response to kind behavior. Reciprocal-minded subjects are willing to reward or punish others at the expense of high costs (Fehr and Schmidt, 2003). Reciprocal tendency and social interaction have a strong influence on individual decision-making in a social context. Reciprocity as a willingness to cooperate with others who have previously demonstrated cooperation is an important behavioral motivation and human action characteristic (Rabin, 1993). Behavior is influenced by perceptions of societal attitudes and behaviors of individuals or institutions. Taxpaying behavior is determined by two dimensions of interaction. One is between the taxpayer and the government, and the other is between the taxpayer and others in the society. Therefore, taxpayers may seek two forms of reciprocity to determine whether or not they will comply. Vertical reciprocity occurs in the relationship between the individual and the public sector (government, political institutions, and tax administration). Reciprocal behavior may also horizontally play a role in the interaction among taxpayers with respect to perceptions on other taxpayers’ compliance levels. (Schnellenbach, 2010:56). Based on these arguments, this study will deal with the vertical and horizontal directions of the socio- political interaction to determine how a taxpayer becomes an ‘insider’ or an ‘outsider’. A taxpayer’s intrinsic motivation to pay taxes can be influenced on one hand, by the interaction with government and society as a general entity, and on the other hand, by the interaction with other individual/categorical members of society. Of course, personal norms, moral appeals, values, and identities that individuals have acquired as a result of these complex interactions are factors influencing tax morale and tax compliance as a moderator. Impacts of socio-political interactions on tax compliance will discuss by following this distinction.

Interaction with Political System and Social Capital Political and societal institutional structure is important in determining whether or not people cooperate with authority and society as a whole entity. First of all, the relationship between the taxpayer and the government contains a partial exchange. Taxpayers expect that the government should produce some important public services for society. Taxpayers may use tax evasion to restore equity when they perceive procedurally unfair treatment and when they do not receive enough government benefits (Schnellenbach, 2010). Experimental evidence by Alm et al. (1992) and Alm et al. (1993) demonstrates that tax compliance is greater when subjects perceive benefits from expenditure programs that they approved. According to findings of Alm et al. (1992), an increase in benefits increases tax compliance even when there is no chance of detection and punishment. If taxpayers do not perceive fiscal exchange as fair, it follows that they do not possess a willingness to pay taxes.

Journal of Applied Business and Economics vol. 16(3) 2014 103 Undoubtedly, an individual’s reciprocal tendency appears not only in fiscal exchange but also in procedural and interactional justice which also shapes taxpaying decisions. Studies of social dilemma and collective action indicate that cooperation with authorities depends on the extent to which people find the exercise of power as fair, legitimate and respectful. People are more willing to cooperate when they feel authorities treat them with impartiality, dignity and respect (Tyler and Lind, 1992). Feld and Frey (2010) emphasize that political decision-making systems should create fair procedures (procedural justice), fair outcomes, (distributive justice) and respectful treatment (interactional justice) in order to persuade citizens to become insides. Smith (1992) and Wenzel (2006) demonstrate that taxpayers have high compliance levels when they feel respectful and fair treatment from authorities. Empirical evidence indicates a number of important determinants of voluntary tax compliance, such as the quality of public governance (Cummings et al., 2009), the quality of government services (Uslaner, 2010),citizens’ political participation rights and confidence in political decision making (Alm et al., 1993; Kucher and Götte, 1998; Feld and Frey, 2002; Feld and Tyran, 2002; Torgler, 2003b; Torgler, 2005; Torgler and Schaltegger, 2005),a fair, well-functioning tax system and respectful treatment by tax authorities (Pommerehne and Weck-Hannemann, 1996; Seidl and Traub, 2001; Torgler, 2004), and trust in political institutions (Kucher and Götte, 1998; Scholz and Lubell, 1998; Torgler, 2003b; Slemrod, 2003). In their study, Torgler et al. (2010) found that the quality of political institutions and trust in these institutions have strong and significant effects on tax morale. Corruption and inefficiency of government institutions crowd out willingness to pay tax. In this kind of a situation, obligation to pay taxes cannot become a social norm for large segments of the society (Torgler et al., 2010). Trust can be seen as a proxy of reciprocal and fairness terms and legitimacy of political institutions in collective action. As long as citizens feel that they can affect political decision-making and that the process works in fair manner, they tend to obey decisions produced via this process. Otherwise, it is more likely that taxpayers will demonstrate negative reciprocal responses. Persons who regard the government and the tax system as legitimate will have higher incentives to comply with tax laws. Otherwise, they incur some psychological costs like guilty feelings (Wenzel 2007). Trust and participation are important factors not only in tax compliance for political decision-making, but also for interaction with society as a whole. Because taxpaying has distributional results in a collective action, the decision whether or not to comply depends on an individual’s attitude toward society and other citizens (Scholz and Lubell, 1998). Literature on social capital emphasizes the importance of trust and participation in collective actions, for both political institutions and societal institutions, to drive cooperation (Putnam et al., 1993; Uslaner, 2002; Hardin, 2006). A lack of trust in society and a feeling of alienation toward society decrease the willingness to contribute to finance collective actions. Therefore, an examination of the compliance decision should consider trust in societal institutions and participation in other collective actions. These factors reflect the need to feel like a part of society and thus, to be an ‘insider’. It can be expected that if taxpayers trust institutions created by societal interactions and are involved in collective actions with the other members of society or are involved in activities created by other members of the society, they feel high cohesion in the society, gain societal identity, and thus, comply with societal norms and decisions. According to social identity theory, people use their membership in social organizations to construct one aspect of their sense of self, and when people define themselves through group membership, they are willing to invest in the group and want it to be successful (Tyler, 2000). At this point, how people identify and classify themselves is important for their social cohesion perception. If people possess a self-identify as autonomous or as a sub- identity, instead of a national identity, they may fail to participate in a national collective action such as taxation. Therefore, trust in others in the society, participation levels of general collective action, and self- identification with society as a whole should be taken into consideration in order to examine the impact of social capital on tax compliance.

104 Journal of Applied Business and Economics vol. 16(3) 2014 Societal Interaction: Social Influence and Social Norms Individuals encounter other members of society, observe them, and learn from them. Social interaction connects individuals’ decisions to each other through social influence and social norms, and thus, a person’s behavior is determined to a certain extent by the behaviors of others in that society. Mostly, in social interaction, individual perceptions of others’ behaviors take effect through social norms. Social norms as rules and expectations that guide and/or constrain social behavior are sustained in part by social sanction or reward (Elster, 1989; Alm et al., 1995). Social interaction regulates the behavior of compliance and cooperation, especially through three social norms: the reciprocity in responding to a positive action with another positive action in a cooperative manner; the equity in distributing resources among group members related to contributions, input, or costs; and the commitment to consistent actions in words and beliefs, even when they are not demonstrated publicly (Kerr, 1995; Ostrom, 1998). The notion of social influence also suggests that individuals tend to conform with their behavior to their peers’ moral values and behaviors. This is observed in daily choices such as fashion, consumption, and political support for a winner as well as engaging in criminal activities such as organized crime, lynching, and looting (Kahan, 1997). With respect to tax compliance, social norms and social influence mean that an individual’s perception about others’ tax compliance behaviors (especially of peers or those in similar situations) is an important determinant in compliance decisions. Falk et al. (2003) demonstrated direct evidence that peer effect or social interaction is important, and that subjects’ contributions are systematically influenced by social interaction in a public good experiment. Thus, taxpayer’s beliefs about the compliance behavior of others shape compliance decisions. Taxpayers are less likely to cheat on their taxes if others behave honestly (Frey and Torgler, 2007; Traxler, 2010). Fortin et al. (2007) show that subjects’ tax evasion levels are influenced by the behavior of others in an experimental design. Frey and Torgler (2007) find a positive correlation between tax morale and compliance level perceptions of others in the society. Those who believe that others are honest consider evasion more morally wrong than those who believe evasion to be widespread. Results from experiments by Fischbacher et al. (2001) suggest that a belief about other’s cheating cause to subjects’ own cheating behavior. It is understood that individuals are much more likely to behave in a non-compliant manner when they perceive that non-compliance is widespread. Three possible reasons can explain this fact. First, if reciprocity and equality are fundamental human attitudes, individuals would be unwilling to pay taxes to restore their own contribution to equitable outcomes or to punish those who do not participate in collective action even if they had an intrinsic motivation initially. Fehr and Gächter (2000) demonstrate that when subjects have an option to punish free riders, they may choose this costly option. Second, if people observe that cheating is widespread among their peers, they can consider being lower the risk of being caught. Kahan (1997) emphasizes that criminal behavior has a strong signaling effect. Finally, people who believe that tax evasion is prevalent among taxpayers may conclude that the psychological costs of guilt, shame, and loss of reputation are low. Alm et al. (1999) show that subjects who learn that others refuse to punish tax evaders demonstrate the low level of compliance, because the outcome may have sent a signal that tax evasion is socially acceptable.

Which Norms?: Personal Value Orientations and Identities Obviously, individuals do not obey all social norms and sometimes even refuse to comply with social norms. On the other hand, it cannot be said that paying taxes is a generally accepted moral standard/social norm for all cultures and/or sub-communities. Also, Elster (1989) emphasizes that in case of conflicting norms, social norms can easily be manipulated by individuals, in accordance with their self-interest. In that case, how and which types of social norms effect individual behavioral decisions on tax compliance? To determine whether or not anticipated social approval or disapproval effects behavior, social norms need to be refined. Cialdini (2007) identifies two types of social norms to determine effects on social influence. ‘Injunctive norms’ refer to perception about others’ beliefs concerning appropriate behaviors, and ‘descriptive norms’ refer to widespread perceptions of the behaviors among others. Injunctive norms

Journal of Applied Business and Economics vol. 16(3) 2014 105 are especially important in terms of the functioning of psychological costs. ‘Subjective norms’ as a particular form of injunctive norms are a person’s perception on expectations of a referent group such as family or friends (Cialdini and Trost, 1998). Individuals are more receptive to the thoughts of significant others than to those of general members of society. Therefore, analyzing the strength of norms impact on personal action/behavior should consider a person’s internalization of a social norm as a personal norm. Once subjects have internalized these norms as their own, self-based behavior standards, these norms can be called personal norms. Thus, subjects punish themselves through this norm when they violate it (Schwartz, 1977; Elster, 1989). In large-scale, social dilemma problems including high anonymity and low solidarity such as taxation, personal norms have more impact than social norms. Therefore, social influence has a special impact if there are internalized norms and shared identity with the group that imposes the norm. It is understood that individuals are especially sensitive to the views of persons who are important to them, and these individuals can easily internalize the thoughts of the referent group. Individuals who identify with the community’s identity internalize the community’s social norms. Therefore, the taxpayer’s perception on views of referent group is important in explaining compliance behavior. Wenzel (2004, p.216) emphasizes that “when influence source (i.e., those others whose taxpaying beliefs and behaviors we are faced with) is not part of one’s self-category, or the group with which one identifies, one does not expect to agree with them.” Bobek et al. (2007) found that especially subjective and personal norms are the most important factors to explain tax compliance, whereas descriptive norms are not significant factors in compliance choice. Wenzel (2004) and Wenzel (2007) found that social norms can influence the behavior of tax compliance only through a process of self-categorization. If persons do not have group identity, the group’s social norms do not have any impact on compliance behavior. Inclusiveness levels of identity, such as a nation, a profession, or an autonomic individual, is also important in participation in collective action. If taxpayers define their identity at the national level instead of a subcategory or autonomy, it is more likely that such taxpayers will have high levels of intrinsic value to participate in national level collective actions like taxation. Wenzel (2002), Torgler (2003a) and Martinez-Vazquez and Torgler (2009) provide evidence of the importance of national pride and national belonging in tax compliance. Apparently, individuals may have different motivations and orientations to comply with social norms (Cialdini and Trost, 1998). Gächter (2007) describes the heterogeneity of people’s cooperation preferences as “types of players.” Falk et al. (2003) found differences in subjects’ inclination to display social interactions in an experiment. Fehr and Schmidt (2003) found that some fair-minded people have a desire to reciprocate, while others are purely self-interested. However, behavior depends on beliefs and the strategic environment in which people interact. Schnellenbach (2010) emphasizes that, in reciprocal relationships with the public sector, while some consider only their own welfare, others consider societal measures of fairness. Again according to Schnellenbach, it may be reasonable to distinguish between reciprocally-minded individuals and self-interested individuals. Although social context and social norms have a strong influence on individual behavior, individuals systematically differ in the manner in which they approach others and in their social value orientation. Some people approach others cooperatively, while others exhibit less cooperation. Social value orientations are important in predicting a willingness to cooperate in social dilemmas (Van Lange et al., 1997). Depending on social context, individual personalities may change with individualistic and collectivistic tendencies in social interaction with regard to self-definition of group membership and interrelationships with others. Some individuals may rely more strongly on features of their personal identity rather than their social identity in social contexts and vice versa (Turner et al., 1987). Liebrand et al. (1986) distinguish value orientations into four types: altruism, cooperation, individualism and competition. They found that subjects with strong cooperative social motives are more sensitive to the moral obligations and social norms in social dilemmas. With respect to compliance literature, Trivedi et al. (2003) found that altruistic value orientation, rather than individualistic orientation, increases tax compliance when audits are completely absent.

106 Journal of Applied Business and Economics vol. 16(3) 2014 It can be concluded several important issue on tax compliance and how taxpayers become ‘insiders’ from discussion in this section. First, individual compliance behavior is influenced by the behavior of others’. However, internalized personal norms and expectations of significant others are especially important in explaining tax compliance behavior. If there are no internalized personal norms, the social context would not have any regulative effect on behavior, even if compliance with laws is a widely accepted social norm. Second, individuals have identities and self-definitions about themselves with respect to others. Although self-categorization and identities are not static, they can encourage or discourage cooperation with others. If individuals define themselves with a national identity rather than with a sense of alienation from society or a sub-identity, they will more likely contribute collective actions at the national level. Third, individuals differ with respect to reciprocally-mindedness, individualistic or collectivist tendencies, and the level of sensitiveness to the views of others. If individuals are strongly reciprocally-minded and sensitive to the expectations of family and friends, they may have a stronger sense of moral obligation to taxation. Again, if individuals have collectivistic orientations and feelings of personal responsibility towards collective welfare, that they may be cooperative in collective action. Persons who have individualistic tendencies may have an autonomist identity instead of a collective identity, and they may show less cooperative intentions. Individuals with collectivist orientation and sensitiveness to subjective norms have more likely internalized tax paying norms.

EMPIRICAL ANALYSIS

Data and Variables The data used in this study were obtained from the five waves of World Values Survey (WVS). The number of observations included in the analyses is 39,966 after excluding missing cases. WVS provides the opportunity to investigate social, political and cultural variables as well as personalities and moral sentiments around the world. The study conducts a logistic regression analysis to predict whether or not there is a correlation between the level of tax morale and variables in a socio-politic context. It uses tax morale (TM) as the dependent variable by following the research of Torgler (for example, 2003b, 2007). The request, with a ten-point possible rating, that indicates tax morale is as follows: “Please tell me for each of the following statements whether you think it can always be justified, never be justified, or something in between… Cheating on taxes if you have a chance” (1 = never justifiable; 10 = always justifiable). An immediate question with data about values or attitudes is whether or not the replies are truthful. This fact may be more important in a sensitive matter such as taxation because of the tendency to overstate compliance. In addition, it can be argued that using a single question to measure tax morale is insufficient. However, because the data set includes wide-ranging questions, a single tax morale question can reduce framing bias. Torgler and Schneider (2007) argue that WVS questions on tax morale have some advantage despite some biases. By following Torgler and Schneider (2007), this variable has been recoded into a dichotomized variable which takes the value 1 for ‘never justifiable’ and 0 for all situations to indicate low tax morale to estimate the odds of high tax morale (TM=1) in logistic regression. Another limitation of the methodology is that some explanatory variables such as tolerance to fiscal offences are likely co-determined with individual tax morale. Results should be interpreted by considering this potential simultaneity. Table 1 defines variables in detail.

Journal of Applied Business and Economics vol. 16(3) 2014 107 TABLE 1 DEFINITIONS OF VARIABLES

Variable WVS Questionnaire and Codes Notes Confidence in “Could you tell me how much confidence you have in them…. Reverse coded Government (CGOV) Government” [E069_11] Feeling of Citizenship “I see myself as citizen of the [country] nation” [G021] Reverse coded (CITZ) Confidence in Societal “Could you tell me how much confidence you have in [the following]” All variables were Institutions (CSOC) … Churches [E069_01] / … The Press [E069_04] / … Labor Unions recoded reverse and [E069_05] / … Television [E069_010] / … Major Companies [E069_013] indexed with standard / … Charitable or humanitarian organizations [E069_40] method. Societal Involvement “Could you tell me whether you are a member, an active member, an Variables were recoded (INV) inactive member or not a member of that type of organization?” 0 for not member, .5 for … Church or religious organization [A098] / … Environmental inactive member and, 1 organization [A103] / … Professional organization [A104] / … for active member. Charitable/humanitarian organization [A105] Then, the scores were summed. Role of Government “People should take more responsibility to provide for themselves vs the (GOV) government should take more responsibility to ensure that everyone is provided for” [E037] Tolerance to Fiscal “Please tell me for each of the following statements whether you think it All variables were Offences (TOLE) can always be justified, never be justified, or something in between…” recoded reverse and … Claiming government benefits to which you are not entitled [F114] / … indexed with standard Avoiding a fare on public transport [F115] / … Someone accepting a bribe method. in the course of their duties [F117] Subjective Norm “One of my main goals in life has been to make my parents proud” [D054] Recoded as a dummy Sensitiveness: Parents (SUB1) Subjective Norm “I make a lot of effort to live up to what my friends expect” [D055] Recoded as a dummy Sensitiveness: Friends (SUB2) Reciprocity: Personal “It is humiliating to receive money without having to work for it” [C037] Recoded as a dummy (REC1) Reciprocity: Societal “Work is a duty toward society” [C039] Recoded as a dummy (REC2) Schwartz: “Would you please indicate for each description whether that person is A mean rating was Collectivist/Individual very much like you, like you, somewhat like you, not like you, or not at all obtained over all values ist (SCHW) like you?” for each respondent. … It is important to this person to think up new ideas and be creative; to Then, each value score do things one’s own way. [A189] / … It is important to this person to be was subtracted from the rich; to have a lot of money and expensive things. [A190] / … Living in mean rating to isolate secure surroundings is important to this person; to avoid anything that the priority of the might be dangerous. [A191] / … It is important to this person to have a respondent relative to good time; to “spoil” oneself. [A192] / … It is important to this person to other values. help the people nearby; to care for their well-being [A193] / … Adventure Collectivist- and taking risks are important to this person; to have an exciting life. Individualist Index: [A195] / … It is important to this person to always behave properly; to (A191 + A193 + A196 avoid doing anything people would say is wrong. [A196] / … Tradition is + A198) – (A189 + important to this person; to follow the customs handed down by one’s A190 + A192 + A195) religion or family. [A198] A positive index value shows collectivistic orientation; a negative value shows individualistic orientation. Obedience to Important child qualities: Obedience [A042] Recoded as a dummy Authority (OBED) NOTE: Expressions in square brackets question codes in the WVS Integrated Questionnaire which can be obtained from www.worldvaluessurvey.org

108 Journal of Applied Business and Economics vol. 16(3) 2014 Four variables have been chosen to demonstrating political interaction and social capital: confidence in government (CGOV), feeling in citizenship (CITZ), confidence in societal institutions (CSOC), and societal involvement (INV). In addition to these variables, two more variables are used to measure the role of government (GOV) and attitudes toward other fiscal offences (TOLE) in the analysis. To determine subjects’ sensitiveness to subjective norms, two variables, making parents proud (SUB1) and as living up friends expectations, are used (SUB2). Two variables are chosen to determine whether or not subjects are reciprocally-minded. The first variable (REC1) indicates subjects’ reciprocal tendencies in personal relations by questioning if they find receiving money without work humiliating. The second variable (REC2) aims to measure if subjects feel reciprocal tendencies in relations with society. To measure personal value orientations as individualistic or collectivistic from Schwartz values, WVS’s ten- item version is used. Schwartz values are commonly used in social sciences to measure basic values. Eight variables are drawn from Schwartz value questions, and an individualistic-collectivistic value index is calculated.. Welzel (2010) is followed to transform the Schwartz value items. A final variable (OBED) related to personal orientation is obedience tendency to authority. This dichotomized variable is obtained from a question related to respondents’ value of obedience in child qualities. If respondents have a norm of obedience to authority and law, they may easily internalize social norms, legal regulations, and political decisions and exhibit high tax morale (Orviska and Hudson, 2002). In addition to these predictors, three demographic variables of gender, age and education are used as estimators. When tax morale is taken as a dichotomous variable (high level=1; low level=0), the probability of a high tax morale as it relates to predictor (independent) variables and the binary logistic regression equation that has been converted into a linear by using the natural logarithm of the odds are estimated as follows: log ( ) = + + + + + + + 1 + 2 + 1 + 2 + + + + + (1) 푇푀=1 0 1 2 3 4 5 6 7 8 푂 훽 훽 퐶퐺푂푉 훽 퐶퐼푇푍 훽 퐶푆푂퐶 훽 퐼푁푉 훽 퐺푂푉 훽 푇푂퐿퐸 훽 푆푈퐵 훽 푆푈퐵 9 10 11 12 13 14 15 In the analyses,훽 푅퐸퐶the forced훽 푅퐸퐶entry method훽 푆퐶퐻푊 (as one훽 block)푂퐵퐸퐷 was훽 used.퐺퐸푁퐷퐸푅 It can훽 be퐴퐺퐸 considered훽 퐸퐷푈 that forced entry can be considered to be more useful to test a theory because a hierarchical entry is influenced by random variation in the data. However, before the final analysis, the model was estimated using a hierarchical entry by blocking subsets of predictors and adding other subsets. Still, any important difference with the forced entry method was not observed in the results.

Empirical Results Goodness-of-Fit and Overall Success of the Model The Hosmer and Lemeshow test, which is based on a chi-square statistic calculated from observed and predicted probabilities, indicates that the full model versus a constant-only model is statistically significant (χ2 = 1374.765; df = 8; p = 000). The predictors as a set reliably distinguish between the levels of tax morale. As is known, the logistic regression has no measurement completely satisfying R2 in ordinal least square (OLS) regression models to provide variance in the dependent variable explained by the independent variables. However, in logistic regression, there are two pseudo R-square measurements to quantify the proportion of explained variation, despite the fact that their magnitudes are not exactly equivalent to those obtained in non-logistic regression. These measurements in the model were .301 for Cox and Snell R2 and .407 for Nagelkerke R2. Pseudo R-square indicates a moderate relationship between prediction and grouping. However, it is typically common to have a low adjusted R2 for logistic regression models. Another way to assess how well the model fits is to look at the ability of the model to accurately predict probabilities to assign cases. The classification table displayed the overall percentage of correctly classification of the model as equal to 76.1%. The overall percentage of cases that are correctly predicted by the model is 59.9% for low tax morale and 86.9% for high tax morale. While the base model with only-constant correctly predicted 60.2 percent of cases, the full model correctly classified 76.1% of cases

Journal of Applied Business and Economics vol. 16(3) 2014 109 which indicates an improvement over the base model. The Omnibus test of model coefficients (χ2 = 14074.565, df = 17, p = .000) also indicates the improvement of the full model over the base model.

Evaluation of Predicted Probabilities for the Model The statistical significance of the individual regression coefficients (βs) was tested using the Wald χ2 in Table 2. The Wald criterion demonstrates that all predictors make a significant contribution to prediction of tax morale. Significant levels are p<0.10 for age (30-49), p<0.05 for education (middle level), and p<0.01 for all other predictor variables. Generally, there is a correlation between tax morale and predictor variables. Since it is difficult to interpret response variable for increments or decrements in natural logs, it can be considered the corresponding multiplicative model that displays the odds rather than the log of the odds. Exponentiated values of odds are presented in Table 2 as ExpB. Table 2 presents results of odds and probabilities for predictors from logistic regression analysis as well as Wald statistics. Exponentiated values of the coefficients indicate the change in odds resulting from a unit change in the predictor. Results for political interaction and social capital factors, political-societal trust, and national belonging and participation are important for high level tax morale. The predictor variable confidence in the government (CGOV) has an odds value of 1.301, which indicates that odds are increased by a factor of 1.301 when CGOV increases, controlling of other variables. A unit change in CGOV is associated with a change in the odds of high level membership in tax morale, with all other variables held constant. For every one point increase in feelings of citizenship, the odds of high tax morale likelihood increase by a factor of 1.179, all other factors being equal. The odds ratio for CSOC is 1.032. As confidence in societal institutions increases by one unit, there is a 3.2% chance of a high level tax morale, controlling for other variables. For membership to societal institutions (INV), an increase in membership scale by one unit increases the share of subjects indicating the high tax morale by 13.4 percentage points. The other two attitudes related to fiscal relations with the government and society also displayed a positive correlation with high tax morale. The odds ratio for the role of government (GOV) is 1.027’dir. An increase in the scale indicating the government should take more responsibility by one unit increases the share of subjects in the category of high tax morale by 2.7 percentages, controlling for other variables. Expectations for government responsibility can be assumed to be a proxy of economic collectivist/ libertarian orientation; the results show that economic collectivists are more likely to have high tax morale. For every one unit increase in a respondent’s attitude against fiscal offences (TOLE), the likelihood of high tax morale increases by 2.292 times after controlling for other factors. This predictor is the most important variable in the model. Porcano (1988) indicates that taxpayers’ general honesty is important in explaining compliance behavior. If individuals find these behaviors that most important part of fiscal exchange with the government to be unacceptable, they would find tax fraud unacceptable as well. As a subject’s sensitiveness to family (SUB1) and friends’ expectations (SUB2) increases, the likelihood of high tax morale increases (1.163 times for parent’s expectations and 1.098 times for friends’ expectations). This finding supports the view that internalized personal norms are important to explain tax compliance by creating psychological costs such as feelings of shame and guilt. According to results, subjects who have high tax morale are more sensitive to their families than friends. Individuals who find it important to make their families proud have a high intrinsic motivation to pay taxes. Reciprocal attitude is raised by one unit, odds ratios are 1.281 and 1.258 times as large, and therefore, persons who have high reciprocal tendencies in personal relations (REC1) and/or feel duty toward society (REC2) are more likely to have high tax morale. The Schwartz index (SCHW) indicates subjects’ collectivist or individualist orientations. According to analysis findings, given an increase in collectivist tendency by one unit, the likelihood of high tax morale increases by 1.047, controlling for other variables. While an individual’s obedient attitude toward authority (OBED) increases, the likelihood of tax morale increases by a 1.127 odds ratio, controlling for other variables. Persons who have a norm of

110 Journal of Applied Business and Economics vol. 16(3) 2014 obedience to authority also have a likelihood of high tax morale. This finding is also consistent with the hypothetical expectations.

TABLE 2 LOGISTIC REGRESSION PREDICTING FOR TAX MORALE (TM=1)

Variables B S.E. Wald df Sig. Exp(B) Confidence in Government (CGOV) .263 .044 36.049 1 0.000 1.301 Feeling Citizenship (CITZ) .164 .021 59.996 1 0.000 1.179 Confidence in Societal Institutions (CSOC) .031 .011 8.339 1 0.004 1.032 Societal Involvement (INV) .126 .015 70.614 1 0.000 1.134 Role of Government (GOV) .027 .005 35.541 1 0.000 1.027 Tolerance to Fiscal Offences (TOLE) .829 .010 7397.319 1 0.000 2.292 Subjective Norm Orientation: Parents (SUB1) .151 .033 20.711 1 0.000 1.163 Subjective Norm Orientation: Friends (SUB2) .093 .026 12.479 1 0.000 1.098 Reciprocity: Personal (REC1) .248 .032 58.222 1 0.000 1.281 Reciprocity: Societal (REC2) .230 .031 55.032 1 0.000 1.258 Schwartz: Collectivist-Individualist (SCHW) .046 .003 240.717 1 0.000 1.047 Obedience to Authority (OBED) .119 .027 19.699 1 0.000 1.127 Gender (Female=1) .072 .025 7.946 1 0.005 1.074 Age 32.078 2 0.000 (1) 15-29 years .190 .035 29.770 1 0.000 1.209 (2) 30-49 years .056 .031 3.229 1 0.072 1.057 Education 36.931 2 0.000 (1) Lower level .208 .036 34.144 1 0.000 1.231 (2) Middle level .070 .032 4.757 1 0.029 1.072 Constant -8.567 .134 4066.836 1 0.000 .000 NOTE: Reference categories are ‘50 and more’ for age and ‘high level’ for education.

CONCLUSION

This paper examines the importance of personal values and identities as well as interactions with political institutions and the society for becoming an insider with high intrinsic taxpaying motivation. This paper suggests three dimensions of being an insider. First, interaction with political institutions shapes a taxpayer’s identity in terms of a feeling of belonging to a collective decision-making system (and the society) and an acceptance of legitimate authority. Second, interaction with others in the society affects behavior through social influence and social norms. Reciprocity and fairness norms are moderators for both interactions. How extensively taxpayers are affected by these relationships depends on personal values and identities acquired from complex social interactions. As it is pointed out by Akerlof and Kranton (2000 and 2010), an individuals’ sense of self and self-definition are important determinants for economic and social structure decisions. This paper suggests that social context has an impact on compliance behavior to an extent, especially in an individual’s self-identity and personal value orientations. The results from the estimated logistic regression based on WVS data present evidence for these hypothetical expectations. Results confirm that respondents’ sensitiveness to expectations of parents and friends, reciprocal tendencies, collectivist value orientation as well as confidence in government and

Journal of Applied Business and Economics vol. 16(3) 2014 111 societal institutions, a sense of belonging to a national identity, and participation in collective action are important for having high tax morale. It should be noted that this study, however, has several limitations because of limited measuring of tax morale and a possible frame effect. Recalling Akerlof and Kranton (2010), a firm will be willing to invest in workers to make them insiders because insiders are willing to work harder despite lower pecuniary utilities. Therefore, it can be suggested that investing in citizens to make taxpayers insiders may be more effective and cheaper than audit activities. Thus, due attention should be placed on any policy aimed at creating insiders.

REFERENCES

Akerlof, G.A. & Kranton, R.E. (2000). Economics and identity. The Quarterly Journal of Economics, 115, (3), 715-753. Akerlof, G.A. & Kranton, R.E. (2008). Identity, supervision and work groups. American Economic Review, 98, (2), 212-217. Akerlof, G.A. & Kranton, R.E. (2010). Identity economics: How our identities shape our work, wages, and well-being, New Jersey: Princeton University Press. Allingham, M.G. & Sandmo, A. (1972). Income tax evasion: A theoretical analysis. Journal of Public Economics, 1, 323-338. Alm, J. & Torgler, B. (2011). Do ethics matter? Tax compliance and morality. Journal of Business Ethics, 101, 635-651. Alm, J., Jackson, B.R. & McKee, M. (1992). Estimating the determinants of taxpayer compliance with experimental Data. National Tax Journal, 45, (1), 107–114. Alm, J., Jackson, B.R. & McKee, M. (1993). Fiscal exchange, collective decision institutions, and tax compliance. Journal of Economic Behavior and Organization, 22, 285–303. Alm, J., McClelland, G.H. & Schulze, W. D. (1999). Changing the social norm of tax compliance by voting. Kyklos, 52, 141-171. Alm, J., Sanchez, I. & De Juan, A. (1995). Economic and noneconomic factors in tax compliance. Kyklos, 48, 3-18. Andreoni, J. (1990). Impure altruism and donations to public goods: A theory of warm-glow giving. The Economic Journal, 100, 464-477. Bobek, D.D., Roberts, R.W. & Sweeney, J.T. (2007). The social norms of tax compliance: Evidence from Australian, Singapore, and the United States. Journal of Business Ethics, 74, 49-64. Cialdini, R.B. (2007). Descriptive social norms as underappreciated sources of social sontrol. Psychometrika, 72(2), pp. 263-268. Cialdini, R.B. & Trost, M.R. (1998). Social influence: Social norms, conformity and compliance. In Gilbert, D.T., Fiske, S.T. & Lindzey, G. (eds.) The Handbook of Social Psychology, 4th Edition Vol. 2, (pp. 151-192). New York and Oxford: McGraw-Hill. Cullis, J. Jones, P. & Savoia, A. (2012). Social norms and tax compliance: Framing the decision to pay tax. The Journal of Socio-Economics, 41, 159– 168. Cummings, R. G., Martinez-Vazquez, J., McKee, M., & Torgler, B. (2009). Tax morale affects tax compliance: Evidence from surveys and an artefactual field experiment. Journal of Economic Behavior & Organization, 70, 447–457. Erard, B. & Feinstein, J. (1994). The Role of moral sentiments and audit perceptions in tax compliance. Public Finance/Finances Publiques, 49 (supplement), 70-80. Falk, A., Fischbacher, U. & Gächter, S. (2003). Living in two neighborhoods – Social interaction in the lab. CESifo Working Papers, No.954, May. Fehr, E & Gächter, S. (2000). Cooperation and punishment in public goods experiments. American Economic Review, 90, (4), 980–994. Fehr, E. & Schmidt, K.M. (1999). A theory of fairness, competition and cooperation. Quarterly Journal of Economics, 114, (3), 817-68.

112 Journal of Applied Business and Economics vol. 16(3) 2014 Fehr, E. & Schmidt, K.M. (2003). Theories of fairness and reciprocity: Evidence and economic applications. In Dewatripont, M., Hansen, L. and Turnovsky, S. (eds), Advance in Economics and Econometrics: Theory and Application Vol.1 (pp. 208-257). Cambridge: Cambridge University Press. Feld, L.P. & Frey, B.S. (2002). Trust breeds trust: How taxpayers are treated. Economics of Governance, 3, 87-99. Feld, L.P. & Frey, B.S. (2005). Illegal, immoral, fattening or what?: How deterrence and responsive regulation shape tax morale. In Bajada, C. and Schneider, F. (eds.) Size, Causes and Consequences of the Underground Economy (pp. 15-37). England: Ashgate. Feld, L.P. & Frey, B.S. (2007). Tax compliance as a result of a psychological tax contract: The role of incentives and responsive regulation. Law & Policy, 29, (1), 102-120. Feld, L.P. & Frey, B.S. (2010). Tax evasion and the psychological tax contract. In Alm, J., Martinez- Vazquez, J. & Torgler, B. (eds.) Developing Alternative Frameworks for Explaining Tax Compliance (pp. 74-94). London and New York: Routledge. Feld, L.P. & Tyran, J. (2002). Tax evasion and voting: An experimental analysis. Kyklos,55, 197–222. Fischbacherr, U. & Gächter, S. (2010). Social preferences, beliefs and the dynamics of free riding in public goods experiments. American Economic Review, 100, (1), 541-556. Fischbacherr, U., Gächter, S. & Fehr E. (2001). Are people conditionally cooperative? Evidence from a public goods experiment. Economic Letters, 71, 397-404. Fortin, B., Lacroix, G. & Villeval, M. (2007). Tax evasion and social interactions. Journal of Public Economics, 91, 2089-2112. Frey, B.S. (1997). A constitution for knaves crowds out civic virtues. Economic Journal, 107, 1043-1053. Frey, B.S. (2003). Deterrence and tax morale in European Union. European Review, 11, (3), 385-406. Frey, B.S. & Torgler, B. (2007). Tax morale and conditional cooperation. Journal of Comparative Economics, 35, 136-159. Gächter, S. (2007). Conditional cooperation: Behavioral regularities from the lab and the field and their policy implications. In Frey, B.S. and Stutzer, A. (eds.) Economics and Psychology: A Promising New Cross-Disciplinary Field (pp.19-50). Cambridge: The MIT Press. Gordon, J.P.F. (1989). Individual morality and reputation costs as deterrents to tax evasion. European Economic Review, 33, 797–805. Hardin, R. (2006). Trus,. Cambridge and Malden: Polity Press. Kahan, D.M. (1997). Social influence, social meaning, and deterrence. Virginia Law Review, 83, 349-394. Kerr, N.L. (1995). Norms in social dilemmas. In Schroeder, D.A. (ed.) Social Dilemmas: Perspectives on Individuals and Groups (pp. 31-48). Westport: Praeger. Kirchler, E. (2007). The Economic psychology of tax behaviour. Cambridge: Cambridge University Press. Kucher, M. & Götte, L. (1998). Trust me: An empirical analysis of taxpayer honesty. Finanzarchiv, 55,(3), 429-44. Liebrand, W.B.G., Wilke, H.A.M., Vogel, R. & Wolters, F.J.M. (1986). Value orientation and conformity: A study using three types of social dilemma games. The Journal of Conflict Resolution, 30, (1), 77-97. Martinez-Vazquez, J. & Torgler, B. (2009). The evolution of tax morale in modern Spain. Journal of Economic Issues, 43, (10), 1-28. Orviska, M. & Hudson, M. (2002). Tax evasion, civic duty and the law abiding citizen. European Journal of Political Economy, 19, 83-102. Ostrom, E. (1998). A behavioral approach to rational-choice theory of collective action. American Political Science Review, 92, 1-23. Pommerehne, W.W. & Weck-Hasnemann, H. (1996). Tax rates, tax administration and income tax evasion in Switzerland. Public Choice, 88, 161-170. Porcano, T.M. (1988). Correlates of tax evasion. Journal of Economic Psychology, 9, 47-67. Putnam, R. D., Leonardi, R. & Nonetti, R.Y. (1993). Making democracy work: Civic traditions in modern Italy. Princeton: Princeton University Press.

Journal of Applied Business and Economics vol. 16(3) 2014 113 Rabin, M. (1993). Incorporating fairness into game theory and economics. American Economic Review, 83, (5), 1281-1302. Schnellenbach, J. (2010). Vertical and horizontal reciprocity in a theory of taxpayer compliance. In Alm, J., Martinez-Vazquez, J. & Torgler, B. (eds.) Developing Alternative Frameworks for Explaining Tax Compliance (pp. 56-73). London and New York: Routledge. Scholz, J. & Lubell, M. (1998). Trust and taxpayers: Testing the heuristic approach to collective Action. American Journal of Political Science, 42, (2), 398-417. Schwartz, S.H. (1977). Normative influences on altruism. In Berkowitz, L. (ed.) Advances in Experimental Social Psychology Vol.10 (pp. 221-279). New York: Academic Press. Seidl, C. &Traub, S. (2001). Taxpayers’ attitudes, behaviour, and perception of fairness. Pacific Economic Review, 6, (2), 255-267. Slemrod, J. (2003). Trust in public finance. In Cnossen, S. and Sinn, H. (eds.) Public Finance and Public Policy in the New Century (pp. 49–88). Cambridge: The MIT Press. Smith, K.W. (1992). Reciprocity and fairness: Positive incentives for tax compliance. In Slemrod, J. (ed.) Why People pay Taxes: Tax Compliance and Enforcement (pp. 223-250). Ann Arbor: The University of Michigan Press. Spicer, M. (1986). Civilization at a discount: The problem of tax evasion. National Tax Journal, 39, 13- 20. Torgler, B. (2003a). To evade taxes or not to evade: That is the question. Journal of Socio-Economics, 32, (3), 283–302. Torgler, B. (2003b). Tax morale, rule-governed behaviour and trust. Constitutional Political Economy, 14, 119–140. Torgler, B. (2004). Cross culture comparison of tax morale and tax compliance: Evidence from Costa Rica and Switzerland. International Journal of Comparative Sociology, 45, 17–43. Torgler, B. (2005). Tax morale and direct democracy. European Journal of Political Economy, 21, 525- 531. Torgler, B. (2007). Tax compliance and tax morale: A theoretical and empirical analysis. Cheltenham: Edward Elgar. Torgler, B. (2008). What do we know about tax fraud? An overview of recent developments. Social Research, 75, (4), 1239-1270. Torgler, B. and Schaltegger (2005). Tax amnesties and political participation. Public Finance Review, 33, (3), 403-431. Torgler, B. & Schneider, F. (2007). What shapes attitudes toward paying taxes? Evidence from multicultural European countries. Social Science Quarterly, 88, (2), 443-470. Torgler, B., Schaffner, M. & Macintyre, A. (2010). Tax compliance, tax morale, and governance quality. In Alm, J., Martinez-Vazquez, J. & Torgler, B. (eds.) Developing Alternative Frameworks for Explaining Tax Compliance (pp. 141-173). London and New York: Routledge. Traxler, C. (2010). Social norms and conditional cooperative taxpayers. European Journal of Political Economy, 26, 89–103. Trivedi, V.U., Shehata, M. & Lynn, B. (2003). Impact of personal and situational factors on taxpayer compliance: An experimental analysis. Journal of Business Ethics, 47, 175-197. Turner, J.C., Hogg, M.A., Oakes, P.J., Reicher, S.D. & Wetherell, M.S. (1987). Rediscovering the social group: A self-categorization theory. Oxford and New York: Blackwell. Tyler, T.R. (2000). Why do people cooperate in groups? Support for structural solutions to social dilemma problems. In Vugt, M.V., Snyder, M., Tyler, T. R. and Biel, A. (eds.) Cooperation in Modern Society: Promoting the Welfare of Communities, States and Organizations (pp. 64-82). London and New York: Routledge. Tyler, T.R. & Lind, E.A. (1992). A relational model of authority in groups. In Zanna, M.P. (ed.), Advances in Experimental Social Psychology Vol. 25 (pp. 115–191). UK: Academic Press. Uslaner, E. M. (2002). The moral foundations of trust. Cambridge: Cambridge University Press.

114 Journal of Applied Business and Economics vol. 16(3) 2014 Uslaner, E.M. (2010). Tax evasion, corruption and the social contract in transition. In Alm, J., Martinez- Vazquez, J. & Torgler, B. (ed.) Developing Alternative Frameworks for Explaining Tax Compliance (pp. 174-190). London and New York: Routledge. Van Lange, P.A.M., Otten, W., De Bruin, E.M.N. & Joireman, J.A. (1997). Development of prosocial, individualistic, and competitive orientations: Theory and preliminary evidence. Journal of Personality and Social Psychology, 73, (4), 733-746. Welzel, C. (2010). How selfish are self-expression values? A civicness test. Journal of Cross-Cultural Psychology, 1, (2), 152-174. Wenzel, M. (2002). The Impact of outcome orientation and justice concerns on tax compliance: The role of taxpayers’ identity. Journal of Applied Psychology, 87, 629–45. Wenzel, M. (2004). An analysis of norm processes in tax compliance. Journal of Economic Psychology, 25, 213-228. Wenzel, M. (2006). A letter from the tax office: Compliance effects of informational and interpersonal justice. Social Justice Research, 19, (3), 345-364. Wenzel, M. (2007). The Multiplicity of taxpayer identities and their implications for tax ethics. Law & Policy 26, (1), 31-50.

Journal of Applied Business and Economics vol. 16(3) 2014 115

Credit Search

Simangaliso Biza-Khupe University of Botswana

INTRODUCTION

Consumer external information search is the time, effort, attention and money that an individual expends towards obtaining environmental information relating to a product being considered for either an imminent or future purchase (Schmidt & Spreng, 1996; Srinivasan & Ratchford, 1991). ‘External search begins when one first considers the purchase seriously and ends with the actual purchase. This effort is affected by information that consumers obtain prior to considering the purchase’ (Srinivasan & Ratchford, 1991, p. 253). Understanding how consumers search and gather information is therefore precursory to explaining consumer rationalisation behaviour. This area of research has gained momentum because of its increasing relevance in today’s dynamic global environment where boundaries of innovation, and hence the order of market operations, are on a constant shift. The role that such micro entities as individuals and households play in the process has been found particularly pivotal in determining market efficiency. For example, to the extent that some researchers have attributed poor credit decisions and choices by individuals and households as one of the causes of the recent Global Financial Crisis, and in particular excessive leverage, increased risk appetite and the global phenomenal rise in personal bankruptcies (Begg, 2009; Kay, 2009), scholars and policymakers have in the recent past become increasingly interested in understanding consumer financial behaviour. The literature is replete with studies on consumer information search, search strategies (Kim & King, 2009; Mattila & Wirtz, 2002), and the myriad of antecedents to information search (Srinivasan & Ratchford, 1991; Punj & Staelin, 1983). Most of these studies have been on consumer durables such as household appliances, clothing and vehicles. Moreover, goods are fundamentally distinct from services and consumers employ different search strategies across product class (Kim & King, 2009; Mitra, Reiss, & Capella, 1999). Research in consumer external search for services has not been as robust as in consumer search for goods (Heaney & Goldsmith, 1999) and therefore is not as well understood.

116 Journal of Applied Business and Economics vol. 16(3) 2014 There has been particular paucity in research on consumer search for financial instruments, i.e. credit and investment. The few studies that have been conducted on credit search, notably; Chang & Hanna (1992), Worden & Sullivan (1987), and Lee & Hogarth (2000a), have not considered an integrated structural model of consumer credit search, in spite of the fact that structural equation modelling (SEM) has been found to be a more powerful analytic technique than the conventional bivariate and regression analysis (Cheng, 2001). For example, SEM makes it possible to investigate considerably complex theoretical models in a relatively robust fashion, and thereby providing an alternative to enhancing theory and practice in personal finance (Tabachnick & Fidell, 2007; Schumacker & Lomax, 2004)). More critically, there is anecdotal evidence to suggest that consumer financial markets are subjected to financial information oversupply, also known as Data Smog. This proposition is born out of an observation that the statutory financial disclosures promulgated by the ‘truth in lending’ legislation, coupled with technological innovations that have made the production, retrieval and distribution of information easy, have led to information propagation and oversupply (Kozup, Howlett, & Pagano, 2008; Waddington, 1997). Thus, borrowers have been observed to be barraged by an aggregate mass of financial data involving disclosures that are extensive and complex (Lee & Cho 2005; Lee & Hogarth 1999b), which they find overwhelming (Bernthal, Crockett, & Rose 2005). In this regard, there are suggestions of a Data Smog in consumer financial markets (Lee & Cho 2005). Considering that individual search behaviour is a function of: (i) memory and cognitive processing capacity (Rischkowsky & Doring, 2008; Datta & Chatterjee, 2008) and; (ii) cost-benefit evaluation of choices (Kulviwat, Guo, & Engchanil, 2004; Guo, 2001; Stigler, 1961), Data Smog is a conceivable antecedent to consumer external search and purchase behaviour. The impact of Data Smog on consumer external search and purchase behaviour has not been investigated a priori. The paper contributes to the literature by pursuing two primary objectives. First, the paper proposes and tests a comprehensive structural model of consumer credit search which is comparable in complexity with those on consumer search for goods. Second, the paper investigates the impact of perceived data smog on consumer external search. The previous section was an introduction. The next section is an overview of literature of information search and a background discussion on Data Smog. Research hypotheses are developed from the literature and a theoretical model of consumer credit search proposed. Thereafter is a deliberation on research method and the results. The paper concludes with a discussion on the research findings and its implications to theory and practice.

CONSUMER SEARCH BEHAVIOUR – AN OVERVIEW OF LITERATURE

Consumer Information Search The literature has two established theoretical streams of consumer external information search behaviour: the psychological (or motivational/information processing) perspective and the economic (or cost-benefit) perspective (Kim & King, 2009; Schmidt & Spreng, 1996; Srinivasan, 1990). The psychological perspective is concerned with consumers’ cognitive abilities and motivation to process information, as predetermining factors to an effortful search. It operates from the premise that consumers’ inclination to actively engage in information gathering is affected by individual constraints, threshold, aspirations and capabilities (Thogersen, 2005). Conversely, the economic perspective projects consumers as sovereign economic entities motivated by rationale to calculate and match their consumption needs to offers in the market in a judicious manner (Rischkowsky & Doring, 2008). In this regard, a consumer is theorised to behave in a manner which maximises his/her net expected gains construed from estimates of the difference between expected positive and negative utility (Kulviwat et al., 2004). On the basis of the two theoretical streams, past research has identified dozens of variables that intervene in consumer external information search behaviour, as many as sixty (60) variables according to other studies (Srinivasan & Ratchford, 1991). Over the past three decades researchers have suggested that consumer external search behaviour is most realistically conceptualised as a series of interrelated activities among variables across the two theoretical streams (for example: Kulviwat et al., 2004; Punj &

Journal of Applied Business and Economics vol. 16(3) 2014 117 Staelin, 1983). Most crucially, it is suggested that the interwoven nature of this myriad of variables and their collective influence on consumer search behaviour could only be realistically captured if their interrelationships are simultaneously tested. The conventional approach by many studies where the exogenous variables are projected as independent of each other has been castigated for redundancy, possessing limited explanatory strength and most critically failing to project a realistic representation of the complexities of consumer behaviour (Kulviwat et al., 2004; Gursoy & McCleary, 2003; Guo, 2001). A study generally considered a pioneer in consumer search behaviour modelling was by Punj and Staelin (1983) who conducted a study on information search for new automobile purchases. The study posited that usable prior knowledge, prior memory structure, desire to seek information, cost of search and size of feasible set directly influence search. Also, the proposed model predicted search outcomes as satisfaction and cost savings. The study results found prior knowledge and search cost to be negatively related to search, and the desire to seek information and feasible set size to be positively related to search. Search was not found to significantly influence satisfaction with purchase decision, but was however found to positively relate to cost savings. A major shortcoming of the study was the validity of measures used, i.e. with four, out of nine, constructs measured by a single item, thereby failing to capture the multiplicity in the dimensions of the projected construct factors (Guo, 2001; Beatty & Smith, 1987), hence the study was subjected to strong criticism (Srinivasan & Ratchford, 1991). Also, the study had an underlying assumption of the non-existence of interrelationships among search predictor variables. These limitations may have contributed to the low squared multiple correlations (R² = 0.11) realised by the structural equation model tested in the study. Srinivasan and Ratchford (1991) expanded and improved on the Punj and Staelin (1983), although their model did not consider the outcomes of external search. The study proposed that search for automobiles was directly influenced by interest in cars, search cost, search benefits, experience and evoked set size, and indirectly influenced by risk, positive experience and product knowledge. The model remains one of the most comprehensive and complex in the literature (Gursoy & McCleary, 2003; Schmidt & Spreng, 1996). Among the direct relationships, size of evoked set, search benefits and search cost were found to be positively related to search effort, while the amount of experience was found to be negatively related to search effort. The study used apt multiple measures for the variables and provides exemplary guidance on consumer search research. In sum, research on consumer search behaviour for goods has been extensive and comprehensive theoretical models have been developed and tested. While this body of literature provides a good theoretical foundation, it is not a substitute for equally robust studies in consumer search behaviour for services (Guo, 2001; Heaney & Goldsmith, 1999). On the backdrop of widely acknowledged measurable differences between goods and services (Kim & King, 2009; Mattila & Wirtz, 2002), and the different search strategies consumers adopt across product classes (Mitra et al., 1999; Murray, 1991), independent studies on consumer search for services are imperative. Considering that the properties of services pose particularly vexing challenges to market players, the construction of service-oriented solutions for service-oriented problems would be the most apt strategy. Heaney and Goldsmith (1999) proposed and tested one of the most comprehensive structural models of consumer search for services. The study examined consumer information acquisition activities for banking services by proposing a model postulating search as influenced by perceived benefits, perceived risk, perceived search cost and prior knowledge. Further, the study proposed interrelationships among the predictor variables. The results indicated that perceived benefits, perceived search cost and prior knowledge significantly related to external search. There was no statistically significant relationship found between search effort and perceived risk. Notwithstanding the extensiveness of the model, in structure, scale measurement and findings, the study had some shortcomings. The use of a convenience sample could have affected external validity. Also, the term ‘banking services’ refers to a cluster of products which could possibly be perceived by consumers as characteristically different, and thus the element of variability may have influenced the outcome.

118 Journal of Applied Business and Economics vol. 16(3) 2014 Credit Search There is a paucity of studies on consumer credit search, and none of these studies are as extensive and rigorous as those in consumer search for goods. Lee & Hogarth (1999b) conducted a study to examine factors that impact returns on information search for consumer mortgage. Interest rate savings were hypothesised to be a function of search effort, term and source of loan, age and education of respondent, household size and region and credit experience. The study results found the mortgage’s APR to be significantly influenced by all the predictor variables. The study had some shortcomings, particularly with relation to the operationalisation of some constructs, for example search effort, which might have affected the results. In a study conducted to investigate consumer search behaviour in the credit card market, Kerr & Dunn (2008) found the size of the credit balance to have a positive and statistically significant effect on the probability of a credit card application rejection and credit search. These findings are contrary to earlier studies that postulated a negative relationship between credit balance and search, based on the notion that high-balance carrying (high default probability) borrowers are insensitive to interest rates, and therefore, not motivated to search (Calem & Mester, 1995). In as far as this author can determine, a study by Chang & Hanna (1992) is the most comprehensive published research in examining antecedents to consumer credit search. The study projected consumer credit search as a function of loan size, time constraint, income, age, prior experience and education. In a regression model, the results found loan size and education to be positively related to credit search. The hypothesis that income is negatively related to credit search was partially supported by curvilinear results. However, the data used was from secondary sources, an element which may have compromised the external validity of some of the variables used. Also, the constructs of credit search, prior credit experience and time constraint were each measured by single and somewhat dubious items. Time constraint was, for instance, measured as the condition of employment for a household. Most notably, though, the Chang & Hanna theoretical model is basic. As already discussed, variables that affect consumer search behaviour are extensive (economic and psychological theoretical streams) and intricately interrelated, an aspect that was not captured by the proposed model of credit search and thereby diminished its explanatory strength (Kulviwat et al., 2004; Gursoy & McCleary, 2003; Schmidt & Spreng, 1996). The model indicated a low R² of 0.10.

Data Smog in Consumer Credit Markets According to classical economics theory, the establishment of financially-informed consumer-driven markets is the pinnacle of market efficiency. In a consumer-driven market, consumers engage in judicious, effective and optimal financial decisions in acquiring debt and investing, and hence lenders compete for market share on the basis of price and other quantifiable market related terms (Lee & Hogarth, 1999a). This paradigm therefore views the problem of consumer over-indebtedness and personal bankruptcy as caused by, or at least symptomatic of, market imperfections (and information asymmetry in particular) (Ferretti, 2007; Malbon, 2001; Lee & Hogarth, 1999a). In a bid to restore market efficiency, many governments (particularly the USA and those in the OECD financial region) have intervened by imposing legislation prescribing minimum financial information disclosures from lenders to borrowers (Buch, Rhoda, & Talaga, 2002; Malbon, 2001; OECD, 1992), laws that have been dubbed ‘Truth in Lending’. Ordinarily, the information required to be disclosed by lenders under the ‘Truth in Lending’ laws includes finance charges as an annual percentage rate, the calculation of interest charges, the amount of repayments, other fees and charges, and other non-price information details (Kerr & Dunn, 2008; Malbon, 2001; Lee & Hogarth, 1999a). These statutory financial disclosures, coupled with technological innovations that have made the production, retrieval and distribution of information easy, have led to information propagation and oversupply (Kozup et al., 2008; Waddington, 1997). Thus borrowers have been observed to be barraged by an aggregate mass of financial data involving disclosures that are extensive and complex (Lee & Cho, 2005; Lee & Hogarth, 1999b). Also, a proliferation of credit product choices and a wide permutation of credit terms at the borrowers’ disposal

Journal of Applied Business and Economics vol. 16(3) 2014 119 have been observed to compound the problem (Bernthal, Crockett, & Rose, 2005). Overall, there is anecdotal evidence of a Data Smog in consumer credit markets (Lee & Cho, 2005). Drawing from the psychological theoretical perspective, the search and buying behaviour of individual is a determinant of memory and cognitive processing capacity. The theory argues that consumer memory and cognitive capacity to process information is limited (Bettman, Johnson, & Payne, 1991; Bettman, 1979), and hence Data Smog would intuitively be expected to have adverse effects on consumer rationalisation and decision-making processes. For example, in studies conducted in different countries information overload was found to play an increasingly significant role in inhibiting optimal decision making in today’s information cluttered marketplace (Walsh, Mitchell, & Hennig-Tharau, 2001; Hiu, Siu, Wang, & Chang, 2001). In a separate study designed to investigate the effect of financial information regulation on the competitiveness of credit markets, a substantial portion of consumers found it difficult and confusing to compare loan products and the available financial information was found not to allow easy comparison of loan products (Malbon, 2001). Moreover, Biza-Khupe (2011) probed underlying factors to the perception of information oversupply in consumer financial markets. The study postulated consumer perceived financial information oversupply in the financial markets to be directly influenced by the four factors of Cost of Search, Prior Memory Structure, Credit Knowledge and Age. Using multiple regression analysis, the results of the study found Cost of Search and Prior Memory Structure to be antecedents of consumer perceived financial information oversupply. In light of these studies, Data Smog is a factor to be reckoned in this field of study.

THE PROPOSED CONSUMER CREDIT SEARCH MODEL

Variables that have been found in the literature to influence consumer search behaviour are voluminous, as already discussed, and thus presenting a vexing challenge. A criterion was therefore formulated to determined variables for inclusion. First and foremost, the importance of incorporating both theoretical streams of consumer external search was recognised and adopted. Previous research has indicated that the two streams are intricately intertwined (Guo, 2001; Fodness & Murray, 1999) and when combined, the proposed model’s explanatory strength is amplified (Kulviwat et al., 2004; Gursoy & McCleary, 2003; Schmidt & Spreng, 1996). Second, it was considered apt to select the most widely used variables in the literature (Heaney & Goldsmith, 1999; Schmidt & Spreng, 1996; Srinivasan & Ratchford, 1991; Punj & Staelin, 1983) as means to allow for juxtaposition to other related studies. Finally, it was important that the selected variables were in synchrony with the study research objectives. Following from the above, five (5) independent variables were selected: i. Prior memory structure ii. Perceived cost of search (search cost) iii. Perceived data smog (data smog) iv. Perceived financial risk (financial risk) v. Prior credit knowledge (prior knowledge) Perceived risk has in the literature been defined as the probability of an anticipated loss compounded by the importance of that loss to the decision-maker (Heaney & Goldsmith, 1999; Srinivasan & Ratchford, 1991). The economic theoretical stream associates consumer uncertainty to perceived risk. Increased consumer uncertainty about a product and/or its domain increases the potential for consumers to commit to costly, suboptimal decisions and choices (Mitra et al., 1999; Murray & Schlacter, 1990). The intangibility, heterogeneity, inseparability and perishability properties of a service, vis-à-vis a good, increase both the variability in the essence of a service and uncertainty in its ascertainment by consumers (Zeithaml, Parasuraman, & Berry, 1985), particularly considering that services are difficult to evaluate even after purchase and consumption (Mattila & Wirtz, 2002; Murray, 1991). It is postulated in the literature that under circumstances of identified uncertainties regarding the outcome of an action and where there exists an identified discrepancy between internal and external information, consumers are inclined to engage in an effortful search as a risk reduction and utility maximisation strategy (Gursoy & McCleary, 2003; Mitra et al., 1999; McColl-Kennedy & Fetter, 1999).

120 Journal of Applied Business and Economics vol. 16(3) 2014 Conversely, lack of adequate information deprives a consumer of an opportunity to profit from good deals, good buys, cost-savings and other related returns of search (Srinivasan & Ratchford, 1991; Punj & Staelin, 1983). In sum, perceived risk induces an effortful search and it is therefore postulated that perceived financial risk is positively related to credit search. Thus, it is hypothesised that:

H1: Consumer credit search is positively related to perceived financial risk.

Prior credit knowledge is defined as credit-specific information that is readily accessible to a decision-maker from long-term memory during the external search process (Heaney & Goldsmith, 1999; Srinivasan & Ratchford, 1991). The retrieval of internal information is low in cognitive effort and hence costs very little relative to external search. Consistent with the information/psychological theoretical stream, consumers are cognitive misers naturally inclined to engage a heuristic approach to external search, effectively substituting internal information for external search (Srinivasan & Ratchford, 1991). For example, it is suggested by researchers that as prior knowledge increases, consumers are more likely to develop routine problem-solving structures that do not require much effort (Fodness & Murray, 1999). Thus, only if internal information is non-retrievable from memory or insufficient to address an impending decision, does a consumer resort to an external search (Gursoy & McCleary, 2003; Ratchford, Talukdar, & Lee, 2001), thereby culminating to a negative relationship between prior knowledge and search effort. It is in this respect that prior credit knowledge is postulated to be negatively related with consumer credit search. Thus, it is hypothesised that:

H2: Consumer credit search is negatively related to prior credit knowledge.

Prior memory structure is defined as the degree of the consumers’ cognitive aptitude in acquiring and processing information (Heaney & Goldsmith, 1999; Punj & Staelin, 1983). Increased prior memory structure facilitates and supports consumers’ extensive information gathering and complex information processing capabilities. Thus, prior memory structure allows consumers to develop more complex cognitive structures of choice decisions and thereby formulate more complex questions that require more information (Gursoy & McCleary, 2003). To the extent that prior memory structure provides an impetus for external search (Punj & Staelin, 1983), it is therefore hypothesised to be positively related with credit search. Thus,

H3: Consumer credit search is positively related to prior memory structure.

The studies using the theory of information processing have determined that consumers have a threshold to the amount of information that they can acquire and efficiently process at any given point in time (Scammon, 1977; Payne, 1976). In particular, information clutter found in consumer markets, compounded with the observed proliferation of products and choices, has been suggested to create data smog. Data smog has been suggested to arouse psychological anxiety and tension, reduced attention span, difficulties in memorising and remembering, and thereby associated with suboptimal financial decisions (Lee & Cho, 2005). Studies found evidence of an association between information oversupply and disjointed and/or dysfunctional rationalisation of situations and hence, suboptimal decisions and choices by consumers (Jacoby, Kohn, & Speller, 1973). With consumer markets subjected to data smog (Biza-Khupe, 2011), it is presupposed that data smog influences consumer credit search. Less obvious though is the relationship between the two variables as this relationship has not been tested a priori. The theoretical streams of consumer information processing and economic analysis offer a lead in this respect. Drawing from the psychological theoretical stream, it could be presupposed that the degree of data smog within consumer credit markets exceeds individuals’ cognitive thresholds, and hence consumers respond by either ignoring or readily discarding information considered superfluous (Payne, 1976). Drawing from this viewpoint, perceived data smog would be postulated to be negatively related with credit search. Under scrutiny, this proposition has a drawback,

Journal of Applied Business and Economics vol. 16(3) 2014 121 which is the (untested) assumption that the data smog levels within the consumer credit market have exceeded consumers’ cognitive threshold. Substantiating this claim would be an arduous task and is beyond the scope of this paper. It is for this reason that this proposition was not further considered. An alternative proposition drawn from the economic theoretical perspective of search is that data smog impedes on the utility maximisation objective of consumers, and hence heightens the perceived risk of attaining suboptimal or undesired credit choices. In view of the fact that individuals are generally regarded as risk averse, heightened perceived risk engenders the adoption of risk mitigating strategies (Schmidt & Spreng, 1996) by, for instance, gathering additional information (Heaney & Goldsmith, 1999; Srinivasan & Ratchford, 1991; Beatty & Smith, 1987). The advantage of this proposition is that the relationship between perceived data smog and perceived risk can be incorporated into the proposed model and explicitly tested. Therefore, data smog is therefore posited to be positively related to perceived risk and credit search. Thus;

H4: Perceived financial risk is positively related to perceived data smog. H5: Consumer credit search is positively related to perceived data smog.

Perceived search cost is defined as the sum total of all subjectively assessed monetary expense; time sacrifice; physical effort; and psychological sacrifice endured during an information acquisition exercise (Kulviwat et al., 2004). In this regard, cost of search are resources that are invested in the normal course of gathering information (Stigler, 1961). It therefore follows that the availability, or lack thereof, of such resources would affect the search activity. In particular, time and financial constraints would curtail consumer credit search. For example, time availability has been found to be positively related to search effort (Beatty & Smith, 1987), while time and income constraints have been found to be negatively related to external search (Weenig & Maarleveld, 2002; Avery, 1996). The net effect of search cost on credit search is hypothesised in this study to depend on its relative positive impact on perceived data smog and perceived financial risk. First, the deficiency in the necessary resources to gather information, thereby resulting in the curtailment of a credit search activity, is expected to exacerbate the perception of data smog. To the extent that time or income constraint, or both, are considered an impediment to consumers’ predisposition to the process of sifting, gathering, organising and analysing credit information from the marketplace. Therefore, and based on the evidence that consumers are generally resource constrained, both financially and temporally (Thogersen, 2005), it is hypothesised that perceived search cost is positively related to perceived data smog. Thus

H6: Consumer perceived data smog is positively related to perceived search cost.

Following from the same argument, a resource constrained consumer is forced to restrict their credit search activity, a condition which effectively impedes on the efficacy of an individual’s risk minimisation strategy (Heaney & Goldsmith, 1999; Punj & Staelin, 1983). To the extent that consumers acquire information to address uncertainties relating to their credit decisions, the lack or limited resource availability to gather such information heightens the risk of committing to a suboptimal decision process and choice. In this regard, the perceived cost of a credit search is hypothesised to be positively related to perceived financial risk. Thus;

H7: Consumer perceived financial risk is positively related to perceived search cost.

Prior memory structure embodies consumer knowledge structure and problem-solving capabilities (Lin & Lee, 2004; Brucks, 1985). To the extent that high prior memory structure proxy higher-order intellectual capabilities, it is postulated that individuals with high prior memory structure are less susceptible to data smog. Further to that, prior memory structure has been found to induce extended information gathering on the part of consumers (Kiel & Layton, 1981), it can reasoned that consumers with a high prior memory structure spend time in credit markets and command high knowledge of the

122 Journal of Applied Business and Economics vol. 16(3) 2014 credit domain. In this respect, such consumers are sharply efficient in gathering and organising the plethora of available financial information and hence less predisposed to data smog. To this end, prior memory structure is hypothesised to be negatively related to data smog. Thus;

H8: Consumer perceived data smog is negatively related to prior memory structure.

Overall, the model proposes that the five variables of financial risk, data smog, prior memory structure and prior credit knowledge have a direct relationship with credit search, whilst the relationship of search cost with credit search is posited to be mediated by perceived financial risk and perceived data smog. The relationship of prior memory structure with credit search is also posited to be mediated by perceived data smog. Another indirect relationship between credit search and data smog, through financial risk, is proposed. The proposed credit search model and the eight hypothesised relationships are presented in diagram 1.

DIAGRAM 1 THE PROPOSED THEORETICAL MODEL OF CONSUMER CREDIT SEARCH

RESEARCH METHODS

A personally-administered questionnaire survey was conducted to empirically test the proposed model of consumer credit search. With a purpose to achieve demographic representation, data was collected from a reasonably large geographic area, covering twelve postal codes of the city of Melbourne, Australia, and targeted a variety of community centres, including schools, social clubs and a mall. Quasi- random sampling was achieved by randomly intercepting potential respondents. The targeted population were individuals who acquired credit to purchase a household appliance for personal use in the past 12 months. A timeframe was imposed on the period after the acquired credit with a purpose to enhance data validity by, for example, minimising on the effect of selective memory and memory decay on the reliability of measures used (Guo, 2001).

Journal of Applied Business and Economics vol. 16(3) 2014 123 The drafting of the questionnaire was informed by the literature in both consumer search behaviour and research methods. Preliminary draft questionnaires were refined through processes that included assessment by doctoral candidates and senior academicians at the School of Commerce and Marketing, Central Queensland University, Australia, and three waves of pilot tests. Through this process, the definition, conceptualisation and operationalisation of key variables were considerably improved. Measures used were predominantly adopted from the extant literature and adapted to the study. The guiding principle espoused the use of multi-item measures to capture the multiple dimensions associated with a variable (construct), thereby increasing scale reliability and validity. Also, to the extent possible, items representing the endogenous and exogenous variables were measured on the recommended 7-point semantic differential scale (Tabachnick & Fidell, 2007; Schumacker & Lomax, 2004), unless if an item could otherwise be better represented and scaled. Credit Search is a multi-dimensional concept which has incorporated the depth of search, breadth of search and patterns of search (Murray, 1991; Beatty & Smith, 1987). In operationalising Credit Search, multi-dimensionality was captured by encapsulated the dual concepts of consumers’ consultation of various financial information intermediaries and the amount of effort expended in procuring information. Other studies have supported and used these two dimensions to represent the depth and breadth of external search (Klein & Ford, 2003; Heaney & Goldsmith, 1999). Perceived Data Smog was adapted from Sproles and Kendall’s (1986) construct of ‘Confusion from Overchoice’ (representing the extent to which a consumer was overwhelmed/confused by information and product choice) and operationalized using the indicator variables suggested and tested by Biza-Khupe (2011). The latent construct of Perceived Data Smog was operationalised as a measure of the degree to which a consumer perceived the proliferation of credit information and products, and the associated degree of confusion. In additional, the dimension of the degree of exposure to financial information by a consumer was considered relevant and hence included in the construct. Perceived Search Cost was operationalised by five items that represented the dual dimensions of perceived monetary and time cost of search, as adapted from previous studies (Moorthy, Ratchford, & Talukdar, 1997; Srinivasan & Ratchford, 1991). The temporal dimension captured time constraint attributable to the urgency of the decision and the general unavailability of time during the credit search process, while the monetary cost dimension captured the perceived costliness of gathering the required amount of information on credit facilities. Combined, these dimensions encapsulated the opportunity cost of time. Perceived Risk has been developed as a construct with the dual dimensions of ‘chance’ and ‘consequence’ of a loss (Srinivasan & Ratchford, 1991; Peter & Ryan, 1976). ‘Chance’ embodies the essence of the probability of a loss occurring or the likelihood of an undesired outcome, while ‘consequence’ encapsulates the essence of the severity of that loss or undesired outcome. Informed by these studies, Perceived Financial Risk was conceptualised as (i) the subjective likelihood of a consumer encountering financial difficulties as a direct result of committing themselves to further debt (thus, the marginal effect of specific credit acquisition on consumers’ overall financial wellbeing) and (ii) the severity of the anticipated financial difficulty. ‘Chance’ was operationalised as the perceived probability of an individual being led into financial difficulties due to high financial commitments and/or low income, while ‘consequence’ was operationalised as the inclination by a respondent to curtail financial commitments and/or increase the income-earning capacity. Cognisant of the importance of cost of credit/debt on consumer credit decisions (Lee & Hogarth, 2000b; Worden & Sullivan, 1987), and by adapting an approach similar to that used by Lee & Hogarth (1999a) in a study of consumer credit knowledge, Prior Credit Knowledge was conceptualised as the level of consumer understanding of the cost of debt. Prior Memory Structure is a complex, multi-dimensional variable operationalised using three dimensions. The first dimension captured the self-assessed adequacy of a consumer’s knowledge of credit facilities (Heaney & Goldsmith, 1999; Punj & Staelin, 1983). The second dimension captured consumer experience with credit (Lee & Hogarth, 2000b; Chang & Hanna, 1992). Thirdly, the highest level of

124 Journal of Applied Business and Economics vol. 16(3) 2014 education of the respondent was used as a measure of the respondent’s ability to learn, develop skills and assimilate relevant information (Kulviwat et al., 2004; Schmidt & Spreng, 1996). In sum, 30 items were used to measure the six variables, as presented in Appendix 1. Also presented is the Cronbach’s coefficient alpha for each variable. The overall results indicated a reliability scale range of 0.64 – 0.86, with one latent constructs below 0.78. Considering that Cronbach’s alpha values greater than or equal to 0.6 are considered acceptable (Robinson, Shaver, & Writghtsman, 1991), with values greater than 0.7 preferred (Hair, Black, Babin, Anderson, & Tatham, 2006), scale reliability was thus confirmed.

THE RESULTS

A total of 600 questionnaires were distributed and completed by respondents, with 245 cases found usable for this research. Descriptive statistics indicated that the average credit amount acquired was AU$1,008. The sample constituted of 56% females and 44% males, with most of the respondents (34.4%) falling in the age-group of 31-40 years. The length of the time span between survey administration and the acquisition of credit by respondents had a mean period of five (5) months. The amount of credit acquired ranged between $50 and $5,000. Details of the sample descriptive statistics are provided in Table 1. TABLE 1 SAMPLE DESCRIPTIVE STATISTICS

Factor Categories Frequency Percent (%) Cumulative % Gender Male 96 43.8 43.8 Female 123 56.2 100 Age Group 20 yrs and below 8 3.3 3.3 21-30 yrs 66 27.4 30.7 31-40yrs 83 34.4 65.1 41-50yrs 56 23.3 88.4 51-60 yrs 22 9.1 97.5 Over 60 yrs 6 2.5 100 Annual Income Less than $13,000 16 6.9 6.9 Before Tax $13,000 - 24,000 35 15.2 22.1 $25,000 - 39,000 56 24.2 46.3 $40,000 - 59,000 69 29.9 76.2 $60,000 - 79,000 40 17.3 93.5 $80,000 - 104,000 9 3.9 97.4 $105,000 or more 6 2.6 100 Highest Level of Secondary school or less 65 27.4 27.4 Education Post-secondary school diploma or certificate 47 19.8 47.2 Trade qualification 51 21.5 68.7 University bachelors degree 49 20.7 89.4 University higher degree 25 10.6 100 Credit amount $500 and less 40 17.4 17.4 $501 - $750 33 14.4 31.8 $751 - $1,000 57 24.7 56.5 $1,001 - $1,500 32 13.9 70.4 $1,501 and more 68 29.6 100 This table shows the demographic descriptive statistics of the sample.

Journal of Applied Business and Economics vol. 16(3) 2014 125 A two-step model-building approach of: (i) assessing the measurement model (CFA) and; (ii) structural equation modelling (SEM), was adopted in this research. The first phase involved testing the integrity of measurement items and the overall model fit (Confirmatory Factor Analysis). In the analysis, each latent construct was related to its predetermined items and allowed to covary with all the other variables (Cheng, 2001). This approach improves the psychometric properties of measures because it takes into consideration the relationship among indicator variables of different constructs (Hair et al., 2006). The results indicated standardised factor loadings ranging between 0.52 and 0.93, which is above the recommended cut-off point of 0.5 (Hair et al., 2006). All items hypothesised to measure a latent construct had statistically significant factor loadings by the critical ratio test (> ± 1.96, p < .05), suggestive of convergent validity. Model respecification was executed by systematically deleting items found not to perform well with regard to model integrity, model fit, or construct validity. In the process, the criterion of reconsidering items that loaded into more than one variable; had high modification indices; had low squared multiple correlations; had low factor loadings; and theoretical reasoning was applied (Cheng, 2001). Also, considering the simultaneous interdependency of the various components of the model, such that the deletion of a single indicator potentially affected other parts of the model simultaneously, model revision was carried out one item at a time. The CFA analysis, through multiple measures of goodness-of-fit, standardised factor loadings and statistically significant factor loadings, supported that the estimated model of the revised CFA model reproduced the sample variance-covariance matrix reasonably well, and thus affirming construct validity in terms of convergent and discriminant validity. The second phase involved specifying hypothesised relationships among variables as posited by theory. Figure 1 shows the structural model that was finally tested.

FIGURE 1 A STRUCTURAL MODEL OF CONSUMER CREDIT SEARCH BEHAVIOUR

This diagram shows the structural model of consumer credit search that was tested using the Amos program.

126 Journal of Applied Business and Economics vol. 16(3) 2014 Model evaluation was ascertained using the multiple indexes of the absolute goodness-of-fit and incremental fit criterions, as presented in Table 2.

TABLE 2 GOODNESS-OF-FIT FOR THE MODELS

CMIN/DF RMSEA GFI TLI IFI CFI Model Revised CFA Model 2.62 .081 .82 .84 .86 .86 Structural Model 2.98 .09 .82 .83 .85 .85 Targeted Benchmark* < 3.0 .05 - .10 > .80 > .80 > .80 > .80 *(Ho, 2006; Chau, 1997; McCullum, Browne, & Sugawara, 1996; Hoyle, 1995; Hair, Anderson, Tatham, & Black, 1995; Browne & Cudeck, 1993) This table shows the results of the goodness-of-fit results for both the Confirmatory Factor Analysis and the Structural Equation model. The literature used for benchmarking is also indicated*.

From, the results, the measure of CMIN/DF (2.98) was within the targeted benchmark score of less than 3.0; RMSEA (0.09) was within the targeted range of 0.05 – 0.10; and GFI (0.82) was above the targeted minimum benchmark of 0.8. The incremental fit indices of TLI (0.83), IFI (0.85) and CFI (0.85) surpassed the targeted minimum benchmark value of 0.80. These indices compared the fit of the hypothesised model to the null or independence model. Given the range of the incremental fit indices, the possible improvement in the fit for the hypothesised model (range: 0.15 - 0.18) appears so small as to be of no practical significance (Ho, 2006). In sum, the results of the absolute and incremental fit indexes were indicative that the structural model was a reasonable fit. The structural model of credit search indicated an R2 of 0.69 for Credit Search, in which seven out of the eight hypothesised relationships were statistically significant by the critical ratio test (> ± 1.96, p < .05), and theoretically meaningful. A summary of the test outcome for each hypothesised relationship is presented in table 3.

TABLE 3 CREDIT SEARCH SEM HYPOTHESES TEST RESULTS

Hypothesis Path Path Critical Hypothesis Coefficient Ratio Test Results H1 Perceived financial risk to credit search 0.17 2.79** Supported H2 Prior credit knowledge to credit search -0.62 -4.37** Supported H3 Prior memory structure to credit search 0.51 5.45** Supported H4 Perceived data smog to perceived financial 0.14 1.64 Not supported risk H5 Perceived data smog to credit search 0.19 2.84** Supported H6 Perceived search cost to perceived data 0.60 5.56** Supported smog H7 Perceived search cost to perceived 0.32 3.95** Supported financial risk H8 Prior memory structure to perceived data -0.17 -2.83** Supported smog This table shows the results of the hypothesis tests derived from the empirical structural equation model. Note that ** denotes significance at p ≤ .01 level

Journal of Applied Business and Economics vol. 16(3) 2014 127 DISCUSSIONS

The paper proposed and tested, simultaneously, (i) the relationship between Credit Search and its direct antecedents, (ii) interrelationships among Credit Search antecedents and (iii) the relationship between Data Smog and its direct antecedents, in a single comprehensive model. From the results, all eight, but one (H4), of the hypothesised relationships were supported. The hypothesised relationship between Financial Risk and Data Smog had a path coefficient of 0.14 and a critical ratio value of 1.64 that was not statistically significant at the 0.05 level. The findings suggested that the degree of perceived data smog within consumer credit markets did not heighten the perceived financial risk of credit acquisition, although respondents found some weak positive association between the two factors. It can be surmised that although the cost-benefit paradigm has associated consumer uncertainty of a product domain with the perceived risk of a purchase decision (Mitra et al., 1999; Schmidt & Spreng, 1996), perceived data smog did not seem to generate sufficiently high levels of uncertainty within the consumer credit market to affect the perceived financial risk of a credit decision. H1 stated that consumer credit search was positively related to perceived financial risk. The findings indicated that consumers searched for more information when they perceived the acquisition of additional credit as posing a risk to their financial wellbeing. The findings of the study are congruent with the results of previous studies in consumer search for goods (Beatty & Smith, 1987; Locander & Hermann, 1979). The only other study found which tested a structural model of consumer search for services did not find a significant relationship between perceived risk and search effort (Heaney & Goldsmith, 1999), and concluded that consumers do not view banking services as highly variable and hence the low perceived risk of making a bad decision. The findings of this study seemed to refute this view, at least in as far as it relates to credit. The findings of this study suggested that borrowers mitigated the perceived financial risk of acquiring credit through an elaborate information search strategy. H2 stated that consumer credit search was negatively related to prior credit knowledge. The findings suggested a substitutive effect between internal information and external search, consistent with the information processing theory of consumer search (Gursoy & McCleary, 2003; Ratchford et al., 2001). The findings were indicative that consumers are cognitive misers who resort to external search when the required information was not previously acquired or unable to be retrieved from memory (Schmidt & Spreng, 1996). H3 stated that consumer credit search was positively related to prior memory structure. The findings suggested that consumers engaged in more search for credit when cognitive structures and knowledge structure, and thus problem-solving capabilities, were high, consistent with the psychological paradigm of consumer search (Klein & Ford, 2003). Overall, the findings supported the theoretical perspective that prior memory structure equips consumers with improved cognitive abilities that allow them to: structure the problem in a more complex manner; gather information more efficiently; process more information; and therefore incur lower search costs; and thereby intensify the external information search activity (Lin & Lee, 2004). H5 stated that consumer credit search was positively related to perceived data smog. The findings suggested that the respondents considered the credit markets to be congested with information and product choices (Lee & Cho, 2005; Lee & Hogarth, 1999b) such that they had to sift through the plethora of data to finally get the relevant information required for the impending credit decision. The concept advocated by the cost-benefit approach of an average consumer being economically rational, goal- directed and a calculated economic entity motivated to gather information (Kim & King, 2009; Stigler, 1961) was vivified in this study. H6 stated that consumer perceived data smog was positively related to perceived search cost. The findings indicated that perceived search cost, which was conceptualised as the curtailment of an investment outlay required in the normal course of gathering information, constrained the respondents’ predisposition to information on credit and search. In this regard, a resource-constrained individual forced by the circumstances to limit their search activity would therefore have had limited prior exposure to the domain of credit markets and hence the increased perception of data smog. The findings provided

128 Journal of Applied Business and Economics vol. 16(3) 2014 supporting evidence to the proposition that the perceived search cost is an antecedent to perceived data smog. Further, the hypothesised indirect association between perceived search cost and credit search, through perceived data smog, was confirmed. H7 stated that consumer perceived financial risk was positively related to perceived search cost. The findings suggested that consumers who were resource constrained (temporal and monetary) had a high propensity to limit their credit search activity, in accordance with the cost-benefit paradigm (Stigler, 1961). Information search is generally regarded as a risk reduction strategy among consumers (Gursoy & McCleary, 2003; Mitra et al., 1999). Therefore, resource-constrained respondents could have had limited ‘internally banked’ information on credit because of the a priori deprived information gathering background. These consumers could have felt ill-equipped to competently address their credit-decision uncertainties, hence the heightened financial risk. Further, the hypothesised indirect association between perceived search cost and credit search, through perceived financial risk, was confirmed. Finally, H8 stated that consumer perceived data smog was negatively related to prior memory structure. The findings suggested that consumers with increasingly advanced knowledge structure and cognitive analytical processes perceived the consumer credit markets as less data-congested and overwhelming. These findings are consistent with the consumer information processing theory which suggests that increased prior memory structure facilitates processing of complex information and efficient information processing, and motivates an extended search (Heaney & Goldsmith, 1999; Srinivasan & Ratchford, 1991), and are congruent with findings of H6.

STUDY IMPLICATIONS AND FUTURE RESEARCH

The paper contributes to theory and practice. First, although services are characteristically distinct from goods, findings of this study suggest that when interrelations among the endogenous and exogenous variables of search are simultaneously tested, they not only provide a better representation of the complexities of consumer behaviour, but also consumer search strategies across product are comparable. In this regard, higher order generic theory of consumer search behaviour which integrates, rather than segregates, the two product classes is suggested as the direction for future research. Second, the study found Data Smog to be a significant role player in consumer credit markets. From these findings, consumers invest in sifting through the masses of data to gather the required information, with the sifting process itself being neither productive nor value-adding in terms of reducing financial risk. From a practical perspective, these findings are suggestive that the provision of ‘perfect’ information through financial information regulation has generated data smog, and that data smog is a liability and detrimental in consumer decision-making processes (Kozup & Hogarth, 2008). An eminent challenge is therefore to reengineer financial information regulation towards providing adequate, comprehendible and comprehensive regulated information in a manner and form that does not congest the market. This paper was not without limitations. There were some items used to measure latent constructs that were found inapt, and hence deleted. Further research will need to be undertaken to determine whether alternative items could have improved the result outcome or new measures need to be developed. Also, there are several other potential relationships between variables that could have been hypothesised in the proposed model of consumer credit search. For example, previous studies in the literature have proposed direct relationship between Perceived Search Cost and Search, and such other interrelationships among search determinant variables as Prior Knowledge and Perceived Risk (Heaney & Goldsmith, 1999; Srinivasan & Ratchford, 1991; Punj & Staelin, 1983). Further, the development of future models of consumer search should expand in comprehensiveness to incorporate other search determinants that were not considered in this paper. Finally, the hypothesised relationship between Financial Risk and Data Smog was not supported. A conceivable explanation for these equivocal results is the role/effect of other factors not incorporated into the model. Clearly, further research is warranted.

Journal of Applied Business and Economics vol. 16(3) 2014 129 APPENDIX 1 OPERATIONALISATION OF KEY VARIABLES

Construct and Item/ Question (abridged version) Cronbach’s alpha Label Perceived Data Conf1 I find comparing credit facilities a confusing task. Smoga Conf2 The more I try to learn about credit facilities, the more confused I seem to get. .86 (3)* Conf3 There are too many different types of credit facilities to choose from. InfAv There is a lot of information available on credit facilities. Acess For me, sources of information on credit facilities are within easy reach.

Perceived Search Time I felt it took a lot of time to search for information on credit facilities. Costb Urgent I had to make a quick purchase decision. .78 (5) NoTime I had little time to search for information on credit facilities. Costy It is costly to obtain information about credit facilities. Mone I need to spend some money in order to obtain adequate information on credit.

Perceived DifIns Acquisition of the credit may lead me to financial difficulties because of the Financial Riskc high weekly instalment. .80 (6) DifPay Acquisition of the credit may lead me to financial difficulties because of my low weekly pay. DifExp Acquisition of the credit may lead me to financial difficulties because of my high weekly expenses. RedCr How important is cutting back on your credit charges? IncrIncm How important is getting additional income to pay for your credit instalments? RedExp How important is reducing your weekly expenses to be able to pay-off your credit?

Prior Memory Know1 I know a lot about credit facilities compared to most of the people I know. Structured Know2 If I had to get a loan today, I would need to gather very little information in .64 (3) order to make a wise decision. Educ Highest level of education attained. CreditExp Credit experience - the number of times a respondent had applied for different loans before and the number of credit cards held.

Credit Search InfSech I searched for a lot of information before using the credit facility. .85 (8)e InfTime Before using the credit facility, I took a lot of time considering my options. ConPS Degree of consultation of friend/ relative/ colleague. ConIND Degree of consultation of independent financial consultant. ConRpt Degree of consultation of consumer report/ article/ guide. ConAD Degree of consultation of commercial advert on radio/ TV/ newspaper/ magazine/ internet. ConBK Degree of consultation of credit provider/ bank. ConStr Degree of consultation of dealer/ salesperson. InfNo I could not be bothered to look for any information before using the credit.

Prior Credit ObjKnw Correctly ranking four instruments of consumer credit in the order of cost Knowledgef using the interest rate charge. This table is a list of item that were used to proxy the six variables used in the proposed and tested structural equation model of consumer credit search. *is the number of items used to represent each variable in subsequent analysis, after reconsiderations informed both the reliability analysis. The items in italics were deleted due to low corrected item-total correlation and whose deletion resulted in a substantial improvement in the overall Cronbach’s coefficient

130 Journal of Applied Business and Economics vol. 16(3) 2014 REFERENCES

Avery, R. J. (1996). Determinants of Search for Nondurable Goods: An Empirical Assessment of the Economics of Information Theory. The Journal of Consumer Affairs, 30(2), 390-420. Beatty, S. E., & Smith, S. M. (1987). External Search Effort: An Investigation Across Several Product Categories. Journal of Consumer Research, 14(1), 83-95. Begg, I. (2009). Regulation and Supervision of Financial Intermediaries in the EU: The Aftermath of the Financial Crisis. Journal of Common Market Studies, 47(5), 1107-1128. Bernthal, M. J., Crockett, D., & Rose, R. L. (2005). Credit Cards as Lifestyle Facilitators. Journal of Consumer Research, 32(1), 130-145. Bettman, J. R. (1979). An Information Processing Theory of Consumer Choice. Reading MA: Addison- Wesley Publishing Company. Bettman, J. R., Johnson, E., & Payne, J. (1991). Consumer Decision Making. In T. S. Robertson & H. H. Kassarjian (Eds.), Handbook of Consumer Research (pp. 50-84). Englewoods Cliffs, NJ: Prentice Hall. Biza-Khupe, S. (2011). Information Regulation in Consumer Credit Markets – Determinants of Financial Information Oversupply. International Research Journal of Finance and Economics, 71 (August), 19-33. Browne, M. W., & Cudeck, R. (1993). Alternative Ways of Assessing Model Fit. In K. A. Bollen & J. S. Long (Eds.), Testing Structural Equation Models (pp. 136-162). Newbury Park, California: Sage. Brucks, M. (1985). The Effects of Product Class Knowledge on Information Search Behaviour. Journal of Consumer Research, 12(1), 1-15. Buch, J., Rhoda, K. L., & Talaga, J. (2002). The Usefulness of the APR for Mortgage Marketing in the USA and the UK. International Journal of Bank Marketing, 20(2/3), 76-85. Calem, P. S., & Mester, L. J. (1995). Consumer Behavior and the Stickiness of Credit-Card Interest Rates. The American Economic Review, 85(5), 1327-1336. Chang, R. Y., & Hanna, S. (1992). Consumer Credit Search Behaviour. Journal of Consumer Studies and Home Economics, 16, 207-227. Chau, P. Y. K. (1997). Reexamining a Model for Evaluating Information Center Success Using a Structural Equation Modeling Approach. Decision Sciences, 28(2), 309-334. Cheng, E. W. L. (2001). SEM Being More Effective Than Multiple Regression in Parsimonious Model Testing for Management Development Research. Journal of Management Development, 20(7/8), 650-667. Datta, P., & Chatterjee, S. (2008). The Economics and Psychology of Consumer Trust in Intermediaries in Electronic Markets: The EM-Trust Framework. European Journal of Information Systems, 17, 12-28. Ferretti, F. (2007). Consumer Credit Information Systems: A Critical Review of the Literature. Too Little Attention Paid by Lawyers? European Journal of Law and Economics, 23, 71-88. Fodness, D., & Murray, B. (1999). A Model of Tourist Information Search Behavior. Journal of Travel Research, 37(3), 220-230. Fox, J., Bartholomae, S., & Lee, J. (2005). Building the Case for Financial Education. The Journal of Consumer Affairs, 39(1), 195-214. Guo, C. (2001). A Review on Consumer External Search: Amount and Determinants. Journal of Business Psychology, 15(3), 505-519. Gursoy, D., & McCleary, K. W. (2003). An Integrative Model of Tourists' Information Search Behavior. Annals of Tourism Research, 31(2), 353-373. Hair, J. F., Anderson, R. E., Tatham, R. L., & Black, W. C. (1995). Multivariate Data Analysis with Readings. Englewood Cliffs, New Jersey: Prentice-Hall. Hair, J. F., Black, W. C., Babin, B. J., Anderson, R. E., & Tatham, R. L. (2006). Multivariate Data Analysis (6 ed.). New Jersey: Pearson Prentice Hall.

Journal of Applied Business and Economics vol. 16(3) 2014 131 Heaney, J., & Goldsmith, R. E. (1999). External Information Search for Banking Services. International Journal of Bank Marketing, 17(7), 305-323. Hiu, A. S. Y., Siu, N. Y. M., Wang, C. C. L., & Chang, L. M. K. (2001). An Investigation of Decision- Making Styles of Consumers in China,. The Journal of Consumer Affairs, 35(2), 326-345. Ho, R. (2006). Handbook of Univariate and Multivariate Data Analysis and Interpretation with SPSS. Florida: Chapman and Hall. Hoyle, R. H. (1995). The Structural Equation Modeling Approach: Basic Concepts and Fundamental Issues In R. H. Hoyle (Ed.), Structural Equation Modeling: Concepts, Issues and Applications (pp. 1-15). California: Sage Publications. Jacoby, J., Kohn, C. A., & Speller, D. E. (1973). Time Spent Acquiring Product Information as a Function of Information Overload and Organisation. Paper presented at the American Psychology Association's 81st Annual Convention. Kay, J. (2009). The Long and the Short of it: A Guide to Finance and Investment for Normally Intelligent People Who Aren’t in the Industry. London: Erasmus Press. Kerr, S., & Dunn, L. (2008). Consumer Search Behavior in the Changing Credit Card Market. Journal of Business and Economic Statistics, 26(3), 345-353. Kiel, G. C., & Layton, R. A. (1981). Dimensions of Consumer Information Seeking Behaviour. Journal of Marketing Research, 18(2), 233-239. Kim, W. J., & King, K. W. (2009). Product Category Effects on External Search for Prescription and Nonprescription Drugs. Journal of Advertising, 38(1), 5-19. Klein, L. R., & Ford, G. T. (2003). Consumer Search for Information in the Digital Age: An Empirical Study of Prepurchase Search for Automobiles. Journal of Interactive Marketing, 17(3), 29-49. Kozup, J., & Hogarth, J. M. (2008). Financial Literacy, Public Policy, and Consumers' Self-Protection - More Questions, Fewer Answers. The Journal of Consumer Affairs, 42(2), 127-136. Kozup, J., Howlett, E., & Pagano, M. (2008). The Effects of Summary Information on Consumer Perceptions of Mutual Fund Characteristics The Journal of Consumer Affairs, 42(1), 37-59. Kulviwat, S., Guo, C., & Engchanil, N. (2004). Determinants of Online Information Search: A Critical Review and Assessment. Internet Research, 14(3), 245-253. Lee, J., & Cho, J. (2005). Consumers’ Use of Information Intermediaries and the Impact on their Information Search Behavior in the Financial Market. The Journal of Consumer Affairs, 39(1), 95-120. Lee, J., & Hogarth, J. M. (1999a). The Price of Money: Consumers’ Understanding of APRs and Contract Interest Rates. Journal of Public Policy and Marketing, 18(1), 66-76. Lee, J., & Hogarth, J. M. (1999b). Returns to Information Search: Consumer Mortgage Shopping Decisions. Financial Counseling and Planning, 10(1), 49-66. Lee, J., & Hogarth, J. M. (2000a). Consumer Information Search for Home Mortgages: who, what, how much, and what else? Financial Services Review, 9(3), 277-293. Lee, J., & Hogarth, J. M. (2000b). Relationships Among Information Search Activities When Shopping for a Credit Card. The Journal of Consumer Affairs, 34(2), 330-360. Lin, Q. C., & Lee, J. (2004). Consumer Information Search When Making Investment Decisions. Financial Services Review, 13(4), 319-332. Locander, W. B., & Hermann, P. W. (1979). The Effect of Self-Confidence and Anxiety of Information Seeking in Consumer Risk Reduction. Journal of Marketing Research, 16(2), 268-274. Malbon, J. (2001). Shopping for Credit: An Empirical Study of Consumer Decision-Making. Australian Business Law Review, 29(1), 44-64. Mattila, A. S., & Wirtz, J. (2002). The Impact of Knowledge Types on the Consumer Search Process: An Investigation in the Context of Credence Services. International Journal of Service Industry Management, 13(3/4), 214-230. McColl-Kennedy, J. R., & Fetter, R. E. (1999). Dimensions of Consumer Search Behavior in Services. The Journal of Services Marketing, 13(3), 242-265.

132 Journal of Applied Business and Economics vol. 16(3) 2014 McCullum, R. C., Browne, M. W., & Sugawara, H. M. (1996). Power analysis and determination of sample size for covariance structure modeling Psychological Methods, 1, 130-149. Mitra, K., Reiss, M. C., & Capella, L. M. (1999). An Examination of Perceived Risk, Information Search and Behavioral Intentions in Search, Experience and Credence Services. The Journal of Services Marketing, 13(3), 208-228. Moorthy, S., Ratchford, B. T., & Talukdar, D. (1997). Consumer Information Search Revisited: Theory and Empirical Analysis. Journal of Consumer Research, 23(4), 263-277. Murray, K. B. (1991). A Test of Services Marketing Theory: Consumer Information Acquisition Activities. Journal of Marketing, 55(1), 10-25. Murray, K. B., & Schlacter, J. L. (1990). The Impact of Services versus Goods on Consumers' Assessment of Perceived Risk and Variability. Journal of the Academy of Marketing Science, 18(1), 51-65. OECD. (1992). Consumer Information About Financial Services. Paris: Organisation for Economic Co- operation and Development. Payne, J. (1976). Task Complexity and Contingent Processing in Decision-Making: An Information Search and Protocol Analysis. Organizational Behavior and Human Performance, 16(2), 366- 387. Peter, J. P., & Ryan, M. J. (1976). An Investigation of Perceived Risk at the Brand Level. Journal of Marketing Research, 13(2), 184-188. Punj, G. N., & Staelin, R. (1983). A Model of Consumer Information Search Behavior for New Automobiles. Journal of Consumer Research, 9(4), 366-380. Ratchford, B. T., Talukdar, D., & Lee, M.-S. (2001). A Model of Consumer Choice of the Internet as an Information Source. International Journal of Electronic Commerce, 5(3), 7-21. Rischkowsky, F., & Doring, T. (2008). Consumer Policy in a Market Economy: Considerations from the Perspective of the Economics of Information, the New Institutional Economics as well as Behavioural Economics. Journal of Consumer Policy, 31(3), 285-313. Robinson, J. P., Shaver, P. R., & Writghtsman, L. S. (1991). Criteria for Scale Selection and Evaluation. In J. P. Robinson, P. R. Shaver & L. S. Writghtsman (Eds.), Measures of Personality and Social Psychological Attitudes. San Diago, CA: Academic Press. Scammon, D. L. (1977). 'Information Load' and Consumers. Journal of Consumer Research, 4(3), 148- 156. Schmidt, J. B., & Spreng, R. A. (1996). A Proposed Model of External Consumer Information Search. Journal of the Academy of Marketing Science, 24(3), 246-256. Schumacker, R. E., & Lomax, R. G. (2004). The Beginner's Guide to Structural Equation Modeling (2nd ed.). New Jersey: Lawrence Erlbaum Associates, Inc. Sproles, G., & Kendall, E. L. (1986). A Methodology for Profiling Consumers' Decision-Making Styles. The Journal of Consumer Affairs, 20(2), 267-279. Srinivasan, N. (1990). Pre-Purchase External Search for Information. In E. V. E. Zeithaml (Ed.), Review of Marketing (pp. 153-189). Chicago: American Marketing Association. Srinivasan, N., & Ratchford, B. T. (1991). An Empirical Test of a Model of External Search for Automobiles. Journal of Consumer Research, 18(2), 233-242. Stigler, G. J. (1961). The Economics of Information. The Journal of Political Economy, 69(3), 213-225. Tabachnick, B. G., & Fidell, L. S. (2007). Using Multivariate Statistics (5 ed.). Boston: Pearson Education. Thogersen, J. (2005). How May Consumer Policy Empower Consumers for Sustainable Lifestyles? Journal of Consumer Policy, 28, 143-178. Waddington, P. (1997). Dying for Information? A Report on the Effects of Information Overload in the UK and Wordwide. Beyond the Beginning: The Global Digital Library Retrieved 23/05/2006, 2006, from http://www.cni.org/regconfs/1997/ukoln-content/repor~13.html Walsh, G., Mitchell, V.-W., & Hennig-Tharau, T. (2001). German Consumer Decision-Making Styles. The Journal of Consumer Affairs, 35(1), 73-95.

Journal of Applied Business and Economics vol. 16(3) 2014 133 Weenig, M. W. H., & Maarleveld, M. (2002). The Impact of Time Constraint on Information Search Strategies In Complex Choice Tasks. Journal of Economic Psychology, 23(6), 689-702. Worden, D. D., & Sullivan, C. A. (1987). Shopping for Consumer Credit: Implications for Market Efficiency Unpublished manuscript. Zeithaml, V. A., Parasuraman, A., & Berry, L. L. (1985). Problems and Strategies in Services Marketing. Journal of Marketing, 49(2), 33-46.

134 Journal of Applied Business and Economics vol. 16(3) 2014

Industrial Clustering Approach in Regional Development: The Case of Turkey

Mustafa Cem Kirankabeş Balıkesir University, Balıkesir, Turkey

Murat Arik Middle Tennessee State University

INTRODUCTION

In 2006, the Public Law 5449 laid out the foundation for establishing 26 Economic Development Agencies across Turkey. The primary goal of this initiative was to promote a competitive economic environment across the regions through public-private partnerships under the leadership of a regional administrator appointed by the central government. Since then, these agencies have initiated a notable number of economic development projects and commissioned studies to uncover the regional economic dynamics.1 What was the underlying approach to economic development behind all these efforts? Around the same time that Public Law 5449 was enacted, “(White Book)-Beyaz Kitap” with a subtitle of “The Project on Developing a Clustering Policy in Turkey” was published with the support of both government research institutes and major chambers of commerce, calling for cluster-based regional economic development policies in Turkey (Beyaz Kitap, 2007). Of course, these concepts of regional economic development and industrial clustering are not new in economic development literature. Gaining wide- recognition with Porter (1990) and his subsequent efforts through the Harvard-based “Institute for Strategy and Competitiveness,” the cluster-based regional economic development policies have become primary tools across many countries including the European Union. Given the increasing emphasis on the cluster-based policies by the European Union leaders, these regional initiatives in Turkey may be considered as a targeted attempt to harmonize regional economic development policies with the European Union policies. The goal of this study is to assess (a) how these Turkish regional economic development policies contributed to cluster development in the 24 manufacturing sectors across 26 newly-established regions,

Journal of Applied Business and Economics vol. 16(3) 2014 135 and (b) whether the extent of cluster formation is related to economic openness, economic development level, and public incentives for investment in these 26 economic development districts. The rest of this paper is organized as follows: First, we provide a background on the clustering approach in regional economic development policies. Second, we offer a detailed discussion about the current practices and issues in literature regarding the industrial cluster policies, as well as the Porter’s (1990) diamond approach. Third, we present the method of inquiry with a conceptual framework, hypotheses, data, and measurement issues. A final discussion and conclusion will follow.

BACKGROUND ON INDUSTRIAL CLUSTERING APPROACH

Although it emerged as a discipline after World War II, regional economy or spatial economics had its roots in the early 19th century with the work of German economist Johann Heinrich von Thuehen (Nijkamp & Mills, 1986, p.1). The Great Depression in the 1930s however, had a profound impact on the macroeconomic dynamics around the world. J. M. Keynes (1936) argued that government intervention, not the free market mechanism, is the way to get out of this depression, ushering a period that has influenced regional economics literature. The real life embodiment of this line of thinking was a success story in the Tennessee Valley region, initiated by the Federal Government in the United States between 1930 and 1940.2 Adapted and expanded by the neoclassical growth models in the 1950s and 1960s, the Keynesian models of regional growth eventually gave way to such concepts as “industrial zones” and “industrial clusters” through the extensive works of F. Perroux, G. Mrydal, and A. Hirschman (Tuyluoglu & Karakas, 2006, pp.197-198) In 1990, Michael Porter advanced the concept of cluster-based regional economic development by investigating the elements of regional competitiveness at the international level. Porter’s study concludes that competitive sectors of countries tend to have cluster formations. Porter's work has added a new vision to economic development and competitiveness approaches by underlining the fact that sustainable competitive advantage depends on a unique mix of internal and external resources in the general business environment. Considered as a new approach or a new way of thinking, clustering has brought a new dimension to industrial policy, regional development policies, innovation, and small and medium enterprises (SME) policies (Quandt & Pacheco, 2000). Since the Porter's diamond model, elaborated in his 1990 study, the literature that has at least cited this model has grown unabated over the years (Figure 1).

FIGURE 1 NUMBER OF JOURNAL ARTICLES INCLUDING WORDS "PORTER" AND CLUSTER

1000 900 800 Number of journal articles containing 700 the words "Porter" and "Cluster" 600 between 1990 and 2012 500 400 300 Number of articles of Number 200 100 0

Source: ABI/INFORM Complete Database

136 Journal of Applied Business and Economics vol. 16(3) 2014 A review of some recent works suggests why the clustering approach to regional development has not lost its appeal over the years. For example, giant clusters, which make a major contribution in terms of regional development, are widely observed in the United States. The notable example is the California's Silicon Valley. Klepper (2010) provides a good history of how Silicon Valley and Detroit have transformed their respective cities in the last 30 years through the semiconductor and automotive industries, respectively (Klepper, 2010, p.15). In addition to broader regional development tool, the clustering approach is often seen as, first, an effective mechanism to promote the emergence of new business opportunities for the small and medium enterprises in the cluster (Quandt & Pacheco, 2000, p.1). Second, the firms in industrial clusters benefit from knowledge and technology transfer. Third, the firms located within a cluster may have access to qualified and/or trained personnel through educational institutions and institutions associated with the labor market in the cluster. Finally, cluster network provides the firms with the opportunity to access specialized suppliers (Isbasoiu, 2007). Businesses located within an industrial cluster will reduce their transaction costs by using local suppliers instead of procuring intermediate goods from a remote supplier. From a marketing perspective, the clustering has two major benefits: on one hand, the enterprises in an industrial cluster will have recognition and prestige, an efficient distribution network, and the opportunity to penetrate deeply into the market. This will in turn increase the number of customers and revenues for the firms not only in the local markets but also in the international markets. On the other hand, the cluster itself is also an important internal market. The clustering approach provides significant opportunities to access facilities and services offered by public institutions to the enterprises which are located within the cluster network. Physical infrastructure provided by the public sector, institutions such as research institutes and testing laboratories, and services such as education are offered to enterprises located within the cluster. All of these facilities increase the success and the competitiveness of enterprises located in industrial clusters. Because of these potential benefits, industrial clustering approach has still been one of the favorite tools used by scholars, professionals and political leaders for regional development and the economic competitiveness.

LITERATURE REVIEW

The industrial clustering approach is mainly based on the brief references related to "industrial districts" proposed by Alfred Marshall in his book "The Principles of Economy" which was published for the first time in 1890. Marshall (1890) refers to three conditions for the formation of industrial clusters. These are the existence of a pool of adequate labor, the presence of specialized suppliers and the possibility of external spill-overs (the rapid transfer of know-how and ideas inside the cluster). Walter Isard (1960) developed the concept of industrial clusters by adding export-oriented industries and links of these industries to other industries in the region. According to Isard, these strong industrial links are a proof of the existence of industrial clusters (Isbasoiu, 2007, p.3). Industrial clusters are defined by many academics as industrial groups concentrated in a particular geographic area, connected to each other both vertically (having relationships with suppliers and customers) and horizontally (sharing common resources such as technology and human resources) (Porter, 1990; Feser & Bergman, 2000; Feser, 2005). Today, many articles on clustering approach are published in scientific journals related to the economy, growth, and regional development (Maskell, P. & Kebir, L., 2006, p.30; see Figure 1 for the general trend). In recent years, there has been a surge in the number of the clustering studies that deals with regional and international competitiveness created by the clusters. For example, ABI/INFORM Complete Search using the “cluster” and “competitiveness” keyword combinations produces 2,454 articles between 1990 and 1999; 8,641 articles between 2000 and 2009; and 4,195 records between 2010 and 2014. As highlighted previously, the pioneering work of Michael Porter (1990) has contributed to this surge. Because of its significance, this section further elaborates Porter’s approach to industrial clustering and economic competitiveness. According to Porter (1990), a cluster as a whole is greater than the sum of its constituent parts. The Porter’s diamond model proposed in this study explains why some societies have a

Journal of Applied Business and Economics vol. 16(3) 2014 137 comparative advantage in certain industries. The diamond model highlights the conditions necessary for a region to achieve an international competitive advantage using four inter-related elements (Porter, 1990, pp.72-74). These are (1) factor conditions, (2) demand conditions, (3) firm strategy and competition, and (4) related and supportive industries (Figure 2).

According to Porter (1990), favorable factor conditions (Figure 2) are necessary to achieve an international competitive advantage. These factor conditions fall under two main categories: simple and advanced. While natural resources, climate, physical infrastructure, unskilled and semi-skilled labor, and financial capital are grouped under the simple category, modern digital data communications infrastructure, highly trained administrative staff such as engineers and computer scientists, and research institutions are cited under the advanced factor conditions. Nowadays, especially advanced factor conditions are very important in achieving international competitive advantage for the firms. To briefly highlight other elements, Porter considers the structure of the household demand for products and services of industries, demand conditions, as an important element for a competitive regional economy. The third element in his model includes related and supporting industries, representing the presence or absence of industries producing internationally competitive products and the first-tier suppliers of these industries in a nation or region. The last element is firm strategy, structure, and rivalry, which are related to how businesses are set up, how they are organized, the structure of local competition, and the level of competition in a given geographical area (Porter, 1990, pp.74-77). An important part of the cluster model developed by Porter is that governments play an important but indirect role in creating internationally competitive sectors. According to Porter, the government should not try to create a competitive advantage on its own, but should indirectly contribute to the creation of a competitive environment by supporting the four main elements in the model. Porter suggests that governments may positively or negatively affect (or may be affected by) the four main elements in the diamond model. In addition to government, there are also the chance factors in the Diamond Model,

138 Journal of Applied Business and Economics vol. 16(3) 2014 described as events beyond anyone’s control but whose occurrence may adversely affect sectors and change their relative positions in the competitive environment (Porter, 1990, p.127). A major research question in the cluster-related papers is about the differential growth rates across the regions (Porter, 1990, 1998a; Saxenian, 1996; Barro & Sala-i-Martin, 1995). Within this context, some studies emphasize the relationship between cluster formation and entrepreneurship. For example, Delgado, Porter & Stern (2010) investigated the impact of industrial clustering or agglomeration at regional industry level on entrepreneurship. Using a novel panel dataset from the Longitudinal Business Database of the Census Bureau and the U.S. Cluster Mapping Project, this study focused on the distinct influences of convergence and agglomeration on the number of start-up firms as well as employment growth in these new firms in a given region The major findings of this study are that (1) clusters have positive impact on entrepreneurship; (2) strong clusters mean more start-ups and employment growth; (3) the strong industrial clusters lead to the expansion of existing companies; and (4) the strong clusters mean a high survival rate for the start-up firms (Delgado, Porter & Stern, 2010, p.2). A somewhat related study looks at the relationship of how clustering is connected with entrepreneurship and economic growth. Wennberg and Lindqvist (2008), for example, compared the performance and chances of survival of newly established enterprises located in industrial clusters with those that are not located in industrial clusters. Using a rigorous methodology, the study concludes that (1) companies located within the cluster were more fortunate in terms of survival; and (2) enterprises located within the cluster created more employment, paid higher wages to their employees, and paid higher taxes as well (Wennberg & Lindqvist 2008, p.2). Today, the clustering approach is regarded as an important program in regional development, sustainable growth, and increasing competitiveness. Because of its significance, it is not hard to find many studies commissioned by developed countries on industrial clustering. For example, European Cluster Policy Group (ECPG) was established in 2008 by a decision of the European Commission aiming to carry out studies to increase coordination and quality. The European Cluster Policy Group (ECPG) is composed of 20 independent experts. ECPG is assigned the task of "producing suggestions on how to better design the clustering policies within the EU."3 The studies conducted by this group demonstrate that companies located within a cluster reach higher levels of productivity and innovation, and clusters provide higher chances of survival and higher growth rates for newly established firms suggesting the continuation of clustering programs across the region. In addition to academic and country specific clustering analyses, it is worth mentioning several institutions and programs that have emerged as a major force in cluster research. Among these programs and institutions, the U.S. Cluster Mapping Project4 carried out by Institute for Strategy and Competitiveness, 5 Harvard University, under the leadership of Michael E. Porter is a major project funded by the U.S. Department of Commerce and the Economic Development Administration. The U.S. Cluster Mapping website created within the scope of the project offers data for researchers and policymakers across the United States allowing them to analyze regional clusters. The European equivalent of this institution is the Center for Strategy and Competitiveness6 located within Stockholm School of Economics and funded by DG Enterprise and Industry of the European Commission. This Center is also the coordinator of the European Cluster Observatory7 and Europe INNOVA Cluster Mapping Project.8 In Turkey, a major source on clustering studies is the so-called “White Book.” The White Book of "The Project on Developing a Clustering Policy in Turkey" is prepared with a participatory approach by the direct and active participation of sixteen stakeholders including the Undersecretariat of the Treasury, the Ministry of Industry and Commerce, the State Planning Organization, TUBITAK (the Scientific and Technological Research Council of Turkey), Middle East Technical University, the Ministry of Agriculture and Rural Affairs, the Ministry of Culture and Tourism, the Ministry of Labor and Social Security, the Ministry of Education, the Undersecretariat of Foreign Trade, TUSIAD (Turkish Industry and Business Association), MUSIAD (Association of Independent Industrialists and Businessmen), TOBB (The Union of Chambers and Commodity Exchanges of Turkey), and Exporters Association of Turkey. The White Book (Beyaz Kitap) includes the principles and the goals of clustering policies, as well as the levels and stages of policy process, policy tools and success factors (Beyaz Kitap, 2007, p.11).

Journal of Applied Business and Economics vol. 16(3) 2014 139 In addition to this general framework, in recent years academic interest on clustering approach is on the rise in Turkey as well (Eraslan, Bulu, & Bakan, 2008, p.5; Alsac, 2010). In the sections that follow, we introduce a cluster density index using a standard cluster identification method in Europe, and then test whether this index is associated with certain elements identified directly or indirectly in the Porter’s Diamond Model.

METHODS

The goal of this study is twofold: (1) developing a clustering map of Turkish manufacturing industry, and (2) assessing the relationship between clustering density and certain variables of interest. For this purpose, available data allows us to map out the industrial clusters for 2008 and 2011. What are the methods used for industrial clustering? These methods range from using qualitative information to sophisticated statistical methods. Some examples of these methods include face‐to‐face interviews, focus groups, 3-Star Analysis, Input-Output Analysis, and Shift Share Analysis. One of the common methods used in Europe to compare cluster formations across the member countries is the 3-Star Analysis.9 This study utilizes the 3-Star Approach in identifying industrial clusters in Turkey. As the name suggests, three different analyses are performed in the 3-Star Analysis. These are the Location Quotient (LQ) analysis, local dominance (sector’s share in local economy), and size (local sector’s share in national sector). Commonly used in "cluster mapping" studies in the European Union (EU), the 3-Star Analysis for the manufacturing sectors at NACE Rev. 2 at the two-digit level (see Appendix 1) is conducted across 26 regions (NUTS-II) in Turkey. These 26 regions (Table 1) overlap with the 26 Economic Development Districts introduced in Turkey in 2006. Because of the data availability issues at the sectoral level across the regions, we used employment by industry data, which was Distribution of Insured Persons and Work Place by Activity Groups and Provinces (Under Article 4-1/a of Act 5510), retrieved from the Social Security Institution of Turkey. There are three important criteria used in the 3-Star Analysis: Size: This measure is defined as the ratio of region’s employment to the national employment in a given industry, specified as

= (1) 푒푖푗

푆푖푧푒 퐸푖 = = 푡ℎ ′ 푡ℎ 푖푗 푤ℎ푒푟푒푡ℎ푒 푖 푖푛푑푢푠푡푟푦′ 푠 푒푚푝푙표푦푚푒푛푡 푖푛 푗 푟푒푔푖표푛 Dominance:퐸푖 푖 푖푛푑푢푠푡푟푦 This measure푠푒푚푝푙표푦푚푒푛푡 is related 푖푛to 푡ℎ푒a sector’s푛푎푡푖표푛 relative strength in a region’s economy, defined as

= (2) 푒푖푗 퐷표푚푖푛푎푛푐푒 푒푗 = = 푡ℎ ′ 푡ℎ 푤ℎ푒푟푒 푒푖푗 푖 푖푛푑푢푠푡푟푦 푠 푒푚푝푙표푦푚푒푛푡 푖푛 푗 푟푒푔푖표푛 푡ℎ ′ 푗 푒Specialization:푗 푟푒푔푖표푛 This푠 푡표푡푎푙 measure푒푚푝푙표푦푚푒푛푡 is the ratio of the share of employment in the sector in the region to the share of employment in the sector in the total employment in the country. Here;

( ) = / (3) 푒푖푗 퐸푖 푆푝푒푐푖푎푙푖푧푎푡푖표푛 퐿푄 � �푗 � � �푛 � = 푒 퐸 = 푡ℎ ′ 푡ℎ 푖푗 푤ℎ푒푟푒푡ℎ푒 푖 ′푖푛푑푢푠푡푟푦 푠푒푚푝푙표푦푚푒푛푡 푖푛 푗 푟푒푔푖표푛 푒푗 푗 푟푒푔푖표푛 푠 푡표푡푎푙 푒푚푝푙표푦푚푒푛푡

140 Journal of Applied Business and Economics vol. 16(3) 2014 = = 푡ℎ ′ 푖 퐸 푖 푖푛푑푢푠푡푟푦 푠푒푚푝푙표푦푚푒푛푡 푖푛 푡ℎ푒 푛푎푡푖표푛 퐸푛 What푡표푡푎푙 are푒푚푝푙표푦푚푒푛푡 some common푖푛 thresholds푡ℎ푒 푛푎푡푖표푛 used to assign a star to an industry? There is no standard measure in the literature. For example, some studies determine a threshold value for each criterion (size, dominance, and specialization). A sector exceeding the threshold value in any one of the ‘size,’ ‘dominance,’ or ‘specialization’ criteria is given one star. Figure 3 establishes the relationship between star levels and maturity of clusters.

FIGURE 3 3-STAR CLUSTER MAPPING APPROACH AND CLUSTER MATURITY

3 Stars (Mature Cluster)

2 Stars (Potential Cluster)

1 Star (Candidate Cluster)

What thresholds are used in this study? Threshold values differ by studies. For example, in the Ketels & Solvell's 3-Star Clustering Analysis study including 10 new EU member countries, the threshold value for ‘size’ and ‘dominance’ criteria was 0.7 and the threshold for ‘specialization’ coefficient was higher than (LQ>) 1.75 (Ketels and Sölvell, 2006, p.24). Cluster Observatory uses a different approach in its study covering all EU countries. Cluster Observatory first calculates the 'size' and 'focus' [‘dominance’ in this paper] values. Then, according to the calculated 'size' value, it allocates a star to the best 10% of the clusters that are located in a region. Similarly, according to the calculated 'focus' value, it allocates a star to the best 10% of the clusters that are located in a region. When evaluating the specialization criterion, Cluster Observatory assumes a Location Quotient value higher than (LQ>) 2. 10 The cluster analysis in this paper is based on the methodology used by Cluster Observatory. However, the assumptions regarding the threshold values were relaxed. We used the NUTS-II regions (26 regions) and NACE Rev. 2 two-digit sector codes (24 manufacturing sectors). First, ‘size’, ‘focus’[dominance], and ‘specialization’ values were calculated. Then, according to the calculated 'size' or ‘focus’[dominance] values, a star was allocated to the best 20% of the clusters located in a region. The threshold value for specialization was set as higher than (LQ >) 1.

Conceptual Framework and Hypothesis Figure 4 below lays out the conceptual framework and hypothesized relationships in this study. The method for the cluster identification phase is already outlined above.

Journal of Applied Business and Economics vol. 16(3) 2014 141

Figure 4 highlights the relationship between cluster development and macroeconomic dynamics in a region. The study proposes a summary measure of cluster dynamics at the regional level and then explores the relationship between this summary measure and macroeconomic indicators. Based on literature review, we have three hypotheses:

Hypothesis 1: Economic openness level is positively related to the high level of cluster density in a region.

We assume that openness in a regional economy promotes a competitive business environment.

Hypothesis 2: Economic development level is positively related to the high level of cluster density in a region.

We assume that economic growth and cluster development are positively associated.

Hypothesis 3: Government incentives are positively related to the high level of cluster density in a region.

We assume that government incentives promote regional cluster development.

Data Manufacturing employment data is obtained from the Social Security Institution of Turkey. The regional level data that includes export, import, population, public incentives, and value-added is from the Ministry of Development (www.dpt.gov.tr).

Measurement and Definitions Cluster Density Index (CDI) is defined as

= [(1 × ) + (2 × ) + (3 × )]/24 (4)

푛 푛 푛 퐶퐷퐼= 퐶퐶 푃퐶 푀퐶 = 푛 푤ℎ푒푟푒 퐶퐶 =푁푢푚푏푒푟 표푓 퐶푎푛푑푖푑푎푡푒 퐶푙푢푠푡푒푟푠 푛 24푃퐶 = 푁푢푚푏푒푟 표푓 푃표푡푒푛푡푖푎푙 퐶푙푢푠푡푒푟푠 푀퐶푛 푁푢푚푏푒푟 표푓 푀푎푡푢푟푒 퐶푙푢푠푡푒푟푠 푁푢푚푏푒푟 표푓 푀푎푛푢푓푎푐푡푢푟푖푛푔 푆푒푐푡표푟푠 퐼푛푐푙푢푑푒푑 푖푛 푡ℎ푒 퐴푛푎푙푦푠푖푠

142 Journal of Applied Business and Economics vol. 16(3) 2014 Openness is defined as per capita trade volume (exports + imports). Economic Development Level is defined as per capita value added in 2008. Public Incentives for Investment is defined as per capita public incentives for investment projects across the region.

Spearman’s Rank-Order Correlations Spearman’s rank-order correlation is the nonparametric version of the Pearson product-moment correlation. Because of the sample size (26 regions) and large outliers across the variables and regions (Table 4 below), we opted for the rank-order correlation rather than a regression analysis. We rank regions by cluster density index (CDI), openness, economic development level, and public incentives per capita from 1 (the highest score value) to 26 (the lowest score value in a given category). The formula for the Spearman’s rank-order correlation is given below

= 1 ( 2 ) (5) 6 ∑ 푑푖 2 휌 − 푛 푛 −1 = = The Spearman’s rank-order correlation ( ) takes a value between +1 and -1; +1 indicating a 푖 푊ℎ푒푟푒perfect association;푑 푑푖푓푓푒푟푒푛 0 indicating푐푒 푖푛 푝푎푖푟푒푑 “no-푟푎푛푘푠association”;푎푛푑 푛 and푛푢푚푏푒푟 -1 indicating표푓 푐푎푠푒푠 a perfect negative association. We used SPSS to analyze the association between 휌the표푟 cluster푟푠 density and variables of interest.

CLUSTERS BY REGIONS

The results of the 3-Star analysis are summarized below in Table 1 for 2008 and in Table 2 for 2011. The analyses were conducted for all NACE Rev. 2 sectors. In this paper, only manufacturing sectors are provided (Appendix 1). The results for all NACE Rev. 2 sectors can be provided by the authors upon request. Tables 1 and 2 present the results of the 3-Star Cluster analysis for 2008 and 2011. By looking at these tables, we would like to offer a few general conclusions about the methodology and findings. First, the many industries in the large regions such as TR10 (Istanbul) tend to get 1 star cluster designation. Second, there has been a dramatic shift across the regions and within the manufacturing industry between 2008 and 2011. For example, the number of industries receiving candidate cluster status through ‘size’ and ‘focus’ increased dramatically while the industries getting the same designation through the ‘specialization’ (LQ) declined by 57 percent. Finally, the number of 3-Star Clusters (mature clusters) increased from 25 to 27 across the regions. A notable shift occurred in Istanbul (TR10) as the region had its three mature clusters in 2011. These were (1) manufacture of wearing apparel, (2) manufacture of rubber and plastic products, and (3) manufacture of fabricated metal products, except machinery and equipment.

Journal of Applied Business and Economics vol. 16(3) 2014 143 TABLE 1 REGIONS WITH TYPE OF CLUSTERS: 2008 g yp

A=Size (1Star) B=Dominance (1Star) C=LQ (1Star) AB=Size and Dominance (2Stars) AC=Size and LQ (2Stars) Region BC=Dominance and LQ (2Stars) and ABC=Size, Dominance and LQ (3Stars) A B C AB AC BC ABC 10, 11, 15, 16, 17, 18, 13, 14, 22, TR10 (İstanbul) 19, 20, 23, 24, 26, 27, 21, 32 25 28, 29, 30, 31, 33 TR21 (Tekirdağ, Edirne, Kırklareli) 25 22, 23, 24, 26, 27 11, 15, 17, 20, 21 10 13,14 11, 15, 16, 20, 28, 29, TR22 (Balıkesir, Çanakkale) 14 10, 23, 24 30, 31, 33 11, 12, 15, 16, 17, 10, 14, TR31 (İzmir) 13, 32 23, 31 18, 19, 20, 22, 24, 25, 28 26, 27, 29, 33 11, 17, 25, 26, 27, 30, 13, 14, TR32 (Aydın, Denizli, Muğla) 16, 32 10 33 23 TR33 (Kütahya, Manisa, Afyon, Uşak) 11, 16, 20, 22, 24, 29 12, 15, 18, 26, 27 10, 13, 25 23 10, 11, 16, 17, 20, 13, 25, TR41 (Bilecik, Bursa, Eskişehir) 18, 33 14 22, 23, 28 11, 16, 17, 19, 20, TR42 (Bolu, Düzce, Kocaeli, Sakarya, 10, 22, 14 27 21, 23, 24, 26, 29, Yalova) 25, 28 30, 31, 33 11, 15, 18, 21, 26, 25, 28, TR51 (Ankara) 16, 20, 23, 24 10 27, 30 33 11, 15, 16, 17, 18, 22, TR52 (Konya, Karaman) 10, 29 23, 25, 28 24, 27, 31 11, 16, 18, 19, 20, 27, 10, 13, 23, TR61 (Antalya, Burdur, Isparta) 32 30, 31, 33 25 11, 12, 16, 18, 20, 22, 10, 13, 14, TR62 (Adana, Mersin) 17, 19 23, 24, 27, 28, 32, 33 25 TR63 (Hatay, Kahramanmaraş, 16, 19, 23, 29 12 10, 13, 25 24 Osmaniye) TR71 (Kırıkkale, Aksaray, Niğde, 11, 13, 16, 18, 22, 24, 10, 14, 23, 19 Nevşehir, Kırşehir) 28, 29, 31, 33 25 17, 20, 22, 24, 26, 27, 10,13,23,2 TR72 (Kayseri, Sivas, Yozgat) 30, 31 28, 33 5 TR81(Bartın, Karabük, Zonguldak) 10, 15, 16, 23, 30, 31 14, 25, 33 24 10, 14, 16, TR82 (Çankırı, Kastamonu, Sinop) 20, 24, 26, 30 23 TR83 (Amasya, Çorum, Samsun, 11, 16, 17, 18, 22, 24, 10, 14, 23, 12, 21 Tokat) 26, 28, 31, 32, 33 25 TR90 (Artvin, Giresun, Gümüşhane, 11, 12, 16, 18, 27, 30, 14 23, 33 10 Ordu, Rize, Trabzon) 31, 32 16, 18, 19, 24, 26, 31, 10, 23, 25, TRA1 (Bayburt, Erzincan, Erzurum) 32 33 10, 15, 18, TRA2 (Ağrı, Ardahan, Iğdır, Kars) 11, 16, 33 23 TRB1 (Bingöl, Elazığ, Malatya, 11, 12, 16, 17, 18, 22, 10, 13, 14, Tunceli) 27, 30, 33 23 12, 16, 18, 19, 24, 30, 10, 14, 23, TRB2 (Bitlis, Hakkari, Muş, Van) 32 33 TRC1 (Adıyaman, Gaziantep, Kilis) 14 17, 19 12, 15 10 13, 22 11, 16, 18, 19, 20, 27, 10, 13, 23, TRC2 (Şanlıurfa, Diyarbakır) 12 28, 32 33 11, 12, 16, 18, 20, 27, 10, 14, 19, TRC3 (Batman, Mardin, Şırnak, Siirt) 28, 32, 33 23

144 Journal of Applied Business and Economics vol. 16(3) 2014 TABLE 2 REGIONS WITH TYPE OF CLUSTERS: 2011

A=Size (1Star) B=Dominance (1Star) C=LQ (1Star) AB=Size and Dominance (2Stars) AC=Size and LQ (2Stars) Region BC=Dominance and LQ (2Stars) and ABC=Size, Dominance and LQ (3Stars) A BC AB AC BC ABC 10, 11, 12, 16, 23, 15, 17, 18, 20, 21, TR10 (İstanbul) 13 14, 22, 25 24, 29, 30, 31 26, 27, 28, 32, 33 TR21 (Tekirdağ, Edirne, Kırklareli) 21 10, 25 26 11, 15, 17, 20, 27 13, 14 11, 16, TR22 (Balıkesir, Çanakkale) 25 10, 23, 24 20, 31 11, 12, 15, 17, 19, TR31 (İzmir) 16, 18, 22, 29 24 10, 25 14, 28 20, 26, 31, 32, 33 TR32 (Aydın, Denizli, Muğla) 11, 16, 33 14 10 13, 23 TR33 (Kütahya, Manisa, Afyon, Uşak) 12, 18 13, 25 11, 16 15, 26, 27 10 23 14, 16, 17, 20, 22, TR41 (Bilecik, Bursa, Eskişehir) 10 11, 23, 27, 28, 31 13, 25, 29 30, 33 TR42 (Bolu, Düzce, Kocaeli, Sakarya, 14, 18, 23, 26, 32, 11, 16, 17, 19, 20, 10 22, 24, 25 Yalova) 33 21, 27, 28, 29, 30 18, 24, 26, 30, 31, TR51 (Ankara) 15, 21 10 27 25, 28, 33 32 TR52 (Konya, Karaman) 15, 16 29 24, 25, 28 10 11, 18, TR61 (Antalya, Burdur, Isparta) 13, 25 16, 32 10, 23 20, 30, 31 16, 20, TR62 (Adana, Mersin) 13, 14, 25 22, 23, 11, 12, 17, 19 10 24, 28, 33 TR63 (Hatay, Kahramanmaraş, 10, 25 19 13 24 Osmaniye) TR71 (Kırıkkale, Aksaray, Niğde, 11, 22, 10,23, 25, 19 Nevşehir, Kırşehir) 24, 31 28 TR72 (Kayseri, Sivas, Yozgat) 13 24, 27 30 10, 25 31 TR81(Bartın, Karabük, Zonguldak) 10 15, 16, 30 14, 23 24 TR82 (Çankırı, Kastamonu, Sinop) 20, 27, 30 10, 14 23 16 16, 17, TR83 (Amasya, Çorum, Samsun, 21 14, 25 24, 31, 10, 23 Tokat) 32, 33 TR90 (Artvin, Giresun, Gümüşhane, 12 14, 25 11, 16, 31 23 10 Ordu, Rize, Trabzon) 11, 16, TRA1 (Bayburt, Erzincan, Erzurum) 25 10, 23, 33 18, 24, 32 TRA2 (Ağrı, Ardahan, Iğdır, Kars) 33 11, 16 10, 18, 23 TRB1 (Bingöl, Elazığ, Malatya, 11, 30, 10, 13, Tunceli) 31, 33 14, 23 TRB2 (Bitlis, Hakkari, Muş, Van) 13, 14 18, 19, 33 10, 23 TRC1 (Adıyaman, Gaziantep, Kilis) 14 17 15 10 13, 22 TRC2 (Şanlıurfa, Diyarbakır) 27, 28 12, 13 10, 23, 33 11, 18, TRC3 (Batman, Mardin, Şırnak, Siirt) 13 19 10, 23, 33 31, 32

Table 3 below highlights regional distribution of mature clusters in 2008 and 2011. The TR31 (İzmir) region experienced a dramatic shift between 2008 and 2011 as its two sectors lost its mature cluster designation in this period. Mature clusters in TR31 (Izmir) Region in 2008 were (1) manufacture of food products, (2) manufacture of wearing apparel, (3) manufacture of fabricated metal products, except machinery and equipment, and (4) manufacture of machinery and equipment n.e.c. In 2011, fabricated metals and food products were no longer part of the mature industrial clusters in this region.

Journal of Applied Business and Economics vol. 16(3) 2014 145 A similar trend is visible across the western regions such as TR32 (Aydın, Denizli, Muğla) region, which lost its competitive advantage in wearing apparel manufacturing industry, and TR42 (Bolu, Düzce, Kocaeli, Sakarya, Yalova) region, which lost its competitive advantage in food products manufacturing industry. On the positive side, there were several regions that made significant progress: TR52 (Konya, Karaman) has gained competitive advantage in food manufacturing industry; TR72 (Kayseri, Sivas, Yozgat) region has gained competitive advantages in furniture manufacturing industry; and TR82 (Çankırı, Kastamonu, Sinop) region has gained competitive advantage in woods, wooden products and cork products manufacturing (excluding furniture) industry.

TABLE 3 MATURE INDUSTRIAL CLUSTERS: 2008-2011

Sectors with three stars in Sectors with three stars 2008 (Mature Clusters) in 2011 (Mature Clusters)

TR10 (İstanbul) 14, 22, 25 TR21 (Tekirdağ, Edirne, Kırklareli) 13,14 13, 14 TR22 (Balıkesir, Çanakkale) TR31 (İzmir) 10, 14, 25, 28 14, 28 TR32 (Aydın, Denizli, Muğla) 13, 14, 23 13, 23 TR33 (Kütahya, Manisa, Afyon, Uşak) 23 23 TR41 (Bilecik, Bursa, Eskişehir) 13, 25, 28 13, 25, 29 TR42 (Bolu, Düzce, Kocaeli, Sakarya, Yalova) 10, 22, 25, 28 22, 24, 25 TR51 (Ankara) 25, 28, 33 25, 28, 33 TR52 (Konya, Karaman) 10 TR61 (Antalya, Burdur, Isparta) TR62 (Adana, Mersin) TR63 (Hatay, Kahramanmaraş, Osmaniye) 24 24 TR71 (Kırıkkale, Aksaray, Niğde, Nevşehir, Kırşehir) TR72 (Kayseri, Sivas, Yozgat) 31 TR81(Bartın, Karabük, Zonguldak) 24 24 TR82 (Çankırı, Kastamonu, Sinop) 16 TR83 (Amasya, Çorum, Samsun, Tokat) TR90 (Artvin, Giresun, Gümüşhane, Ordu, Rize, 10 10 Trabzon) TRA1 (Bayburt, Erzincan, Erzurum) TRA2 (Ağrı, Ardahan, Iğdır, Kars) TRB1 (Bingöl, Elazığ, Malatya, Tunceli) TRB2 (Bitlis, Hakkari, Muş, Van) TRC1 (Adıyaman, Gaziantep, Kilis) 13, 22 13, 22 TRC2 (Şanlıurfa, Diyarbakır) TRC3 (Batman, Mardin, Şırnak, Siirt)

As highlighted in Tables 1-3, there have been changes in the fortunes of the regions. What might be some of the factors that are related to the cluster density across the regions? Does relatively high trade volume promote a competitive business environment? How about the role of government? How does initial development level contribute to clustering in future periods? These are some of the critical questions that should be explored in Turkey. Because of data limitations, in the next section, we will look at the nonparametric relationship between the cluster density index and these variables without implying any causal relationships.

146 Journal of Applied Business and Economics vol. 16(3) 2014 STUDY RESULTS

Summary Data Table 4 below summarizes major variables used in this paper. For each variable, the five best performing regions (green) and the five worst performing regions (red) are highlighted. There are significant variations across the regions in terms of economic development, openness, cluster density, and public investment incentives. The coefficient of variations are higher than 0.4 across all variables suggesting regional uneven economic development. One notable observation is that the coefficient of variation for per capita trade volume is 1.24 suggesting that a few regions account for a large portion of the trade volume in the nation. Indeed, a review of data shows the extreme concentration of trade activities in a few regions, with Istanbul topping the list as an extreme outlier.

Spearman’s Rank-Order Correlations As Table 4 clearly shows, data is highly skewed to conduct regression analysis. Spearman’s rho seems to be the best approach to look at the association between the Cluster Density Index and regional macroeconomic indicators. Table 5 reports the strength of association between the Cluster Density Index and per capita trade volume. Spearman rho correlation coefficient suggests a statistically significant relationship between the cluster density and openness in the region (rs(26)=.575, p<0.002). If we square the correlation coefficient, we can argue that 33.1 percent of the variance in cluster density is accounted for by per capita trade volume, and likewise, 33.1 percent of the variance in per capita trade volume is accounted for by the cluster density. Test result confirms our first hypothesis.

TABLE 4 SUMMARY DATA: CLUSTER DENSITY INDEX, VALUE ADDED,

PUBLIC y INVESTMENT y INCENTIVES AND TRADE VOLUME

Per Capita Cluster Per Capita Public Per Capita Value Added in Density Incentives for Trade Volume Region 2008 (TL) Index 2011 Investment 2011 (TL) in 2011 (US$) TR10 (İstanbul) 14,591 1.67 3,586 13,379 TR21 (Tekirdağ, Edirne, Kırklareli) 12,243 0.83 6,892 1,227 TR22 (Balıkesir, Çanakkale) 9,000 0.46 4,682 698 TR31 (İzmir) 11,568 1.46 4,636 4,667 TR32 (Aydın, Denizli, Muğla) 9,076 0.54 3,753 2,232 TR33 (Kütahya, Manisa, Afyon, Uşak) 8,256 0.71 3,731 2,972 TR41 (Bilecik, Bursa, Eskişehir) 12,983 1.17 3,971 6,850 TR42 (Bolu, Düzce, Kocaeli, Sakarya, Yalova) 13,265 1.54 6,668 8,570 TR51 (Ankara) 12,598 1.04 3,078 3,729 TR52 (Konya, Karaman) 7,213 0.54 3,259 1,177 TR61 (Antalya, Burdur, Isparta) 10,334 0.63 5,388 761 TR62 (Adana, Mersin) 7,363 0.83 7,680 1,919 TR63 (Hatay, Kahramanmaraş, Osmaniye) 5,937 0.33 4,543 3,131 TR71 (Kırıkkale, Aksaray, Niğde, Nevşehir, Kırşehir) 6,789 0.58 4,359 535 TR72 (Kayseri, Sivas, Yozgat) 6,813 0.50 2,620 1,438 TR81(Bartın, Karabük, Zonguldak) 8,734 0.46 3,640 2,599 TR82 (Çankırı, Kastamonu, Sinop) 6,676 0.50 4,114 218 TR83 (Amasya, Çorum, Samsun, Tokat) 6,914 0.54 2,084 689 TR90 (Artvin, Giresun, Gümüşhane, Ordu, Rize, Trabzon) 7,059 0.46 3,078 921 TRA1 (Bayburt, Erzincan, Erzurum) 5,520 0.50 3,184 100 TRA2 (Ağrı, Ardahan, Iğdır, Kars) 3,601 0.38 1,748 225 TRB1 (Bingöl, Elazığ, Malatya, Tunceli) 5,517 0.50 2,516 295 TRB2 (Bitlis, Hakkari, Muş, Van) 3,419 0.38 1,577 217 TRC1 (Adıyaman, Gaziantep, Kilis) 4,597 0.50 4,698 3,853 TRC2 (Şanlıurfa, Diyarbakır) 3,724 0.50 1,336 200 TRC3 (Batman, Mardin, Şırnak, Siirt) 3,812 0.54 2,326 955 Average 7,985 0.70 3,813 2,444 Standard Deviation 3,211 0.37 1,567 3,022 Coefficient of Variation (CV) 0.40 0.53 0.41 1.24 Source: Authors' calculations from official government statistics, Ministry of Development, www.dpt.gov.tr

Journal of Applied Business and Economics vol. 16(3) 2014 147 TABLE 5 CORRELATIONS \ SPEARMAN’S RHO: CLUSTER DENSITY INDEX VS. OPENNES (PER CAPITA TRADE VOLUME) Cluster Density Per Capita Trade Index Volume (Opennes) Spearman's rho Cluster Correlation Coefficient 1.000 .575** Density Index Sig. (2-tailed) .002 N 26 26 Per Capita Correlation Coefficient .575** 1.000 Trade Volume Sig. (2-tailed) (Openness) .002 N 26 26 ** Correlation is significant at the 0.01 level (2-tailed).

Our second hypothesis was about the positive association between economic development level and cluster density index. Table 6 presents correlation coefficients for this relationship. In this analysis, we used per capita value added in 2008 as a proxy for economic development level of regions. The correlation coefficient is statistically significant and has the expected positive sign (rs(26)=.74, p<0.000). The squaring the correlation coefficient suggests that 54.8 percent of the variance in the cluster density is accounted for by economic development level, and similarly, 54.8 percent of the variance in the economic development is accounted for by the cluster density in a region.

TABLE 6 CORRELATIONS \ SPEARMAN’S RHO: CLUSTER DENSITY INDEX VS. PER CAPITA VALUE ADDED Cluster Density Per Capita Value Index Added

Spearman's rho Cluster Correlation Coefficient 1.000 .740** Density Index Sig. (2-tailed) .000 N 26 26 Per Capita Correlation Coefficient .740** 1.000 Value Added Sig. (2-tailed) .000 N 26 26 ** Correlation is significant at the 0.01 level (2-tailed).

Our last hypothesis was about the role of government in cluster development. We hypothesized that government incentives for investment are positively associated with the cluster density index. Table 7 below suggests that there is a statistically significant relationship between the per capita government incentives for investment and the cluster density index, and the correlation coefficient has the expected sign. However, the relationship is not as robust as we reported in the previous tables (rs(26)=.401, p<.042). The squaring of the coefficients in this case suggests that only 16.1 percent of the variance in the cluster density is accounted for by the per capita government incentives for investment, and similarly, only 16.1 percent of the variance in the per capita government incentives for investment is accounted for by the cluster density in a region. This finding suggests that the government incentives for investment have different macroeconomic dynamics.

148 Journal of Applied Business and Economics vol. 16(3) 2014 TABLE 7 CORRELATIONS \ SPEARMAN’S RHO: CLUSTER DENSITY INDEX VS. PER CAPITA GOVERNMENT INCENTIVES Cluster Density Per Capita Index Government Incentives Spearman's rho Cluster Correlation Coefficient 1.000 .401* Density Index Sig. (2-tailed) .042 N 26 26 Per Capita Correlation Coefficient .401* 1.000 Government Sig. (2-tailed) incentives .042 N 26 26 * Correlation is significant at the 0.05 level (2-tailed).

CONCLUSION AND RECOMMENDATIONS

In this study, we attempted to achieve several goals. First, we provided a brief review of the importance of clustering in regional economic development. Second, we highlighted critical literature and institutions in clustering efforts. Third, adopting a commonly used clustering methodology, we performed a 3-Star Cluster Mapping for the manufacturing industry across the 26 Economic Development Districts in Turkey. Finally, we tested several relationships between the regional cluster density and macroeconomic indicators. Our findings about the relationship among macroeconomic indicators are in line with the findings in literature. In other words, regional economic clustering has a close relationship with (a) international openness of the region, and (b) economic growth in a region. Although targeted government incentives for investment should promote cluster type industrial concentration in the regions, the findings show a weak relationship between the cluster density and government incentives. This may be because (a) the government efforts alone to create industrial clusters are not enough, or (b) the government efforts are directed towards the less developed areas in which case it may take some time to assess the impact of those efforts on mature cluster formation. For future studies, we recommend the further refinement of the cluster density index developed here as a broader regional cluster summary measure. We also recommend that businesses should strategically position themselves in areas of international trade as the per capita trade volume seems to promote competitive economic dynamics and cluster formation.

ENDNOTES

1. For a review of activities of each agency, see the Ministry of Development at http://www.mod.gov.tr/en/SitePages/mod_aboutus.aspx 2. The Tennessee Valley Authority, Access Date 05.05.2013, http://www.tva.com/abouttva/history.htm 3. European Cluster Policy Group Portal, Access Date 05.05.2013, http://www.proinno-europe.eu/ecpg 4. U.S. Cluster Mapping Portal, Access Date 05.05.2013, http://mvp.clustermapping.us/ 5. Harvard University, Institute for Strategy and Competitiveness , Access Date 05.05.2013, http://www.isc.hbs.edu/econ-clusters.htm 6. The Center for Strategy and Competitiveness (CSC), Access Date 05.05.2013, http://www.hhs.se/csc/Pages/default.aspx 7. The Cluster Observatory Portal, Access Date 05.05.2013, http://www.clusterobservatory.eu/index.html 8. The new Europe INNOVA Portal, Access Date 05.05.2013, http://archive.europe-innova.eu/index.jsp 9. Access Date 05.05.2013, http://www.clusterobservatory.eu/index.html#!view=aboutobservatory;url=/aboutobservatory/methodology/indicators/

Journal of Applied Business and Economics vol. 16(3) 2014 149 10. Access Date 05.05.2013, http://www.clusterobservatory.eu/index.html#!view=aboutobservatory;url=/aboutobservatory/methodology/indicators/

REFERENCES

Alsac, F. (2010). Bölgesel gelişme aracı olarak kümelenme yaklaşımı ve Türkiye için kümelenme destek modeli önerisi [Clustering approach as a regional economic development tool and a proposal for Turkey], DPT Uzmanlığı Tezi, Ankara: DPT. Andersson, T., Serger, S.S., Sörvik, J., and Hansson, E.W. (2004). The cluster policies whitebook, International Organization for Knowledge Economy and Enterprise Development, Sweden. Barro, R.J. and Sala-I-Martin, X. (1995). Economic growth, Cambridge: The MIT Press. Beyaz Kitap [White Book]. (2007). Türkiye için kümelenme politikasının geliştirilmesi projesi [A project for development of industrial cluster policy in Turkey], Ankara: Başbakanlık Dış Ticaret Müsteşarlığı (Access Date 05.05.2013, http://www.smenetworking.gov.tr/userfiles/pdf/dcp/Beyaz%20Kitap/Beyaz_Kitap.pdf. Delgado, M., Porter, M.E. and Stern, S. (2010). Clusters and entrepreneurship, US Census Bureau Center for Economic Studies, Paper No. CES-WP-10-31 http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1689084. Eraslan, H., Bulu, M. and Bakan, İ. (2008). Kümelenmeler ve inovasyona etkisi: Türk turizm sektöründe uygulamalar [Clusters and their effects on innovation: Applications in Turkish tourism sector], Seyahat ve Otel İşletmeciliği Dergisi, Yıl:5 Sayı:3. Feser, E. J., and Bergman, E.M. (2000). National industry cluster templates: A framework for applied regional cluster analysis, Regional Studies, 34, 1-19. Feser, E. J. (2005). Benchmark value chain industry clusters for applied regional Research, Regional Economics Applications Laboratory, University of Illinois at Urbana-Champaign. Isbasoiu, G. M. (2007). Industrial clusters and regional development: The case of Timisoara and Montebelluna, MPRA Paper No. 5037, 3. (Access Date 05.05.2013, http://mpra.ub.uni- muenchen.de/5037/). Isard, W. (1960). Methods of regional analysis: An introduction to regional science. Cambridge: Published jointly by the Technology Press of the Massachusetts Institute of Technology and Wiley, New York. Ketels, C., and Sölvell, Ö. (2006). Innovation clusters in the 10 new member states of the European Union, Europe INNOVA Paper, No:1. Keynes, J. M. (1936). The General theory of employment, interest and money. London: Pelgrave Mcmillen. Klepper, S. (2010). The origin and growth of industry clusters: The making of Silicon Valley and Detroit, Journal Of Urban Economics, 67, 15-32. Marshall, A. (1890). Principles of economics. London: Mcmillen and Co., Ltd. Maskell, P. and Kebir, L. (2006). 'What qualifies as a cluster theory?'. In: Asheim, B.T., Cooke, P. and Martin, R.L., (Eds.) Cluster and regional development: Critical reflections and explorations, London: Routledge, 30-49. Nijkamp, P. and Mills, E. S. (1986). Handbook of regional and urban economics, Volume I, Regional Economics, Elsevier Science Publishers. Porter, M.E. (1990). The competitive advantage of nations. New York: Free Press. Porter, M.E., (1998). Clusters and competition: New agendas for companies, governments and institutions. In M.E. Porter (ed.). On competition. Boston: Harvard Business School Press, 197- 299. Porter, M.E., (2003). The Economic performance of regions,” Regional Studies 37, pp. 549-578. Saxenian A. (1996). Regional advantage, culture and competition in Silicon Valley and Route 128. Cambridge: Harvard University Press.

150 Journal of Applied Business and Economics vol. 16(3) 2014 TURKSTAT. (2011). Regional gross value added at current basic prices- By kind of NUTS1 (2004- 2008), Press Releases, No.10704, 26/04/2011 (Access date 05.05.2013, http://www.turkstat.gov.tr/Start.do). Tuyluoglu, S. ve Karakas, D. N. (2006). Bölgesel kalkınma ve ekonomik durgunlaşma süreci: Zonguldak örneği, [Regional development and the process of economic stagnation: The case of Zonguldak province], Amme İdaresi Dergisi, Cilt 39, Sayı:4. S:195-224. Quandt, C. and Pacheco, L. (2000). Fostering the growth of innovation clusters for regional development: Building a network of software clusters in Paraná, Brazil, 4th International Conference on Technology. Wennberg, K. and Lindqvist, G. (2008). How the entrepreneurs in cluster contribute to economic growth, SSE7EFI Working Paper Series in Business Administration No:2008:3.

APPENDIX

APPENDIX 1 NACE REV.2 STATISTICAL CLASSIFICATION OF ECONOMIC ACTIVITIES IN THE

EUROPEAN COMMUNITY (2-DIGIT LEVEL MANUFACTURING SECTORS)

10 Manufacture of food products 11 Manufacture of beverages 12 Manufacture of tobacco products 13 Manufacture of textiles 14 Manufacture of wearing apparel 15 Manufacture of leather and related products Manufacture of wood and of products of wood and cork, except 16 furniture; manufacture of articles of straw and plaiting materials 17 Manufacture of paper and paper products 18 Printing and reproduction of recorded media 19 Manufacture of coke and refined petroleum products 20 Manufacture of chemicals and chemical products 21 Manufacture of basic pharmaceutical products and pharmaceutical preparations 22 Manufacture of rubber and plastic products 23 Manufacture of other non-metallic mineral products 24 Manufacture of basic metals 25 Manufacture of fabricated metal products, except machinery and equipment 26 Manufacture of computer, electronic and optical products 27 Manufacture of electrical equipment 28 Manufacture of machinery and equipment n.e.c 29 Manufacture of motor vehicles, trailers and semi-trailers 30 Manufacture of other transport equipment 31 Manufacture of furniture 32 Other manufacturing 33 Repair and installation of machinery and equipment

Journal of Applied Business and Economics vol. 16(3) 2014 151

Effectiveness of Corporate Social Responsibility in Enhancing Company Image

Siphiwe P. Mandina Midlands State University, Zimbabwe

Christine V. Maravire Midlands State University, Zimbabwe

Victoria S. Masere Midlands State University, Zimbabwe

INTRODUCTION

Corporate Social Responsibility (CSR) has become an area of interest for many organizations. In recent times there has been much debate about whether corporations should be socially responsible or not and to what extent they should be responsible, (Visser, 2009). The phrase CSR is often hard to pin down because of the fact that there are several schools of thought concerning this notion. According to Robbins (2003) a business is a part of a large society and therefore it has a responsibility other than just maximizing profits. Organizations operate in a society therefore they have to be concerned about the society`s well-being. Robbins (2003) further states that, the core objective of a business is to make a profit, however if a business is socially responsible it has to be concerned about the well-being of the society it operates in. However scholars who include Friedman (1970) questions if organizations are required to take responsibility of social issues, he stresses that the sole social responsibility of a business is to boost its profits through legal ways and that donating an organization`s funds to the society is harmful to the organization`s profits as this reduces the organization`s profits. Porter and Krammer (2011) indicates that addressing social concerns could increase the levels of company productivity, with subsequent positive effects on profitability, share value and company image. Visser (2010) identified the

152 Journal of Applied Business and Economics vol. 16(3) 2014 five dimensions of CSR and these include; the Economic dimension, Legal dimension, Ethical dimension, Philanthropic and Environmental management.

Global Perspective on CSR Looking closely at international mining companies, it will be seen that almost all of them, consider CSR and its effects on their business operations particularly as it pertains to their corporate image and competitive advantage, (Mining Weekly June 2012). Davis (1973) in his work asserts that engaging in corporate social responsibility can improve a company`s image and finances. According to The Post, (October, 2012) the Open Society Initiative for Southern Africa stated that developing countries with rich mineral resources deserve solid CSR strategies for the people to begin to see tangible benefits from the exploitation of their country`s mineral wealth. It further stated that if the community cannot get adequate compensation for the exploitation of their mineral wealth by private mining companies, be it local or foreign, how will people be expected to move out of poverty. In Southern Africa, especially South African mining Corporations have come to realize that they cannot operate in isolation to the community that good governance and social involvement go beyond the work performed in the office. South African mining companies have stepped up their CSR activities with big corporates like, Optimum Coal, Great Basin Gold, Implants, Impala and Harmony Gold taking the lead, (www.miningwatch.com). However other mining companies are neglecting the community, and turning a blind eye to the plight of the community. These companies are falling short in terms of their commitments to the surrounding communities and their implementation of corporate social responsibility programs, (Mining Weekly June 2012). Neglecting the community has resulted in negative media reviews, which has led to strikes in the South African mining industry and also negative perceptions by stakeholders because of deteriorating social conditions related to a lack of service, crime, ill health and a deteriorating environment in the area which the mining companies operate, (Mining Weekly June 2012).

Zimbabwean Perspective on CSR The Zimbabwean economy mainly relies on agriculture and mining. There are many mining companies in Zimbabwe, some locally owned with some of them foreign owned. The major mining companies include Mbada diamonds, Todal Mine, Marange diamonds, Zimplants, Marowa, Mimosa and Zimasco. Even though most of these mining companies are involved in some certain CSR activities, stakeholders’ opinions are that these organizations are not doing enough for their local communities, CSR is said to be mostly limited to their mining workers with the communities surrounding the mine being largely ignored, (www.miningwatch.com). The Zimbabwe Environment Law Association has challenged the government to come up with comprehensive law governing the mining industry which compels companies to plough back to the community they mine. (www.miningwatch.com).

Unki Mine’s Perspective on CSR Unki Mine (UM) is a subsidiary of the Anglo American Platinum group with its head office in South Africa. The Anglo American Platinum group was the majority shareholder of Unki Mine until November 2012 when the group agreed to cede 51% of its stake to the locals in line with the indigenization act. The company is situated 60km from Gweru on Zimbabwe`s great dyke. It began its development operations in 2005, with its mining operations commencing in 2008, with Unki Mine`s concentrator commissioned in December 2010. Its main focus is on the mining of platinum. A twin decline shaft system provides access to the underground working for personnel and material. The life-of-mine of the current operations at Unki extends to 2041. When it began its operations Unki was expecting to produce 120 000 tone per month with potential for expansion, (www.angloplatinum.com) The end product is used as anti-pollutants in motor vehicles, jewelry, glass fuel cells and also as a refining catalyst in the petroleum industry, (www.miningwatch.com). In Zimbabwe there are several companies which are into mining of platinum these include among others Zimplants, Todal mine, and Mimosa Mine. Most foreign companies in the mining sector have experienced image problems, thus the community has a negative view towards these

Journal of Applied Business and Economics vol. 16(3) 2014 153 firms. The local community, the government and environmental groups are concerned about the exploitation of the country`s mineral resources by the foreign companies. “We are sitting on the richest land but our minerals are not benefiting us, there is a lot of secrecy by these mining companies”. (Zimbabwe Environment Law Society, Chairman Professor Tumai Murombo, May 2013). UM, through its Community Engagement and Development department, has been involved in numerous CSR activities since 2009. According to the Unki Mine newsletter (2013) the CSR activities include among the following; • Philanthropic Projects – community water and sanitation, resettlement of the displaced, infrastructure development, donations and bursaries, community health and community education. • Employing locals • Investment incentives • Environmental management – UM is ISO1400 certified • Adherence to the legal systems of the country- indigenization act

According to their Newsletter, Unki CED Newsletter (2013) UM has invested US$ 1 Million in community projects in 2013. In 2012 the researcher found out that UM made donations to the community of $21 183.00

TABLE 1 UNKI MINE DONATION REGISTER, (2012)

UNKI MINES PRIVATE LIMITED CED - CSI DONATIONS REGISTER Ref 003/2012 Donation Description Beneficiary Date Approved Value of donation USD Donation towards Independence Day Celebration 04.04.12 800 Donation to Midlands Province 300 Donation to DA's Office 300 Donation towards installation of Headman TRDC 06.07.2012 250 Mhangami DSTV Subscription for Thornhill Air Base 10.05.12 1033 Donation to Midlands Heroes Acre 07.06.12 1065 ZRP & ZNA Donation Cost ZRP & ZNA 3792.82 Donation Cost Midlands Show Society Midlands Show 1745 Society Midlands Show Society Midlands SS 2160 Donation towards Heroes 2012 Donation 2012 400 Financial Assistance For MSU students MSU 04.04.12 5538 2012 Heroes Commemorations Shurugwi Shurugwi 06.08.12 400 District ZNA Donation ZNA 22.05.12 1000 Donations towards Shurugwi Environments Fair 08.08.12 5 Computers Bizzy Bee Nursery BB Nursery 750 11 Computers Selukwe school 1650 TOTAL 21183.82 Source: Unki Mine CED Newsletter (2012)

154 Journal of Applied Business and Economics vol. 16(3) 2014 CSR is concerned with the relationship between a corporation, its stakeholders and the local society in which it operates. Since UM began its operations the Shurugwi community has experienced problems including displacement of families, thus people are being forced from their homes and families, assault to the environment, and also the social dynamics of the society has been changed, (Sunday News September 2012). Most importantly UM transports their concentrate, by road to their smelting plant in Polokwane, South Africa. This has resulted in an outcry from stakeholders claiming that foreign owned mining companies are exploiting Zimbabwean`s natural resources for their own benefit leaving Zimbabweans with nothing, (www.miningwatch.com) All this has resulted in a bad company image for UM, as is the case with all foreign owned mining companies in Zimbabwe. UM is involved in numerous CSR activities to try to improve their image. This study therefore seeks to determine the effectiveness of the current CSR projects by UM in enhancing their company image. The study will enable the researcher to investigate if the CSR activities being carried by UM are effective in enhancing their company image. CSR is an area of interest because of several reasons; firstly CSR has become a legal requirement, (www.miningwatch.com). In Zimbabwe the indigenization act in the mining sector, which leads to CSR in terms of the Community Share Ownership Trust, has become a legal regulation. Secondly CSR is an area of interest since most organizations are now practicing the societal marketing concept which gives emphasis to the fact that organizations must not only be concerned about themselves but also about the society`s well-being and lastly organizations in the 21st century are becoming more involved in the social welfare of the community, this has motivated this research, as it will enable the researcher to gain more insight on this concept of CSR.

DEFINING CSR

CSR has progressed from an irrelevant and often discriminated concept to one that is today well- known and established in businesses around the globe (Lee, 2008). CSR can be viewed as an umbrella phrase that considers the various means and ways that corporation embarks on in trying to act morally and ethically .CSR can also be referred to as Corporate Conscience, Corporate Citizen or Sustainable Responsible Business. Visser (2009) has rephrased the term corporate social responsibility to corporate sustainability and responsibility. In the last couple of decades, CSR has become widely known, (Campbell, 2007). Carroll, (1979) and Kantanen, (2005) posit that the first book on CSR was written in 1953 by Howard Bowen, under the tittle Responsibility of the Businessman. Defining CSR has proven to be a complex task since it has different meanings to different people. This is due to the fact that there is no agreed definition and as such organizations that are meticulous in their goals of incorporating CSR activities into their business are faced with compound problems. Since there is no agreed definition, organizations translate it to suit them depending on their state of affairs, (MacLagan 1999; Campbell 2007). Stakeholders make use of different definitions that are in line with their business operations, goals and aims. Bowen (1953) defines CSR as the obligations of businessmen to purse those policies, to make those decisions, or to follow those lines of action which are desirable in terms of the objectives and values of our society. Aguilera et al (2007) emphasizes that corporations should not border their CSR activities on stipulated legislation regarding such issues but should also make provision for activities not stipulated in any legislation they adhere to. Aguilera et al (2007) also asserts that CSR is a company`s consideration of and response to issues beyond the narrow economic, technical and legal requirements of the company to accomplish social and environmental benefits along with traditional economic gains. Carroll (1991) went further and identified aspects of CSR, he states that CSR consists of four aspects; legal, economic, ethical and philanthropic responsibility, Carroll (1991) further states that corporations striving to be seen as good within the society should fulfill these four aspects. The four aspects of CSR as postulated by Carroll (1991); (see figure 1).

Journal of Applied Business and Economics vol. 16(3) 2014 155 FIGURE 1 CSR PYRAMID

Source: Carroll(1991)

Carroll (1991) cites renowned economists Friedman’s assertion in trying to explain the relationship of the four aspects. Carroll (1991) sates that on the part of Friedman(1970) , he was only interested in the three parts of CSR stating that corporations exist to make as much money as possible while conforming to the basic rules of the society, both those embodied in the law and those embodied in ethical custom, In saying this, Friedman (1970) meant that the usual economic standpoint only acknowledges legal, ethical and economic responsibility as a crucial principle while taking part in altruistic activities do not yield incentives for corporations, (Carroll 1991). Crowther and Aras (2008) defines CSR in two different ways, they state that the broadest definition of CSR is concerned with what is or should be the relationship between global corporations, governments of countries and individual citizens. Crowther and Aras (2008) further state that more locally the definition is concerned with the relationship between a corporation and the local society in which it resides or operates. These two definitions by Crowther and Aras (2008) are pertinent and each represents a dimension of the issue of CSR. From the above definitions it is clear that, the exact nature of CSR is understood in various ways, with differences evident in the comprehending or presentation of the concept. Corporate social responsibility means something, but not always the same thing to everybody, (Vataw and Seth, 2003). Different scholars view CSR differently but one thing is clear, they all agree that companies must be concerned about the wellbeing of the society. Although governments, because of their nature, earn more than businesses, if a company becomes socially responsible it will increase competitiveness and enhance its image This paper however deems the definition by Watts (2004) to be applicable in the study. Watts (2004) states that CSR is the continuing commitment by business to behave ethically and contribute to economic development while improving the quality of life of the workforce and their families as well as the society at large. Businesses must not only be in business for the business but must also be in business for the community in which it operates, which means it must be concerned about the welfare and wellbeing of the society.

156 Journal of Applied Business and Economics vol. 16(3) 2014 Perspectives on CSR According to Clarke (1998) and Lantos (2001) the role of a business in a society can be viewed in two perspectives which are; the “classical view”, shareholders and the “stakeholder view”, which is created on stakeholder theory, which says that companies have a social responsibility that requires them to consider the interests of all parties affected by their actions, (Rodrigues and Branco 2007).

The Classical View Lantos (2001) identified two perspectives in the classical view: the “pure profit-making view”; and the “constrained profit-making view”. The “pure profit-making view” is based on the perception that some degree of dishonesty is acceptable because business people have a lower set of moral standards than those in the rest of society. The major proponent of the “constrained profit-making view” is Friedman (1970), who states that the purpose of the company is to make profits for shareholders, he further states that the only responsibility of business is to use its resources to engage in activities designed to increase its profits so long as it stays within the rules of the game be companies should behave honestly: that is, they do not engage in deception and fraud, (Friedman 1970 in Rodrigues and Branco 2007). Friedman (1970) further urges that because managers are agents of the shareholders they have a responsibility to conduct business in accordance with their interest. With authors like Barry, (2002); Coelho et al., (2003); Henderson, (2005) Jensen, (2001); supporting this notion. Barry (2002) argues that companies can only engage in social responsibility activities the less competitive the markets in which they operate are, and that such engagement is a form of rent-seeking by managers. Therefore, Barry’s (2002) assessment of the stakeholder perspective is that it tries to make the business system operate like the political system. Other contemporary authors defend shareholder value maximization as the one objective function to all companies but are not necessarily against the social responsibility actions by companies (Jensen, 2001); Coelho et al., (2003).

Stakeholder View The stakeholder view is based on the stakeholder theory. The stakeholder theory is based on the notion that beyond shareholders there are several agents with an interest in the actions and decisions of companies, (Branco and Rodrigues, 2007). Freeman (2004) explains that not only the owners of a corporation have genuine concern about it but also groups of persons that might be affected or can possibly be affected by the corporate`s doing. Stakeholders are groups and individuals who benefit from or are harmed by, and whose rights are violated or respected by, corporate actions, (Freeman 2006). In addition to shareholders, stakeholders include creditors, employees, customers, suppliers, and the communities at large, the government, financial analyst, employee unions and investors, competitors (Freeman 2004, Baker 2004, Khoury et al 1999, and Hopkins 1998). Baker (2001) states that stakeholder groups have a right not to be treated as a means to some end, and therefore must participate in determining the future direction of the firm in which they have a stake. Lantos (2001) identified two perspectives in the stakeholder view. The “socially aware view” which postulates that businesses should be sensitive to potential harm of its actions on various stakeholder group and the “social activism” which postulates that business must use its vast resources for social good and therefore must be concerned about the well-being of the society, (Rodrigues and Branco 2007). Supporting this notion is Werhane (2007) who states that the goal of any company is or should be the flourishing of the company and its principal stakeholders. De Wit and Meyer (2002) state that the main features of the perspective are emphasis on responsibility over profitability, organizations are seen as joint-ventures, organizational purpose is to serve all parties involved, the measure of success is the satisfaction among stakeholders.” However the socially aware view has been criticized by Capron (2003) who stated that the problem with theory is that it does not take into consideration “mute” stakeholders (the natural environment) and “absent “stakeholders (the future generation). The classical view and the stakeholder view are in clear contradiction of each other. However it has been noted that the stakeholder view and particularly the social activism view, takes into consideration the fact that businesses must wary about the well-being of the society. Which implies that a business must not

Journal of Applied Business and Economics vol. 16(3) 2014 157 only be in business for its self but it must also be in business for the community or society in which it operates, which means it must be concerned about the welfare of the society. The socio activism view is in support of the concept of CSR.

COMPANY IMAGE

The success of an organization is depended on the image of the organization. A good company is desired by all organizations. Roger (2005) states that company image can also be referred to as corporate image. There are several definitions of company image, it can be thought of as a mental picture that springs up at the mention of a firm`s name (Stewart 2004). Whereas Beliu (2001) defines company image as the perception people have of a business when they hear a company name. The image is composed of an infinite variety of facts, events, personal histories, advertising and goals that work together to make an impression on the public. Hatch and Schultz (2003) give a more comprehensive definition of corporate image; they defined it as the views of the organization developed by its stakeholders; the outside world’s overall impression of the company, including views of the customer, shareholders, the media and the general public. Belieu (2001), states that a good corporate image backs up the corporate culture that has been established inside and outside firm. The importance of a good company image cannot be overlooked. Belieu (2001) went on to identify five benefits of good company image as follows; • A good image is an efficient marketing and promotional tool • Credibility and integrity comes with a good name • A good company image establishes trust, confidence, loyalty and superb client relationship. • Company image is instrumental in increasing business opportunities. • A good image can stand the test of time.

Barret (2005) supports the importance of good image and stressed that a good image does not just grow, active efforts must be employed in order to achieve this. Endorsing this notion is Hayward (2007) who asserts that organisations may actively attempt to shape the image by communications, brand selection and promotion, use of symbols, and by publicizing its actions. A corporate image is as stated by LaReau (2005) the sum total of impressions left on the company's many publics. Hayward (2007) however goes further, and states that the overall image is a composite of many thousands of impressions and facts. Hayward (2007) further identified six major elements of company image which are; the core business and financial performance of the company; the reputation and performance of its brands ("brand equity"); its reputation for innovation or technological powers, usually based on concrete events ; its policies toward its salaried employees and workers; its external relations with customers, stockholders, and the community, and the perceived trends in the markets in which it operates as seen by the public. The theory of the corporate image holds that, all things equal, a well-informed public will help a company achieve higher sales and profits, whereas a forgetful or poorly informed public may come to hold negative impressions about the company and may ultimately shift more of its patronage toward competitors, (Hudson 2004). The significance of a good company image cannot be disregarded this is because a positive image increases profitability in the long run. A good image establishes confidence, loyalty, trust, and stronger relations with stakeholders.

Enhancing Company Image Through CSR Fritz (2009) argues that organizations today are getting involved in CSR to enhance their company image. It is often argued that the reason why corporations engage in CSR is a certain level of self-interest, not considering if the act is strategically motivated by commercial reasons alone or whether it is also motivated by what might seem as an altruistic interest, (Moon 2001). Veradajan and Menon (2002) states that there are several objectives of CSR for corporate giving beyond altruism. Companies seek to enhance their image in order to create a positive reputation that may also relate to higher long-run organizational performance. Some of the marketing objectives of CSR are increasing visibility, enhancing corporate

158 Journal of Applied Business and Economics vol. 16(3) 2014 image and thawing negatives publicity. In support of this assertion is Bennet et al., (2006) who explains that the main advantage of CSR are improvement of company image, attracting media attention, altering attitudes and helping the company’s relationship with the government and impressing key decision makers. More CSR practices leads to an improved (or at least maintained) reputation, which causes the firm to continue to be a target of activism, the consequence of which is more commitment to CSR. From the point of view of the company, however, having a good reputation can be a “double edged sword” or at least a potential liability when facing activists who seek the public limelight, (Rhee and Haunschild 2006). Corporate social responsibility (CSR) has become an important focus of attention among companies. A McKinsey global survey shows that 76% of executives believe that corporate social responsibility contributes positively to long-term shareholder value, and 55% of executives agree that sustainability helps their companies build a strong reputation, (McKinsey 2010). The theory of firm claims that an organization's interest is to maximize its shareholders value. Observing CSR from this point, it can be said that it is an answer to the ever increasing competition in the environment coupled with excessive demands on executives from different stakeholder group, (McWilliams and Siegel, 2006: Menon and Menon, 2007). Gray et al (2005) asserts that including stakeholders in the business of a corporation and embarking on sustainability reporting can be seen as mechanisms by which the organizations satisfy and manipulate stakeholders. In other words, CSR reporting can be seen in the light of corporate image management, marketing and a public relations tool which corporations use in order to show that they perform some sort of CSR activity. These tools are used adequately in order to foster a healthy competitive status by passing on information created to preserve an excellent image (Adkins, 2004; Darby, 2009). As such, many organizations pay particular attention to the image the public sees of them because it helps them do business effectively, anything that affects their image can possibly hinder their sales and even affect their licenses or funding, (Reich, 2008) Khanifar (2012) states the one of the benefits CSR is reputable for is its ability to enchase, if not build a business’s image and reputation. Concurring with this notion is Barney (2010) who states that, firms seek to enhance their public image to gain more customers, better employees, access to money markets, and other benefit. Porter and Kremer (2006) also supports the assertion by stating that reputation is used by many companies to justify CSR initiatives on the grounds that they will improve a company’s image, strengthen its brand, enliven morale, and even raise the value of its stock. Since the public considers social goals to be important, business can create a favorable public image by pursuing social goals, a poor social responsibility image can lead to sell outs of company shares by large investment funds, which can in turn negatively impact financial performance, (Chatterji, 2006; Levine, and Toffel 2009). Katsoulakos and Katsoulacos (2007) upholds that in recent times, organizations have been taking steps to ensure that they are not only tops at what they do but are projected in a positive light, by striving to be socially responsible either by making sure they are ethical in their dealings or transparent in their accounts to their stakeholders. Embedding Corporate Social Responsibility activities into organizations, otherwise known as mainstreaming, is a step that a growing number of organizations are beginning to take in making sure that every area of their business operations is linked with CSR, (Portney, 2008). From the assessment by different scholars above, it is clear that they all acknowledge the fact that corporate social responsibility activities can indeed enhance the company’s image. They also concur on the fact that CSR can go beyond enhancing company image, and also bring other benefits to the organisation which will; all if summed up will; improve the overall performance of the organisation and also improve the bottom line of the organisation in all its areas of focus. Although Bhardwag (2001) and Black (2007) state that CSR on its own is not enough to create and maintain a positive image; it is however clear that these authors do not dispute the fact that CSR enhances company image. It can be noted that CSR improves company image. CSR actually portrays the image of the firm itself. It shows what the company has done to fulfill its corporate duty to ensure the firm is not only good in providing the service but also plays its, roles by contributing something to the community. The issue of an increase in costs from the researcher’s point of view may only be experienced in the short run

Journal of Applied Business and Economics vol. 16(3) 2014 159 however in the long run CSR activities are beneficial to the organization. Organizations are concerned with survival in the long run.

DEFINING CORPORATE PHILANTHROPY

Corporate philanthropy is a phenomenon which associates the business sector with the social sector. Philanthropy provides an opportunity for corporations to establish an ethical and moral mantra within the organization, (Gan, 2006; Madrigal and Boush, 2008). An organization is comprised of people who assume the responsibility of cultivating and maintaining a culture supportive of philanthropy and its range of objectives. Successful philanthropy – achieving the goal is as vital to an organization as the “core business” (Bruch and Walter, 2005). Philanthropic initiatives are complex and thus need to be developed, communicated, implemented, monitored, and lastly sustained, in order to guarantee its viability as a strategic tool, (Mullen 2000). According to Carroll (1979) corporate philanthropy is a subset of corporate social responsibility. Lee (2007) also gave evidence about the relationship between CSR and corporate philanthropy, Lee (2007) articulates that, corporate philanthropy is one component of corporate social responsibility, albeit an important, highly visible component. Dayton (2004) declares that as with ‘Corporate Social Responsibility, ‘corporate philanthropy’ is an umbrella term which encompasses a number of different values, interests, mind-sets and alternative approaches. These, in turn, are based on a variety of perceptions shaped by cultural, contextual and professional factors, (Carroll 2009). Carroll (2009) further articulates that corporate philanthropy is more discretionary or voluntary on the part of businesses even though there is always the societal expectation that businesses provide it. Schwartz (2003) pronounces that although the primary purpose of corporate philanthropy is altruistic, it can also generate positive ‘moral capital’ among communities and stakeholders beyond the company’s direct business relationships. It can also strengthen the motivation of employees by making them proud of their company, although there are several dimension of CSR; Carroll (1991) placed philanthropy at the top of the pyramid of CSR, as the most common and most practiced dimension of CSR. Corporate Philanthropy is said to be the first step in building a robust CSR program, (Porter in Visser 2010). Carroll (1991) defines philanthropy as encompassing those corporate actions that are in response to society`s expectations that businesses be good corporate citizens. Carroll (2000) further states that this includes actively engaging in acts or programs to promote human welfare or goodwill. Communities desire firms to contribute their money, facilities, and employee time to humanitarian programs or purposes, but they do not regard the firms as unethical if they do not provide the desired level. Whereas Visser (2010) concurred with Carroll by defining Philanthropy as intervening in the lives of others for their benefit not merely for own. This dimension involves active involvement in activities that promote human welfare and goodwill in other words; it refers to business contribution to society by making the local community a better place to live and addressing sound concerns and problems, organized around a central theme driven by a collectivist culture of economic, legal, and social purpose, (Lee 2007). Authors, (Dayton 2004, Gan 2006 and Halme & Laurila, 2009) also share the same view by defining philanthropy as a means by which public organizations externally exhibit corporate social responsibility – widely. Moreover, the term as simply put by Drucker in Visser (2010) who states that, philanthropic, is the love of his fellow men. Alternatively, Luo and Bhattacharya (2009) suggest, a “Friedman-esque view” of CSR as an acknowledgment of a more traditional economic or capitalistic perspective. To stay in business we have to make profit, to succeed in business, we have to share some of that profit for the public good. (Garvin, in Mescon and Tilson, 1987). Trots (2006) define corporate philanthropy as an active effort to promote human welfare in form of cash or non-cash related corporate donation. In support of this assertion is Schultz (2005) who postulates that although firms donate money and aids to charities, schools and individuals, it may be for philanthropic purposes or to portray a good image to consumers. According to WBCSD (2000), corporations engage in philanthropic activities because it is easy and very public relations friendly,

160 Journal of Applied Business and Economics vol. 16(3) 2014 corporate giving is more easily dismissed as a public relations exercise than other forms of CSR. In an effort to respond to this criticism companies are shifting to making larger donations to a smaller number of charity 'partners' and combining giving with other activities. Lerner and Frywell (2005) also gave their own definition of corporate philanthropy, they define it as an activity above and beyond what is required of an organization and which can have a significant impact on the communities in which the company operates It is important to assist voluntarily those projects that enhance a community`s quality of life. CSR is sustainable in that CSR actions become part and parcel of the way in which a company carries out its business, whereas philanthropy is whimsical, It simply depends on the whims of the company directors at a particular time (Hopkins 2004). Johnson (2008) supports Hopkins (2004) in his view by saying that, companies should abandon all philanthropy which is outside of a CSR framework, thus companies should work hand-in-hand with governments to promote economic and social development. Johnson (2008) further states that government should help those people who cannot be helped to help themselves through a subsidy. Government should look after vulnerable groups and not just await the whim of corporate philanthropy: if a charity fails because a company fails then this is a disaster for all the vulnerable groups and people concerned, (Johnson 2008). This therefore means that, a company that is philanthropically generous but is not aware of, or engaged in, its broader CSR role will not be in business for very long. Porter (2007) fully supports this assertion by stating that if companies are just being good and donating a lot of money to social initiatives then they will be wasting shareholders' money. That is not sustainable in the long-run, and shareholders will quickly lose interest. Scholars (Sharfman 2008, Bremner 2007 Gladden 2005, Sternberg 1979) are against the concept of corporate philanthropy .The question of whether companies should engage at all in charitable giving has long been the subject of heated debate, (Sharfman 2008). Sternberg (1979) urges that business are owned by shareholders, any money spend on so called social responsibility is effectively theft from those shareholders who can after all, decide for themselves if they want to give to charity, she further states that it deprives shareholders of their property rights. In support of this view is Friedman (1970) who postulates that the sole responsibility of a business is to make profits for its shareholders. Debates, about the legitimacy of corporate philanthropy, have led to the rejection of the notion of corporate donations as being tainted or defiled (Bremner 2007, Gladden 2005). Sharfman (2004) claims that “it was immoral for companies to give away stockholders’ money; increasing scrutiny of corporate activities by journalists as well as the federal government; and a proliferation of charitable organizations, which made it increasingly difficult for companies to ascertain criteria for donations or to choose between solicitors”. Social issues are the responsibility of the politicians to deal with it all; it’s not the role of corporates, (Sternberg 1979). Barnett and Salomon (2006) questioned the role of corporate philanthropy and they stated that corporate philanthropy does not improve company performance but rather reduces it by increasing costs. CSR efforts such as corporate philanthropy are merely tools for public relations or legitimization, (Chen et al 2008). One of the most powerful arguments against corporate philanthropy is based on claims about the rights of property owners. It is argued that, as owners of the corporation, shareholders are entitled to the full value of their investment. Donations by public companies amount to a kind of 'expropriation' of shareholders' property and are therefore morally objectionable, (Wren 2003). The discussion on corporate philanthropy is embedded in a broader phenomenon of Corporate Social Responsibility (CSR), which in itself is a hotly debated issue. Here some argue that corporations should steer clear of social issues altogether (Friedman1970); others criticize CSR efforts such as corporate philanthropy as being merely tools for public relations or legitimization (Chen et al 2008). Friedman in Friedman (2008) completes the argument by stating that, when management spend money on matters which does not maximize profit, Friedman contended, they should rightly return the money to investors. From the arguments above it is clear that they are different views on corporate philanthropy. Those who support the notion (Dayton 2004, Gan 2006, Halme & Laurila, 2009, Visser 2010, Carroll 1991, Lee 2007, Trots 2006, Lerner and Frywell 2005 and Johnson 2008) summarily concur that organization must assist the society and assist in improving their well fare; this can be done through donations, in cash or kind, offering bursaries and infrastructure development. Some authors (Sharfman 2008, Bremner 2007

Journal of Applied Business and Economics vol. 16(3) 2014 161 Gladden 2005, Sternberg 1979, Friedman 1970, Wren 2003 and Chen et al 2008) are clear in their assertion that corporates must not be involved in these issues as they are not their responsibility and also that corporate philanthropy might be detrimental to the organisation’s performance. They further argue that corporate philanthropy was essentially an “agency cost,” which may bring benefits to individual executives and managers by improving their personal reputations or opportunities for advancement, (Galaskiewicz 2007), but this ultimately comes at the cost of shareholder wealth (Brown 2005;Helland and Smith 2006). Arguments cited in the literature are on the issue that corporate philanthropy is not the responsibility of corporates but that of the government and also that it increases costs for the organization It has been noted that the concept of corporate philanthropy cannot be overlooked. This is because getting involved in corporate philanthropy can improve the organisation’s chances of not only attracting but also retaining employees, customers, partners and investors, while also making a difference in the community an organisation operates in. Corporate philanthropy also enhances the social wellbeing of the community, Corporations wants to operate in well-developed communities in order for them to be able to function properly and to achieve the corporation’s objectives.

Types of Corporate Philanthropy Social responsibility is the newest of the three dimensions of corporate social responsibility and it is getting more attention than it has previously had, (Lee 2007). Many organizations are becoming increasingly active in addressing social concerns social responsibility means being accountable for the social effects the company has on people even indirectly, ( Visser 2010). They are different types of corporate philanthropy activities which organisations might get involved in. Corporates giving range from employee engagement programs, financial support for local causes and capital campaign donations by businesses to the society, (Manson 2002). Visser (2009) also identified several activities which include community grants, employee matching gifts, corporate sponsorship and non-cash contributions. According to Garriga and Melé (2004) Corporate philanthropy, includes direct cash giving, foundation grants, stock donations, employee time, product donations, and other gifts in kind. Hussein and Hussein (2005) also state the variety of corporate philanthropy an organisation an organisation can undertake they state that any companies become involved in community causes, for example by providing additional vocational training places, recruiting socially excluded people, sponsoring local sports and cultural events, and through partnerships with communities or donations to charitable activities. Brush and Walter (2005) took a different approach in explaining the elements of corporate philanthropy, they state that there are four fundamental elements that build and enhance a company’s image and role as a good corporate citizen. They term them the four business philanthropy elements; these are: engage, educate, empower and enrich. There are several activities which a company may carry out in line with corporate philanthropy and Carroll (1991) presents a more detailed outline of philanthropic activities. Philanthropic components or responsibilities as postulated by Carroll (1991) are as follows; • It is important to perform in a manner consistent with the philanthropic and charitable expectations of the society. • It is important to assist the fine and performing arts. • It is important that employees and management participate in voluntary and charitable activities with their local community. • It is important to provide assistance to private and public educational institutions • It is important to assist voluntarily those projects that enhance a community`s quality of life.

It is therefore clear that corporate philanthropy is one of the most diverse dimensions of CSR. These different types of corporate philanthropy give corporates a choice to choose that elements which is most effective not only for their organization but also for the stakeholders.

162 Journal of Applied Business and Economics vol. 16(3) 2014 Fostering Relations with the Community Through Corporate Philanthropy Philanthropy, in a business perspective, is through the lens of the social sector (Collins, 2009). Alternatively, according to Gan (2006) “Philanthropy, by its definition and in its early forms, assumes a certain degree of altruism and magnanimity”. This often is referred to as “generosity of spirit” which creates a dichotomy for corporations today. Harley (2001) suggests that corporate philanthropy by its very definition creates the sense of social responsibility with no strings attached. Firms utilizing philanthropic initiatives as part of an overall market development strategy must not look for an absolute monetary return, but to a certain extent a balance of returns comprised of social, ethical, and financial measures (Davis, 2003; Lockett, Moon, & Visser, 2006). Berger, Cunningham, and Drumwright (2007) furthered this notion and professed, that corporate philanthropy does appear to make business sense for some, but not all companies, notwithstanding, firms can use philanthropy as a means to an end through an ethical, enterprise-wide, and cogent focus. Hopkins (2008) articulates that corporations seemingly have a duty to align themselves with philanthropic causes in a strategic investing behavior – with an eye on charitable good and the hope (or intent) of some business return. Burch and Walter (2005) reported two distinct categories of corporate philanthropy. “Marketing orientation” represents the external strategies and tactics employed and readily focuses on the customer and other stakeholders who place demands and expectations on the firm. Alternatively, “competence orientation” suggests the need for internal strategies and assessments to ensure “alignment of corporate philanthropic initiatives with their companies’ abilities and core competencies”, (Burch and Walter 2005). Barrel (2006) asserts that corporate philanthropy can provide competitive advantage when they are well designed Mirvis (2008) further supports this perception by stating that, charitable contributions can increase name recognition, support by consumers and also the creation of a reputation and an improved image which will all result in the success of the organisation in the long run. Brammer and Millington (2005) states that corporate support of local causes improves the quality of life in communities where the company operates and that contribution help managers to build relationships with the community leaders and also the government and this can reduce regulatory and special group obstacles. Barker (2011) states that businesses engaging in community relations or community involvement typically conduct outreach to the community aiming to prevent or solve problems, foster social partnerships, and generally contribute to the community quality of life. They also participate in community relations to help improve their business by getting valuable community and other stakeholder input, (Baker 2011). Levy (1978) also supports Barker (2011), he declares that businesses have relationships in their local communities, sharing common interests, as such; it is valuable to spend some time considering how to leverage relationships on mutually beneficial initiatives. It is possible to enhance business performance, profitability and your reputation through your corporate philanthropy effort. Godfrey (2005) suggested that corporate philanthropy creates moral capital and act as insurance for the firm by building a strong relationship with stakeholder groups.

Matching Corporate Philanthropy Activities with Expectations of Stakeholders A key priority for a socially responsible business is to develop and maintain strong and mutually beneficial relationships with stakeholders. According to Wilkerson (2001) consulting and engaging key stakeholders is key to success. Whereas Donaldson and Preston (2005) state that when corporations manage their stakeholders accordingly their performance will improve tremendously in relation to their stability, growth, image and profitability. According to Bhattacharya (2011), if companies are to benefit from corporate philanthropy activities, it must understand how stakeholders think and feel about the programs. Bhattacharya (2011), further states that, if companies are to benefit from this initiative they must actively involve their stakeholders in decision making. According to Noir (2009), corporate philanthropy provides an opportunity to strengthen relationships between a company and its key stakeholders; he further urges that for corporate philanthropy activities to generate value for a company it must not only reinforce the company’s core values, but also fulfill some of the most basic needs of its stakeholders. Visser (2010) provided a linkage to stakeholder theory by noting the “natural fit between the idea of CSR and an organization’s stakeholders.” Furthermore, the concept of stakeholder personalizes

Journal of Applied Business and Economics vol. 16(3) 2014 163 social responsibilities by specifying groups or persons to whom companies are responsible and should be responsive. Clarkson (1995) holds that a stakeholder management framework is more useful to the analysis and evaluation of corporate social performance than models and methodologies based on concepts of social responsibilities and responsiveness. Clarkson (1995) further states that it is necessary to distinguish “between stakeholder issues and social issues because corporations and their managers manage relationships with their stakeholders and not with society. However, it is vital to understand that being responsive to stakeholders’ expectations implies the need to consider prevailing social norms and dominant views of corporate responsibilities. Stakeholders’ expectations of companies are intertwined inextricably with society’s views or expectations of business performance which evolve over time, (Hillman and Keim, 2009). Stakeholders are also part of the society; therefore what they expect is usually what the society expects. The basic framework is that the contribution of corporate philanthropy initiatives to stakeholder-company relationships hinges on the benefits they provide to the stakeholder. Bhattacharya et al, (2009) argue that for corporate philanthropy activities to provide returns the company initiative must provide a return to individual stakeholders. Bhattacharya et al, (2009) further states that research indicates that corporate philanthropy initiatives are successful in generating returns to the company to the extent that they foster strong and enduring relations with stakeholders. The key element of corporate social responsibility initiatives is societal alignment; Morris (1999) defines societal alignment as strategies and programs that meet society’s expectations. According to Kennedy (2009) over time, society’s expectations of businesses as responsible expectations of business responsibilities broaden as a society passes through the phases of economic development and its population increasingly seeks to meet not only physical but also social and personal-growth needs. Although expectations for corporate philanthropy differ from society to society it is important for organizations to make sure that their activities meet the expectations of the society and other stakeholders. According to Morris (2009) failure by an organization to carry out activities that meet expectations of the stakeholders will result in the activities to be disregarded. Morris (1999) further states that corporate philanthropy activities must also be channeled to the right beneficiaries otherwise the organization might suffer from image problems. A study by Porter and Kramer (2002) pointed out the importance of strategic assessment of philanthropic actions into the firm’s core capabilities in which philanthropy is an instrument to achieve the ultimate goal of profit maximization by improving the Competitive context of the firm, fostering collaboration and influencing local market features, their premise emphasizes competitive advantage through the alignment of philanthropic and business activities and stakeholder expectations. Morris (1999), one of the proponents of the societal alignment strategies states that, expectations of the society can be realized through pubic researches which will enable a company to know what the stakeholders expects from their corporate philanthropy activities. In support of the assertion of societal alignment are Willard (2006) states that aligning corporate philanthropy activities with the expectations of the stakeholders creates and improves relations between a company and its community. Lee (2006) affirms that all the four dimensions of CSR should meet the expectations of the society.

RESEARCH METHODOLOGY

Two research designs were used such that the weaknesses of one might be overcome by the other. The research designs used are Descriptive and Exploratory. Qualitative research was used to allow free response and qualitative research was used to avoid too much variation in responses.

Exploratory Research This research design gave a clear understanding of the subject and the researchers were able to clearly understand the problem. Through the use of exploratory research the researcher was able to gain more knowledge of the problem especially through the use of questionnaires and interviews. Personal interviews where the most effective in gaining an insight into the problem at hand, as it enabled UM

164 Journal of Applied Business and Economics vol. 16(3) 2014 management to give a more detailed explanation on the problems they have been facing as far as their image especially to the local community is concerned. Both personal and focus group interviews made it possible for the researcher to get the full corporation of the respondents, as it enabled a high degree of response. Exploratory research was also seen as the most suitable method since the researchers made use of secondary, secondary data involved analyzing internal reports, newspapers and also the internet.

Descriptive Research Descriptive research because it was seen suitable because of the use of questionnaires and interviews as data collection tools. Questionnaires made it easy for respondents to disclose information they deemed sensitive and where not able to express if interviews where to be used. Descriptive research also enabled the researchers to achieve the research objectives because it gave data on attitudes, opinions and perceptions. Attitudes opinions and perceptions of respondents where part of the objectives, respondents include UM employees and management were able to express their opinions and perceptions on the CSR and corporate philanthropy activities they are carrying out for the community and also if they perceive these activities to be beneficial both the UM and the community. Through descriptive research the community and Tongogara rural council were able to express how they perceive UM, their opinions and attitudes towards UM was also expressed.

Population and Sampling The target population breakdown was as 7842, the breakdown of the population is depicted in table 2 below:

TABLE 2 TARGET POPULATION FOR THE STUDY

Participants Target population Unki Mine management 31 Unki Mine employees 920 The community, (Chironde 3485 and Tongogara 3384) 6869 Tongogara rural council 21 Total 7842

A total of respondents 208 were used as the sample. Stratified sampling was used and it involved the division of the total population into different groups or segments or strata. The four stratums in which the population was classified are; UM management, UM employees, the local community and lastly Tongogara rural council employees were used. Systematic sampling was also applied in the strata whereby every 5th element was selected as part of the sample. The research also made use of convenience sampling. Convenience sampling was directed to the UM employees and management and Tongogara rural council employees. Here respondents were selected because of their convenient accessibility and proximity to the researchers. Convenience sampling is a useful tool in the exploratory phase of a research project, a phase in which ideas and insights are more important than scientific objectivity. The sample frame was made up of UM employees and management, local Chironde community and Tongogara rural council, in Shurugwi.

Research Instruments Interviews These are face to face meetings and are most versatile and flexible of all the available communication methods. Interviews were used because they are a fast way of getting information and also the researcher was able to collect confidential and honest information from the respondents. A combination of personal

Journal of Applied Business and Economics vol. 16(3) 2014 165 and focus group interviews were used. Personal interviews were conducted on 9 UM managers. Focus group interviews were used on the Chironde community. Focus group interviews made the respondents feel comfortable because they were able to express their views freely since they were in groups. From the 100 community respondents; 10 focus groups where created with each focus group being made up of 10 individuals.

Questionnaires The researchers found this research instrument appropriate since questionnaires are used when factual information is required or when opinion rather than facts are desired. Open and closed ended questions were asked. Closed ended questions were used when responses were known, the study made use of; dichotomous questions, multiple choice questions, rankings and likert scales as types of closed ended questions. Open ended questions allowed the respondents to give their views and opinions pertaining to the topic understudy. Questionnaire were used on UM employees and Tongogara rural council employees.

FINDINGS AND CONCLUSIONS

The research findings revealed that CSR activities enhance company image, as evidenced by 85.5% of the respondents who supported this notion. The study also revealed that corporate philanthropy activities enhance relations between a company and the community. The findings attributed good relations between a company and the community to corporate philanthropy. The study also showed that relations with the community results in a social license to operate as was reflected by 83% of the respondents who agreed with this view. Importance of matching corporate philanthropy activities with expectations of stakeholders was found to be important as it enables the organization to achieve its objectives.

CSR and Company Image After analyzing the responses it can be concluded that CSR activities have a positive impact on a company’s image. A positive image creates good relations with stakeholders, attracts investors, and the company will enjoy positive media reviews. The researchers can thereby conclude that CSR activities by UM have played a significant role on the positive company image they are currently enjoying. This was clearly supported by 85.5% of the respondents agreed that CSR activities enhance company image.

Role of Corporate Philanthropy in Fostering Relations Corporate philanthropy activities play a crucial role in fostering relations between a company and the community around its operations. 83% of the respondents agreed that corporate philanthropy activities have fostered relations between UM and the community of Chironde and Tongogara. From these research findings it can be concluded that corporate philanthropy activities go a long way in fostering relations between a company and the community. It is therefore concluded that businesses that take an active interest in the community’s wellbeing can generate support, loyalty and goodwill from the community.

Matching Corporate Philanthropy Activities Stakeholder Expectations From the research findings, It can be concluded that corporate philanthropy activities by UM match the expectations of the stakeholders. This was evident as a total of 89% of the respondents were of the view that the corporate philanthropy activities meet their expectations. UM engages their stakeholders through the community engagement development department through quarterly stakeholder engagement forum meetings. This means that the community and the Tongogara rural council as stakeholders have an input on the type of corporate philanthropy activity to purse. It can then be concluded that for the corporate philanthropy activities to be effective they must be designed with the input of stakeholders.

166 Journal of Applied Business and Economics vol. 16(3) 2014 RECOMMENDATIONS

The researchers recommend the following to UM in order for the company to maintain a positive image and to remain competitive in the market. UM could publicize their CSR activities more, so that they become known more to their stakeholders and also throughout the country. Through publicizing their activities throughout the country UM may be able to achieve a good company image from the general public. They can do this by introducing an external house journal whereby they update the public on their activities, facility visits and a company website which has an up to date coverage of their CSR activities . It is recommended that UM management involve employees in choosing the type of corporate philanthropy activity to implement. Employees are also stakeholders and some of them are part of the community, therefore their contributions are important. Involving employees in decision making will result in them taking ownership of not only the corporate philanthropy activities of the company but the company as a whole. UM could also put more investment on the other dimension of CSR as they are doing with corporate philanthropy. The other dimensions of CSR are just as important as corporate philanthropy. Coordinating all the four dimensions of CSR will indeed benefit the company. UM is also recommended to complete their projects within a reasonable period of time, as taking too long to finish the projects might end up tarnishing the image of the organization. The last recommendation to UM is that they can take their philanthropic activities beyond their community, thus they must also consider the wellbeing of other communities.

Further Study The research was limited to one dimension of CSR, corporate philanthropy. Further study can be carried out on the other dimensions of CSR and their role in enhancing company image.

REFERENCES

Aaker, D, A (1976), Essentials of Marketing Research, John Wiley and Sons, New York. Adkins, S. (1999) Cause Related Marketing: Who Cares Wins, Heinemann Butterworth, London. Aguilera, R., Rupp, D., Williams, C., Ganapathi, J. (2007) Putting Back CSR: A Multi-level Theory of Social Change in Organizations. Academy of Management Review. Alexander, G. and Bulcholz, R. (1978), “Corporate social responsibility and stock market performance”, Academy of Management Journal, Vol. 21. Barnett, M.L. (2005) Stakeholder influence capacity and the variability of financial returns to corporate social responsibility. Academy of Management Review, 32(3). Baron, D. (2007) ‘Corporate social responsibility and social entrepreneurship', Journal of Economics and Managerial Strategy, 16: 683-717. Baron, D.P. (2000) Business and its Environment, 3rd edition. Prentice Hall, New Jersey. Beesley M, and Evans, T. (1978) Corporate Social Responsibility. Croom Helm, London. Bowen, H.R. (1953), Social Responsibilities of the Businessman, Harper & Row, New York, NY. Branco, M.C. and Rodrigues, L.L. (2006) ‘Corporate Social Responsibility and Resource Based Perspectives', Journal of Business Ethics 69: 111-132. Bryman A. and Bell E. (2007) Business Research Methods, 2nd edition, Oxford: University. Burke, L. & Logsdon, J.M. (1996). How Corporate Social Responsibility Pays off. Long Range Planning. Campbell, J.L. (2007) ‘Why Would Corporations Behave In Socially Responsible Ways? An Institutional Theory Of Corporate Social Responsibility', Academy Of Management Review 32(3): 946-967. Carroll, A. (1991) ‘The Pyramid Of Corporate Social Responsibility: Toward The Moral Management Of Organizational Stakeholders', Business Horizons, 39-48. Carroll, A. B. (1979), “A Three-Dimensional Conceptual Model of Corporate Performance” The Academy of Management Review, Vol. 4, No. 4, October.

Journal of Applied Business and Economics vol. 16(3) 2014 167 Carroll, A. B. (2000) ‘A Commentary and an Overview of Key Questions on Corporate Social Performance Measurement', Business & Society 39(4): 466-478. Carroll, A.B. (1999) Corporate Social Responsibility. Business and Society 38(3) CECP. Clarkson, M.B.E. (1995), “A Stakeholder Framework For Analysing And Evaluating Social Corporate Performance”, Academy of Management Review, Vol. 20 No. 1. Cochran, P. and Wood, R. (1984), “Corporate Social Responsibility And Financial Performance”, Academy Of Management Journal, Vol. 27 No. 1. Collis, J. and Hussey, R. (2003) Business Research: A Practical Guide for Undergraduate and Postgraduate students, 2nd edition. Basingstoke: Palgrave MacMillan. Cooper, D.R., Emory, C. W. and Schindler, P.S. (2003) Business Research Methods, 7th edition. McGraw Hill. Crowther D. and Rayman-Bacchus L. (2004) Perspectives on Corporate Social Responsibility, Hants- England: Ashgate Publishing Company. Davis K. & Blomstrom R. L. (1975) Business and Society: Environment and Responsibility, McGraw- Hill, New York. Davis, K. (1973) ‘The case for and against Business Assumption of Social Responsibilities', Academy of Management Journal. Denzin, N.K. (1970) The Research Act: A Theoretical Introduction to Sociological Methods, Chicago: Aldine. Donaldson, T. and Preston, L.E. (1995) ‘The Stakeholder Theory of the Corporation: Concepts, Evidence, and Implications', The Academy of Management Review 20 (1): 65-91. Drucker P. F. (1968), Management Tasks, Responsibility and Practices, Harper and Row, New York. Ellen, P.M., Webb, D.J. and Mohr L.A. (2006). Building Corporate Associations: Consumer Attributions for Corporate Socially Responsible Programs. Journal of The Academy of Marketing Science. Emory, C and Cooper, D R. (1993) Business Research Methods. 5th Edition. Homewood. European Commission (2007) Corporate Social Responsibility National public policies in the European Union. Evan, W. M. and Freeman, R. E. (1988) A Stakeholder Theory of the Modern Corporation:. Fombrun, C. and Shanley, M. (1990) what’s in a name? Reputation Building and Corporate Strategy. Fombrun, C. J. (1996). Reputation: Realizing Value from The Corporate Image. Business School Press, Boston, Massachusetts. Frederick, W.C. et al. (1992), Business and Society, McGraw-Hill International, New York, NY. Freeman, R.E and Evan, W.M. (1990) ‘Corporate Governance: A stakeholder interpretation', Journal of Behavioural Economics. Freeman, R.E. (1984), Strategic Management: A stakeholder approach. Boston: Pitman. Friedman (1970) “The Social Responsibility of Business is to Increase its Profits. New York Times Magazine, (September 13). Friedman M. (1962), Capitalism and Freedom, University of Chicago Press, Chicago. Friedman, A.L. and Miles, S. (2006) Stakeholders, Theory and Practice, USA, Oxford University Press. Ghauri, P. and Gronhaug, K. (2005) Research Methods in Business Studies: A Practical Guide, 3rd edition. Harlow, Financial Times Prentice Hall. Godfrey, P.C. (2005). The Relationship Between Corporate Philanthropy and Shareholder Wealth: A Risk Management Perspective, Academy of Management Review. Graafland, J.J., Eijffinger, S.C.W., and SmidJohan, H. (2004) ‘Benchmarking of Corporate Social Responsibility: Methodological Problems and Robustness', Journal of Business Ethics. Gummesson, E. (1991) Qualitative Methods in Management Research, Newbury Park: Sage. Hayward, R (2005) Maintaining Corporate Image. Prentice Hall. Hill, I. (2006), “How Finance Can Help Move CSR up the Agenda”, The Cost and Management, ISSN- 1817-5090 Vol 34, No. 6, November – December. Hoffman, A.J. (2001) From Heresy to Dogma: An Institutional History of Corporate Environmentalism, Stanford Business Books.

168 Journal of Applied Business and Economics vol. 16(3) 2014 Hopkins, M. (1998) A Planetary bargain; Corporate Social Responsibility Comes Of Age. Macmill. Hossain, D.M., Salat A. and Amin A (2005),”Voluntary Disclosure on Corporate Social Responsibility: A Study on the Annual Reports of Bangladeshi Companies”, The Bangladesh Accountant, Vol. 47, No. 20 (April-June). Hussain B., and M. Hussain (2005), “Corporate Social Responsibility: Do Customers Get What They Expect?” Journal of Business Studies, Southeast University, ISSN- 1815- 3267 Vol 1 No. 1, July- December. Kabemba C (2012) Mining Companies Should Plough back into Communities. The Herald, (Harare) 8 May Katsoulakos, T. and Katsoulacos, Y. (2007) ‘Integrating Corporate Responsibility Principles; Corporate Social Responsibility For Irresponsibility. Working Paper, University Of California-Santa Barbara. Knox, S. & Maklan, A. (2004) Corporate Social Responsibility: Moving Beyond Investment Towards Measuring Outcomes. European Management Journal. Kotler, P (1998) Marketing Management; Analysis, Planning, Implementation and Control. Englewood Cliffs, Prentice Hall, NJ. Levitt, T. (1958) ‘The Dangers Of Social Responsibility', Harvard Business Review, September - October, 41- 50. Mahoney, L.S. and L. Thorne (2006), “An Examination of the Structure of Executive Compensation and Corporate Social Responsibility: A Canadian Investigation”, Journal of Business Ethics, ISSN- 1573-0697, Vol. 69, No. 2, December. Manakkalathil, J., and Rudolf, E. (1995) ‘Corporate social responsibility in a globalizing market', SAM Advanced Management Journal (Winter), 29-47. Marcus, A., Geffen, D. and Sexton, K. (2002): Corporate Self-Regulations and Multi stakeholder Dialogue, In E. Croci, (Ed), Handbook of Voluntary Environmental Agreements, Kluwer Academic Publisher, Netherlands. Martins J, H, and Louber M, (1998) Marketing Research. Unisa Press. Pretoria. McWilliams A. and Siegel D. (2001) ‘Corporate Social Responsibility: A Theory of the Firm Perspective', Academy of Management Review 26(1), 117-127. Mintberg, H. (1993) The Case Of Corporate Social Responsibility. Journal of Business Strategy. Available @ www.mintbrg.com. [Accessed 15 June, 2013] Mohr, L.A., D.J. Webb, and K.E. Harris (2001), “DO Consumers Expect Companies to be Socially Responsible? The Impact of Corporate Social Responsibility on Buying Behavior”, Journal of Consumers Affairs. Moon, J. (2001) ‘Business Social Responsibility: A Source of Social Capital?', Reason in Practice Vol 1(3), 35-45. O‟Brien, D. (2001). Integrating Corporate Social Responsibility with Competitive Strategy, The Center for Corporate Citizenship at Boston College, Boston, MA. Orlitzky, M., Schmidt, F.L., and Rynes, S.L. (2003). Corporate social and financial performance: A meta- analysis. Organization Studies. Oshagbemi P. (1983) Small Business Management in Nigeria, London: Macmillan. Paine, L.S., (2002): Value Shift, New York, McGraw- Hill. Peloza, J. (2005). Corporate Social Responsibility as Reputation Insurance, Paper presented at the 2nd Annual Corporate Social Performance Conference, Haas School of Business, University of California, Berkeley, April. Porter, M. E. and Van der Linde, C. (1995) ‘Toward a New Conception of the Environment Competitiveness Relationship', Journal of Economic Perspectives 9 (4). Porter, M.E. & Kramer, M.R. (2002). The Competitive Advantage of Corporate Philanthropy. Harvard Business Review. Reich, R. B. (1998) ‘The New Meaning Of Corporate Social Responsibility', California Management Review, 40(2): 8-17.

Journal of Applied Business and Economics vol. 16(3) 2014 169 Reynolds, P.D. (1971), A Primer In Theory Construction, New York: Macmillan Publishing. Robson, C. (1993) Real World Research: A Resource for Social Scientists and Practitioner - Researchers, Wiley Blackwell. Rollinson, D. (2002) Organisational Behaviour: An Integrated Approach, Harlow: Pearson Education. Saunders, M., Lewis, P. and Thornhill, A. (2000) Research Methods for Business Students, 2nd edition. Harlow: Financial Times Prentice hall. Sekaran, U. (1992) Research Methods for Business - A Skill Building Approach, 2nd Edition, New York: John Willey and Sons, Inc. Sen, A. (1999) Development as Freedom. UK: Oxford University Press. Shrivastava, P. (1995) ‘Ecocentric Management For A Risk Society', Academy Of Management Review, 20:118-137. Stahl, M.J. (1999) Perspectives in Total Quality, Oxford: Wiley-Blackwell. Stoner A. F. and Wankel C. (1988), Management, USA: Prentice Hall. Strauss, A., Corbin, J. (1990), Basics of Qualitative Research: Grounded Theory Procedures and Techniques, Sage, Newbury Park, CA. Swanson, D.L. (1995), “Addressing a Theoretical Problem by Reorienting the Corporate Social Performance Model”, The Academy of Management Review, Vol. 20, No. 1, January. Thorne, D.M.; Ferrell, O.C. and Ferrell, L. (2008) Business And Society: A Strategic Approach To Social Responsibility, 3rd edition. Boston, New York: Houghton Mifflin Company. Van Marrewijk, M (2003) ‘Concepts and Definitions of CSR and Corporate Sustainability: Between Agency and Communion', Journal of Business Ethics. Visser W. (2008) The A to Z of Corporate Social Responsibility. Available @ www.aynevisser.com [Accessed May 25 2013] Visser, W., Matten, D., Pohl, M., and Tolhurst, N. (2008) The A to Z of Corporate Social Responsibility: A Complete Reference Guide to Concepts, Codes and Organisations, Wiley. Vyakarnam, S. (1992), “Social Responsibility: What Leading Companies Do”, Long Range Planning, Vol. 25 No. 5. Waddock, S. A. and Graves, S. B. (1997) ‘The Corporate Social Performance-Financial Performance Link', Strategic Management Journal. Wartick, S. and Cochran, P. (1985), “The Evolution Of The Corporate Social Performance Model”, Academy of Management Review, Vol. 10 No. 4. Wolfe, R. and Aupperle, K. (1991) ‘Introduction to Corporate Social Performance: Methods for Evaluating an Elusive Construct', In J. E. Post (ed.), Research in Corporate Social Performance and Policy, Vol. 12. Wood, D. (1991), “Corporate Social Performance Revisited”, Academy of Management Review, Vol. 16 No. 4. Xueming, L. and Bhattacharya, C.B. (2008) ‘Corporate Social Responsibility, Customer Satisfaction and market value', Journal of Marketing, 70(4): 1-18. Yin, R.K. (1993) Case Study Research: Design and Methods, California: Sage. Zahra S.A., and M.S. LaTour, (1987), “Corporate Social Responsibility and Organizational Effectiveness: A Multivariate Approach”, The Journal of Business Ethics, ISSN- 1573-0697, Vol.6, no. 6, August. Zhuwakingu (2013) Making South African Mining Companies Take Social Responsibility Seriously. Mining weekly (online) available @ www.miningweekly.com Zikmund, W. (2002), Business Research Method, 7th edition South Western College, Ohio.

170 Journal of Applied Business and Economics vol. 16(3) 2014

Effect of Automated Teller Machine (ATM) on Demand for Money in Isolo Local Government Area of Lagos State, Nigeria

Fatai Abiola Sowunmi University of Ibadan, Ibadan, Nigeria

Zakariyah Olayiwola Amoo Lagos State University, Ojo, Lagos

Samuel Olasode Olaleye Lagos State University, Ojo, Lagos

Mudashiru Abiodun Salako Moshood Abiola Polytechnic, Abeokuta, Ogun State, Nigeria

INTRODUCTION

Automated Teller Machine (ATM) is a cash dispenser that enables bank customers to enjoy banking services without coming in-contact with bank tellers (cashier). ATM helps to perform the duties of the cashier in term of payment services. It is a computerised telecommunication device that provides the clients with access to financial transaction in a public place without the need for a cashier, human clerk or bank teller. ATM is also known as cash point or cash machine (Ogunsemore, 1992). ATM was introduced into the Nigerian economy by Central Bank of Nigeria (CBN) in 1989. Now defunct Societe General Bank of Nigeria (SGBN) was the first commercial bank in Nigeria to introduce ATM in 1990. The trade name for SGBN’s ATM was ‘‘cash point 24”. This was followed by “First Cash” ATM introduced by First Bank Plc. in 1991. The aim of CBN on the introduction of ATM was to reduce the rate of cash withdrawal from the counter through the use of tellers as well as to prepare Nigerians for the incoming cashless economy. However, as part of banking reforms that started in July 2004, the Central Bank of Nigeria (CBN) in its quest to improve bank services, achieve cashless economy and decongest the banking halls; mandated

Journal of Applied Business and Economics vol. 16(3) 2014 171 commercial banks operating in Nigeria to install Automated Teller Machines (ATMs) in the their premises and other strategic locations to serve their customers. As fallout on this directive by apex bank, studies (Fanawopo, 2006; Olatokun and Igbinedoin, 2009) revealed that the Nigeria’s debit card transactions rose by 93 percent between January 2005 and March 2006 over previous years owing to aggressive roll out initiatives by the Nigerian banks, powered by interswitch network. Specifically, the number of ATM transactions through the interswitch network also increased from 1, 065, 972 in 2004 to 14, 448, 615 between January 2005 and March 2006. This represents 92.6 per cent increase over the previous year. The number of ATMs has also grown from just over 500 ATMs in 2006 to over 8,000 ATMs in 2009. In Lagos State (the commercial nerve centre of Nigeria) alone under what is called ‘operation cashless Lagos’ there has been massive and aggressive deployment of ATMs in all the banks for customer’s use. Eni (2011), Enyimaya (2011) and Okwe (2011) projected that banks would have deployed about 40,000 ATMs in various parts of Lagos, 75,000 ATMs by June 2012 and 150,000 ATMs per 100 persons by December 2012. Past studies (Ogunsemor, 1992; Hone et al. 1998; Fasan, 2007; Akrani, 2011) affirmed that ATM has made it possible for bank customers to access cash at any time irrespective of bank business hours. This includes cash demands to meet day-to-day transactions and emergency needs by households on weekends, public holidays and during national strikes. Despite the various benefits of ATM, there are still customers who prefer to transact with withdrawal slip within the banking hall which have not made the “queue system” to come to a halt in banks. Since ATM came into use in its present form in 1974 in developed countries, the device has attracted widespread patronage offering wide range of services to the customers (Kolodinsky et al., 2004; Agboola, 2006; Bellis, 2007). Conversely, challenges associated with its use have made the machine unattractive and no-go areas for some banks’ customers (Adeloye, 2008; Obiano, 2009; Omonkhanlen, 2009). In Nigeria, the effect of ATM on demand for money has not been widely discussed in the literature. The study is aimed at highlighting this aspect and also to contribute to the existing literature on this topic. The purpose of this study is to determine the variations in the withdrawal frequency and cash demands (if any) that exist between the users and non-users of ATM. The novelty of the study is to determine the probability that a resident of the study area chosen at random uses ATM. The paper also tries to identify the factors influencing the probability of using ATM in the study area.

Conceptual Framework and Literature Review The study’s conceptual and theoretical framework is based on Keynes Liquidity Preference Theory which identified three primary reasons for holding on to cash; these are: 1. Transactions Motive: Money is a medium of exchange, and people hold money to buy goods. Money is demanded in order to fulfil day-to-day needs. As income rises, people have more transactions and people will hold more money. 2. Precautionary Motive: People hold money for emergencies (ill health, savings for unexpected job loss, labour strikes among others). Since this also depends on the amount of transactions people expect to make, money demand is again expected to rise with income. 3. Speculative Motive: Money is also a way for people to store wealth. Keynes assumed that people stored wealth with either money or bonds. When interest rates are high, rate would then be expected to fall and bond prices would be expected to rise. Bonds are more attractive than money when interest rates are high. When interest rates are low, they then would be expected to rise in the future and thus bond prices would be expected to fall. However, money is more attractive than bonds when interest rates are low.

The inventory theoretic model of demand for cash proposed by Baumol (1952) and Tobin (1956) opined that when deciding how frequently and, equivalently, how much to withdraw, consumers take two factors into account: the cost incurred per withdrawal (possibly including the opportunity cost of the time

172 Journal of Applied Business and Economics vol. 16(3) 2014 required per withdrawal) and forgone interest. Baumol and Tobin showed that the optimal withdrawal amount is proportional to the square root of the total value of transactions and indirectly proportional to the square root of interest rates. Stix (2003) revealed that the effect of ATM on cash demand depends on the user groups. On the one hand, if the proportion of people using ATMs frequently is high, ATMs have a negative effect on cash demand. On the other hand, if the proportion of active ATM users is low, ATMs do not affect cash demand. According to Attanasio et al. (2002), ATM transactions and cashless payments do not only affect optimal cash holdings but also are likely to reduce the time-cost per withdrawal. They reasoned that with ATM, customers withdraw cash more frequently and so hold smaller amounts of cash on average. The small cash demands by most ATM users are enough to address the transactionary and speculative motives of demand for money. Goodhart and Krueger (2001) found that the demand for small bank notes is positively related to the number of ATMs. People may visit ATMs more often and withdraw small amounts of cash, which would increase the demand for small bank notes. Despite the increase in the acceptance of ATM in Nigeria, Echekoba and Ezu (2012) observed that 68.2% of the respondent still complained about long queues in the bank, 28.9% complained of bad attitude of teller officers (cashiers) while 2.89% complained of long distance of bank locations to their home or work places. Generally, most studies (Hancock and Humphrey, 1998; Boeschoten, 1998; Goodhart and Krueger, 2001; Drehmann et al., 2002) showed that the effects of ATM on cash demand are not highly significant.

METHODOLOGY

The study was carried out in Isolo Local Government Area (LGA). Isolo LGA is one of the local government areas in Lagos state (Lagos state is the business nerve centre of Nigeria and former Federal Capital Territory). It is located between Amuwo-Odofin and Oshodi local government areas. Isolo Local Government Area is made up of the following towns, namely: Ilasamaja, Aswani, Ire-akari Estate, Okota, Isolo, Ajao Estate and Industrial Estate Isolo. The choice of Isolo local government area is based on its vibrant economic activities, the diverse population and location of many branches of commercial banks. The study adopted a two-stage sampling technique which comprises of purposive sampling and stratified random sampling. Using purposive sampling, only the residents of the seven towns that made up Isolo local government area operating commercial bank account(s) are considered. The study area was stratified based on the seven major towns and samples were selected randomly. A sample of 30 respondents which comprises of users and non-users of ATM were selected from each stratum. The study utilised primary data collected using structured questionnaire. A total of 210 questionnaires (the questions captured socioeconomic characteristics of respondents, the average amount collected through teller or ATM, frequency of collections, average time spent while collecting money among others) were administered. Out of these, 180 were returned to time by respondents. Data were analysed using descriptive, difference of means and probit analyses. Specifically, the difference of means was used to determine the variations in frequency of withdrawal; time spent withdrawing money and average amount withdrawn between the users and non-users of ATM. Since the sample size (number of respondents) used for the study is greater than 30, large sample formula was applied. The Z-calculated is obtained using the formulae below:

− µµ Z = US NUS σ (1) US −µµ NUS

Journal of Applied Business and Economics vol. 16(3) 2014 173 2 σσ 2 σ US += NUS US −µµ NUS (2) US NN NUS

Where: Z = N > 30 (large sample) N = numbers of respondents

µUS = Average parameter (amount withdrawn and time spent) for ATM card users

µNUS = Average parameter (amount withdrawn and time spent) for non-user of ATM card. σ US = standard deviation of the parameter for the ATM card users and non-users.

σ NUS = standard deviation of the parameter for the non-users US = users of ATM card NUS = non-users of ATM card.

The probability and the determinants that a resident in the study area uses ATM were estimated using probit analysis. The independent variables in the model are age (year), marital status, frequency of withdrawing money per week and average time spent during transaction (either through ATM or through teller). According to Spermann (2008) probit is based on a latent model:

* >== i i xyPxyP )|0()|1( ' xP εβ ii >+= x)|0( ' −>= ii βε xxP )|( ' −−= xF iβ )(1 (3)

Where: y = 1 represents ATM users y = 0 represents non-users of ATM xi represents the independent variables X1 represents age (year), X2 represents the marital status X3 represents frequency of cash withdrawal per week, X4 represents average time spent during transaction.

RESULTS AND DISCUSSION

The age distribution of the respondents shows that majority of the ATM users falls within the age bracket of 26 – 30 years. On the other hand, majority of the non-users of ATM card falls within the age bracket of 21 – 25years. These age brackets constitutes 33.3% and 45.5% of ATM users’ and non-users’ respondents respectively. The average age of the users of ATM (29.7years) is greater than the average age of non-users of ATM (27.8years). Among the ATM users, 59.2% are male while 40.8% are female while

174 Journal of Applied Business and Economics vol. 16(3) 2014 54.5% and 45.5% of non-users of ATM are male and female respectively. The study revealed that 62.41% of commercial bank customers in the study area use ATM. The bank officials posited that the congestion of customers in the banking hall has reduced significantly. Moreover, 42% of the ATM users possess Higher National Diploma (HND)/BSc certificates while 36.4% of the non-users of ATM are General Certificate of Education (GCE)/OND holders. Only 6.1% of ATM users possess postgraduate certificates (MSc/MBA/MA). Comparatively, majority of ATM users have better educational qualifications. Well educated respondents find it easy to use ATM unlike the less educated respondents. Also, 59.2% of the respondents who are users of ATM are married while 66.7% of non-users of ATM are married. This result indicates that there are more of single ATM users than the non-user of ATM. The study showed that 33.3% of ATM users are self-employed, 19.7% and 12.9% are civil servants and the unemployed respectively. Among the non-users of ATM, 45.5% are self-employed, 3.0% and 15.2% are civil servants and the unemployed respectively. The various problems encountered by the ATM users and the percentages of respondents that encountered each identified problem are shown in figure 1.0. The figure shows that 45.5% of the respondents ATM users had problem of no service displayed while 12.2% complained of ATM card seizure.

FIGURE 1 DISTRIBUTION OF PROBLEMS ENCOUNTERED BY ATM USERS

Expired Others, 5.8% card, 6.6%

No cash, 30.0%

No service, 45.4%

Seizure of card, 12.2%

Figure 2 shows that 36.1% and 1.4% of ATM users withdraw cash four and six times per week respectively while majority of the non-users (54.5%) withdraws cash twice per week.

Journal of Applied Business and Economics vol. 16(3) 2014 175 FIGURE 2 FREQUENCIES OF CASH WITHDRAWAL PER WEEK BY USERS AND NON-USERS OF ATM

60% 54.5%

50%

40% 36.1% 30.6% 30% 24.2% 25.2% Users of ATM 21.2% Non-users of ATM 20% of Respondents Percentage 10% 6.8%

0% 1.4% 0% 1 2 3 4 5 6 Frequency of Cash Withdrawal

The result of the inferential analysis showed that the average amount of cash withdrawn by ATM users is significant less than the average amount withdrawn by the non-users (p<0.05). This means that on average, small amount of money is withdrawn by the ATM users. Also, the average frequency of cash withdrawal per week by ATM users is significantly greater compared to the frequency of cash withdrawal by non-users of ATM (p<0.01). The results are in agreement with Attanasio et al. (2002) that with ATM, consumers withdraw cash more frequently and so hold smaller amounts of cash on average. Specifically, the average amount withdrawn by users and non-users of ATM are N10, 700.68 and N72, 106.06 respectively. The study revealed that ATM has made it possible for customers to demand for cash beyond normal banking hours, weekends, public holidays and during national strikes. Bearing in mind the small amount of money withdrawn at a time, ATM has enables customers to fulfil their immediate cash needs (transactionary motive) and unexpected expenses (precautionary motive) which are in agreement with Keynesian theory of demand for money. The security implications of withdrawing substantial amount of money and the pegging of the maximum amount that can be withdrawn through ATM CBN and unrestricted access to ATM anytime by respondents may be adduced for small cash demand by ATM users. Moreover, the result also revealed that the average time spent (minutes) to withdraw cash by ATM users is significantly less than the time spent by non-users of ATM (p<0.01). Specifically, the ATM users spent on average 7.91minutes compared to 16.95minutes spent by non-users of ATM in the banking hall. The implication of this result is that the use of the ATM saves customers from wasting precious time that can be put into productive use. The study showed that customers prefer using ATM in order to reduce time wasting which has helped to reduce traffic volume of customers in the banking hall and by extension it has helped to reduce insecurity in banks. The probit analysis result showed that the coefficients of three out of the five variables considered were significant at different levels of significance. The significant variables are the age (p<0.05), frequency of withdrawing money per week (p<0.05) and average time spent withdrawing money

176 Journal of Applied Business and Economics vol. 16(3) 2014 (p<0.10). From the foregoing, the age (year), frequency of withdrawing money per week and average time spent withdrawing money are the factors influencing the probability of using ATM in the study area (see table 1).

TABLE 1 PROBIT ANALYSIS RESULT

Parameter Coefficient std. error z Slope

Constant 1.08319 1.32249 0.8191

AGE 0.102412 0.0506119 2.023** 0.0174082 MARITAL STATUS - 0.503227 0.459306 - 1.096 ns - 0.0927277 GENDER - 0.377554 0.0388678 - 0.9840 ns - 0.0616770 FREQUENCY/WEEK 0.855367 0.367255 2.329** 0.187949 AVERAGE TIME - 0.355465 0.202945 - 1.752 * - 0.0604226 Source: Result of data analysed (2012) Note: * means significant at 10% ** means significant at 5% ns means not significant

Specifically, the result showed that an infinitesimal increase in age (year) of respondent selected at random from the study area raises the probability of using ATM card by 1.74% (0.017). That is, as one ages, the probability of using ATM increases. The precautionary and transactionary motives for demanding for money are higher among the married adults; most especially married couple with children. Moreover, an infinitesimal increase in the frequency of withdrawing money by a resident selected at random from the study area raises the probability of using ATM by 18.79% (0.1879) (see table 1). The result also posited that an infinitesimal increase in the average time spent while withdrawing money by a respondent selected at random, the probability of using ATM reduces by 6.04% (0.060). Furthermore, the study revealed that the probability that a person chosen at random from the study area will use ATM is 0.9215. That is, taking hundred residents of the area, it is very likely that 92 of the residents will use ATM.

CONCLUSION AND RECOMMENDATION

The study examined the effect of automated teller machine on demand for money in Isolo local government area. The study showed that ATM has significantly increased the frequency of demand for money when compared with the non-users of ATM. However, the average volume of money withdrawn through ATM is significantly less than the amount withdrawn through teller or cheque. The study showed that the ability of customers to meet their emergency need for cash is improved through the use of ATM. This is because apart from the normal working hours, customers have access to the use of cash during weekends and during national strikes. The study found that the use of ATM did not only help to reduce significantly long queues in the banking hall but also reduced the average time spent while withdrawing cash. Out of the various problems encountered, majority of ATM users (45.5%) complained of ‘no service’ displayed by ATM due to technical fault or power outage. From the foregoing, it is recommended that stakeholder should ensure improved performance of their Automated Teller Machine by making sure

Journal of Applied Business and Economics vol. 16(3) 2014 177 that they invest in quality ATM. While the stakeholder has a role to play in stable power supply not only for the machine but the environment also, the bulk of problem of stable power supply in the countries lies with the federal government.

REFERENCES

Agboola, A. (2006). Information and communication technology (ICT) in banking operations in Nigeria: An evaluation of recent experiences. Available on http://www.iisit.org/Vol6/IISITv6p373- 393Olatokun631.pdf Akrani, G. (2011). Automated Teller Machine ATM – Advantages of ATM. Available on, http://www.kalyan-city.blogspot.com/2011/02/automated-tellermachine-atm-advantages.html Attanasio, O., Guiso, L. and Jappelli, T. (2002). The demand for money, financial innovation, and the welfare cost of inflation: An analysis with household data. Journal of Political Economy, 110, (2), 317–351. Baumol, W. J. (1952). The transaction demand for cash: an inventory theoretic approach. Quarterly Journal of Economics, 66, 545—556. Bellis, M. (2007). The ATM machine of Luther Simjian, Available on http://www. inventors.about.com/od/astartinventions/a/atm.htm Boeschoten, W. (1998). Cash management, payment patterns and the demand for money. The Economist, 146, (1), 117–142. Central Bank of Nigeria (2012). Monetary policy reform 2006 – 2011. Available on http://www.cenbank.org/monetarypolicy/reform.asp) Drehmann, M. Goodhart, C. and Krueger, M. (2002). The challenges facing currency usage: will the traditional transaction medium be able to resist competition from the new technologies? Economic Policy, 17, (34), 195–227. Echekoba, F. N. and Ezu, G. K. (2012). Electronic retail payment systems: User acceptability & payment problems in Nigeria. Arabian Journal of Business & Management Review, 5, 60 – 63. Encyclopaedia, the free dictionary.com (2011). Meaning of ATM. Available on http://www.thefreedictionary.com/automated) Eni, H. (2011). Cashless Lagos: Will it work. TELL, 49, 40-42. Enyinnaya, C. (2011). Where Sanusi missed cashless banking. Available on, http://www.ngrguradiannews.com Fanawopo, S. (2006). World without cash-Nigeria’s payment card grows significantly. Available on http://www. iisit.org/Vol6/IISITv6p373-393Olatokun631.pdf Fasan, R. (2007). Banks, customer relation and use of ATM cards. Business Day Newspapers, Available on http://www.businessdayonline.com/ Goodhart, C. and Krueger, M. (2001). The impact of technology on cash usage. Discussion Paper 374, LSE Financial Markets Group, Discussion Paper Series. Hancock, D. and Humphrey, D. (1998). Payment transactions, instruments, and systems: A survey. Journal of Banking and Finance, 21, 1573–1624. Hone, K. S, Graham, R. Maguire, M.C., Baber, C. and Johnson, G. I. (1998). Speech technology for automatic teller machines: An investigation of user attitude and performance. Ergonomics, 41, (7), 962-981. Isolo Local Government Area (LGA) – Report on the development of Isolo LGA (Oct. 7th 2011). Kolodinsky, J. M., Hogarth, J. M. and Hilgert, M. A. (2004). The adoption of electronic banking technologies by US consumers. International Journal of Bank Marketing, 22, (4 and 5), 238-259. Obiano, W. (June 21St, 2009). How to fight ATM fraud. Nigerian Daily News: 18. Ogunsemor, A. O. (1992). Banking services: the emergence and impact of electronic banking. The Nigerian Banker, Jan-March. Okwe, M. (2011). The threat of bank-less Nigeria.” Available on http://www.ngrguradiannews.com

178 Journal of Applied Business and Economics vol. 16(3) 2014 Olatokun, W. M. and Igbinedoin, L. J. (2009). The adoption of automated teller machines in nigeria: an application of the theory of diffusion of innovation. Issues in Informing Science and Information Technology, 6, 373-393. Omonkhanlen, O, (2009). ATM fraud rises: Nigerians groan in Nigeria. Daily News June, l, (21), 8-10. Snellman, H. (2006). Automated Teller Machine network market structure and cash usage. Scientific monographs. E: 38. Sperman, A. (2008). The Probit model. University of Freiburg, SS 2008. Available on www.zew.de Stix, H. (2004). The impact of ATM transactions and cashless payments on cash demand in Austria. Monetary Policy and the Economy, Q1/04: 90-105. Tobin, J., 1956. The Interest-Elasticity of transactions demand for cash. Review of Economics and Statistics, 38, (3), 241—247.

APPENDICES

APPENDIX 1 VARIATION IN AVERAGE AMOUNT OF MONEY (N) WITHDRAWN BY ATM USER AND NON-USERS

Amount Withdrawn Parameter Non-User User Mean 72106.06061 10700.68027 Variance 29020043087 56194040.63 Observations 33 147 Hypothesized Mean Difference 0 Df 178 Z Stat 2.070236665 P(T<=t) one-tail 0.023290226 t Critical one-tail 1.693888703 P(T<=t) two-tail 0.046580453 t Critical two-tail 2.036933334 Source: Result of data analysed (2012)

Journal of Applied Business and Economics vol. 16(3) 2014 179 APPENDIX 2 VARIATION IN AVERAGE TIME SPENT BY ATM USERS AND NON-USERS

Average time spent (min.) Parameter Non-User User Mean 16.95454545 7.911564626 Variance 59.56818182 5.317468083 Observations 33 147 Hypothesized Mean Difference 0 Df 178 ZStat 6.664275007 P(T<=t) one-tail 6.92321E-08 t Critical one-tail 1.962360258 P(T<=t) two-tail 1.38464E-07 t Critical two-tail 2.034515287 Source: Result of data analysed (2012)

APPENDIX 3 VARIATION IN FREQUENCY OF WITHDRAWAL PER WEEK BY ATM USERS AND NON-USERS

Frequency of withdrawal per week

Parameter Non-users Users

Mean 2.03030303 3.836734694

Variance 0.46780303 0.863572826

Observations 33 147

Hypothesized Mean Difference 0

Df 62

ZStat -12.7573066

P(T<=t) one-tail 2.7041E-19

t Critical one-tail 1.669804163

P(T<=t) two-tail 5.4082E-19

t Critical two-tail 1.998971517

180 Journal of Applied Business and Economics vol. 16(3) 2014

Optimizing Patient Flow and Resource Utilization in Out Patient Clinic: A Comparative Study of Nkawie Government Hospital and Aniwaa Health Center

John Mensah Kwame Nkrumah University of Science and Technology

David Asamoah Kwame Nkrumah University of Science and Technology

Akua Amponsaa Tawiah Kwame Nkrumah University of Science and Technology

INTRODUCTION

Out Patient Department (OPD) Services are one of the important aspect of Hospital Administration. OPD is the mirror of the hospital, which reflects the performance of the hospital being the first point of contact between the patient and the hospital staff. Ensuring high standard for medical treatment as well as other auxilliary services will be essential in preserving the efficacy of the current first line treatment for the patients who visit the various OPD centers in Ghanaian hospitals. Satisfaction refers to a state of pleasure or contentment with an action, event or service, especially one that was previously desired (Hornsby and Crouther, 2000). When applied to medical care, patient satisfaction can be considered in the context of patient’s appraisal of their desires and expectations of health care. One of the factors that influence patient satisfaction is efficiency of services rendered to patients (Santillan, 2000). The “ efficiency” of service refers to the promptness of the care given to patients, including issues like waiting time before consultation, duration of consultation, amount of time spent with the doctor, quick response to emergencies, quick dispensation of drugs, fast and accurate laboratory tests (Santillan, 2000). Out-patient healthcare service as the term connotes is simply the provision of healthcare to patients at hospitals, polyclinics, doctor’s offices etc where they are treated and released on the same day. Out- patient service has stood the test of time as a vital component in the delivery of healthcare in the Ashanti

Journal of Applied Business and Economics vol. 16(3) 2014 181 Region. It registered one of the strongest growth areas in the industry. Average out-patient visits per member per year were between 1.40 and 1.50 in 2009 for Ashanti Region alone, as against a national average of 0.18 (Ministry of Health 2010). Among the many challenges confronting the healthcare sector in a developing nation such as Ghana, out-patient healthcare service is in no way immune and has its own fair share of the challenges. Notwithstanding the successes and improvements chalked by the out-patient healthcare service in the nation, this part of the service continues to grapple with concerns such as inadequate funding, lack of qualified professionals and medical equipments, etc. Furthermore, there is still the wave of overcrowding due to longer waiting times in out-patient clinics and this has been a source of concern for health workers and policy makers (Coté, 1999). In Ghanaian healthcare, queues are characterized structures formed to maintain order and create a hold on time; capital and human contribution towards development and efficient performance of any hospital system usually rely on queues to coordinate the delivery of health care (Ministry of Health 2010). Even though healthcare delivery systems in private hospitals are perceived to be better than that of government hospitals, majority of Ghanaians seem to rely often on government hospitals for healthcare services. This could probably be due to the higher cost of assessing private healthcare facilities, which in turn may be attributed to the low income level of many Ghanaians. The large number of patients patronizing government hospitals often results in overcrowding and longer waiting times in out-patient clinics. Longer waiting times have been reported to affect patient satisfaction and choice of hospitals (Nketiah-Amponsah and Hiemenz, 2009). Because of the growth in out-patient health care, the operational and planning activities associated with its delivery have been quite sensitive. Consequently, there is a recognized need for more research in this area. To the best of the researchers' knowledge, there is currently no publication on optimizing patients flow in order to reduce waiting times in public hospitals in Ghana. Thus the main motivation for conducting this research was to come out with a suitable and appropriate system for the management of out-patient departments in various hospitals in Ghana. As a comparative study, the research sought to examine the patients' flow into consulting rooms at Nkawie General Hospital (government owned) and Aniwaa Medical Center (privately owned) and to determine the traffic intensity at both hospitals. Additionally, the average time spent by physicians on patients in consulting rooms at both hospitals would be estimated and finally, a system to reduce the length of waiting periods and average time spent in consulting rooms by patients in public hospitals would be proposed. The rest of the paper is organized as follows: Section 2 presents relevant literature review of Optimizing patient flow and resource utilization in outpatient clinics, to be followed by the research methodology in Section 3. Section 4 is a presentation of the results and discussion. Finally we conclude the paper by highlighting the findings, implications and potential recommendations in Sections 5 and 6.

LITERATURE REVIEW

Much healthcare research has been devoted to the occurrence of the length of time which patients have to wait in order to be given healthcare services (Appleby et al., 2005). After sometime, waiting lists were viewed (often implicitly) as arising from a “backlog” in the need for care (Harrison, 2000). From his earlier research (Gross, 1992), established that management scientists view waiting as arising from the vibrant buffering or smoothing of demand. Queuing theory, which has been far and widely applied in areas such as the design of call centers (Koole and Mandelbaum, 2002) but which also has a long-recognized applicability to the healthcare. It is thus promising to think of waiting as a “time price” as opposed to a “money price” (Gravelle et. al. 2002) Coté (1999) shows that similar to many hospital or in-patient applications, such as those described by Badri and Hollingsworth (1993), Blake et al. (1996), Cohen et al. (1980), and Hancock and Chan (1988), the movement and control of patients and the utilization of resources is also of paramount importance to the operations of outpatient clinics. Keller and Laughun (1973) formulated an objective function to study the effect of patient congestion on physician capacity. Cox et al. (1985) sought later to control patient

182 Journal of Applied Business and Economics vol. 16(3) 2014 flow in order to ``optimize'' an appointment system for an outpatient clinic's operations. Rising (1977) formalized the outcome of earlier research to study how to allocate patient arrivals such that an outpatient facility's workload was balanced. Stafford and Agarwal (1979) introduced a demand function for health services based on classical economic theory of the firm in order to predict daily patient load as a function of the calling population. With their recent research Bevan and Morton (2008) found that, the central idea of queuing theory is that variability in the waiting times arises from variability in the arrival process which is either unpredictable, or predictable but unmanageable (as happens in systems which experience “rush-hours”). According to Baurerle (2001), p represent system utilization and in queuing systems, it states that if p < 1, i.e. the number of arrivals is less than the number of potential departures and for that matter the system is stable (positive recurrent). If p > 1 the system is unstable (transient) and if p = 1 it is null recurrent. Simple and compound waiting times are implicated in an attempt to access treatment through queues from the perspective of the patient (David, 2005). However, various methods have been adopted to reduce queues to the barest minimum in some hospitals. This has led to several techniques employed by health care facilities to queuing and its characteristics on queuing systems, service or server efficiency, service space and service point(s) provided (Kolobe, 2006). According to Q-Matic (2009), Customer Flow Management process is central and it is only when this process has been fully understood, is it possible to design solutions that maximize the benefits for the service provider.

METHODOLOGY

Two main types of data will be used for this study, primary and secondary data. The primary data consisted of information gathered on long waiting times in the out-patients departments in both government and private hospitals in the country-Ghana. The selection of sample respondents from the two hospitals was based on their OPD (Out-patient Department) attendance. Available records from the two hospitals showed that, OPD attendance constitute an average of five hundred and eighty per day (580) for Nkawie Hospital and four hundred and eighty (480) for Aniwaa Medical Centre. Based on these figures, the sample size for the study was pegged at three hundred and twenty-four (324) which constitute sixty percent (60%) of the total average number of the OPD attendance and this, the researchers believe, is sufficient and adequate for the purpose of this study. Questionnaires administration, interviews as well as observation constituted the forms of data collection mechanisms. The data gathered was analyzed by applying the basic queuing theory formula for a single-phase, single-channel system to patients in both hospitals. Quantitative Methods (QM) software version 2.2 was also used for the analysis of the study.

RESULTS PRESENTATION, ANALYSIS AND DISCUSSION

The Flow of Patients into Consulting Rooms at Nkawie General Hospital and Aniwaa Medical Center Concerning the flow of patients at the Nkawie and Aniwaa hospitals, the total average length of cases for the month of the study was estimated to be 5040 and 8991 patients respectively for the two hospitals. The mean arrival rate (λ) was determined by dividing the number of actual cases by the number of hours the hospital operates in a day for a month and this was:

5040 Nkawie λ)( = =× 2330 (λ) = 5040/(12 hours per day x 30) = 23 (1) hours 12( hours per day)

8991 Aniwaa (λ) = =× 2530 8991/ (12 hours per day x 30) = 25 (2) (12 hours per day)

Journal of Applied Business and Economics vol. 16(3) 2014 183 Therefore, patients arrive at an average rate of 23 and 25 per hour at the Nkawie and Aniwaa hospitals respectively. Service rate (μ) was estimated using the average number of patients served per an hour period within which we took the sample cases. This was obtained after observing the patients right from the ‘records’’ section up to the time they obtain their drugs from the dispensary. Average time spent by the sample patient gave us the service rate for the two out patients departments under the discussion

The service rate (μ) at Nkawie General Hospital = 25 patients per hour (3)

The service rate (μ) at Aniwaa Medical Centre = 27 patients per hour (4)

From the data gathered the average time patients were:

1 Waiting in the entire OPD unit: Ws = − λ (5) µ 1 Waiting in the entire OPD unit of Aniwaa Hospital: Ws = − 25 27 = 0.50 1 Waiting in the entire OPD unit of Nkawie Hospital: Ws = − 23 25 = 0.50

From the data gathered, the waiting time in the entire OPD unit before the patient see the doctor is 0.5 of an hour or 30 minutes for both Aniwaa Medical Centre and Nkawie Government hospital. This shows that Aniwaa Medical Center has a better service rate than the Nkawie Government Hospital. This is due to the number of OPD cases they are able to serve within its operational time - 5040 patients for Nkawie Hospital and 8991 patients for Aniwaa Medical Center. The number of patient flow into each of the hospital has resulted in more waiting time for the patients at the OPD units. From the data gathered, waiting at the OPD units at both Aniwaa and Nkawie hospitals was estimated to be 30 minutes. These consist of the wait period at records, revenue and vital signs checking sections. This is however, quite a substantial time to be wasted in these hospitals. Concerning the Out-Patient Department (OPD) flow of patients at Nkawie and Aniwaa hospitals, patients arrive at an average rate of 23 and 25 per hour in the two hospitals respectively excluding the emergency wards and the patients who have appointments especially at Aniwaa Medical Centre where scheduling of appointments is practiced. The patients arriving at the OPD unit with Scheduled Appointments are not considered as part of the estimated waiting hours, even if they wait to see their medical doctors. Most of the cases at the out-patient unit were non-emergency cases and therefore we used Poisson distribution for arrival process in the study. After an extensive statistical analysis of the collected data, the service rates for the two hospitals were 25 and 27 patients per hour for Nkawie Hospital and Aniwaa Medical Centre respectively. Considering the time patients had to wait at the OPD unit, it was observed that both Nkawie hospital and Aniwaa Medical Centre had two consulting rooms each serving as their servers. This obviously created the long queues at the OPD units due to large number of arrival of patients. Both hospitals maintained only two servers because of the variability in the arrival process which is unpredictable or predictable yet unmanageable. These situations are consistent with the work of Morton and Bevan (2008) and Coté (1999).

The Traffic Intensity Within the OPD Unit at Nkawie and Aniwaa Hospital Traffic intensity is a measure of the average occupancy of a server or resource during a specified period of time, normally a busy hour. From the above discussion, the volume of traffic in the OPD was

184 Journal of Applied Business and Economics vol. 16(3) 2014 measured by the number of service requested per unit time and the time that a section in OPD satisfies each patient request and were recorded as follows (Table 1 and 2).

TABLE 1 WAITING LINE RESULTS FROM THE NKAWIE HOSPITAL

Parameter Value Parameter Value Minutes Single-channel system Average server utilization 0.9200 Arrival rate(lambda) 23 Average number in the queue(Lq) 10.5800 Service rate(mu) 25 Average number in the system(Ls) 11.5000 Average time in the queue(Wq) 0.4600 27.6000 Average time in the system(Ws) 0.5000 30.0000 Source: Field Survey, 2013

TABLE 2 WAITING LINE RESULTS FROM ANIWAA MEDICAL CENTRE

Parameter Value Parameter Value Minutes Single-channel system Average server utilization 0.9259259 Arrival rate(lambda) 25 Average number in the queue(Lq) 11.57407 Service rate(mu) 27 Average number in the system(Ls) 12.5 Average time in the queue(Wq) 0.462963 27.7778 Average time in the system(Ws) 0.5 30.0000 Source: Field Survey, 2012

The traffic intensity for both Nkawie and Aniwaa are all approaching one (1) indicating some amount of traffic in the system by the two hospitals. Figures recorded by the two hospitals-Nkawie and Aniwaa was 0.92. That is the average server utilization or the traffic intensity in the Public Hospital-Nkawie and that of the private Hospital-Aniwaa Medical Center in Ghana are all approaching null recurrent which is consistent to Coté (1999). This implied that the queue in OPD units will grow out of bound if nothing is done on the flow of the patients to these hospitals as well as the service rate of the hospitals. In other words, the figure 0.92 of both hospitals is approaching one (1) and if it is checked, very soon it will be more than one (1). This will increase the server traffic intensity, and operations at the units and planning activities at the OPD will all become complex and quite sensitive in its management (Coté, 1999). The ideal traffic intensity has to always fall well below one for the system to be stable or positive recurrent (Bauerle, 2001). This shows that there are a lot of inefficiencies in the operations at the OPD units in both hospitals that need immediate attention for smooth patients' flow. Again from the data gathered, the average number of patient found in the queue at a time at Nkawie Hospital is eleven (11) patients as compared to the Aniwaa Medical Centre which on the average is twelve (12) patients at a time. With this average number in the queue, each patient at Nkawie Hospital spent an average of 27 minutes in all the queues that he/she joins at the hospital and 28 minutes at the Aniwaa Medical Center. But considering the total number of patients attending these two hospitals, Aniwaa Medical Centre is more operationally efficient than the Nkawie Hospital due to the less relative average time used to serve its patients.

To Determine the Average Time Spent by Patients in a Consulting Room at Nkawie General Hospital and Aniwaa Medical Center The average time patients spent in the consulting room is

Journal of Applied Business and Economics vol. 16(3) 2014 185 11 At Nkawie the service time == (6) µ 25

µ = 25 therefore the average time spent in a consulting room

Lq 6.101 1 = =+=+ 5.0 = 0.5hours or 30 minutes. (7) µλ 23 25 11 At Aniwaa the service time = = (8) µ 27

µ = 27

Therefore the average time patients spent in the consulting room is

Lq 5.111 1 = =+=+ 49.0 or 30 (9) µλ 25 27

As at the time of the study, both hospitals had two (2) consulting rooms serving the entire patients. These consulting rooms are represented by m meaning the servers for the two hospitals. Therefore the Average server utilization for the two hospitals is 0.46. The figure also depicts the probability of how much a patient is likely to wait in the OPD Unit. In all, it was observed, from the data gathered that the average time of patients waiting in line to be served are 27 minutes and 30 minutes for Nkawie hospital and Aniwaa medical centre consulting rooms respectively. The times could be detrimental for patients who do not have a stable condition. Average time spent in the queue (Wq) is the proportion of the system's resources used by the traffic which arrives at the OPD unit. From the data gathered, the average time in the queue spent by patient in each of the section visited at the OPD unit at both Nkawie and Aniwaa hospitals was 0.46 hour. Therefore, a patient who visits records section, vital signs section, consulting room, and dispensary section will spend an average of 0.46 hour in each of these sections and the total time spent for these sections is 1.86 hours or 110.4 minutes. For the two hospitals, the value indicate the growth of queuing in the OPD operation system which is due to an increase in time the patients spent in the queue. Measures must be put in place to check these delays and provide efficient services. In general, a lower utilization corresponds to less queuing of patients and for that matter more efficient utilization of the system. Service times are also often assumed to be random. The mean service rate is mu, and the expected service time is E(s) = 1/mu. Hence, when the OPD system capacities were measured, it was found that as the mµ increases, the system utilization for a given Patients arrival rate decreases. This means that, if the capacity of the OPD units increases, then the various sections within the OPD unit will be able to reduce the time used to serve patients who used the facility. It was also realized from the study that there are increases in waiting times especially at Nkawie General Hospital as a result of a fall in service capacity. It therefore rests on the Hospital Administrators to consider the trade-off between OPD capacity increase and the service delay, implying that they should weigh the cost of providing a given level of service against the potential cost of having patients wait as a result of having low capacity for the service provided as suggested by Mazumdar (2007).

Developing a System to Reduce the Length of Waiting Periods of Patients From the observation made from the study, as a patient enters the hospital OPD, that patient has to either first join the queue leading to the records section or go straight to the records section when there is

186 Journal of Applied Business and Economics vol. 16(3) 2014 no queue. The record section personnel determines if the patient ever receives the service at the hospital or not and if the former is true, they pull up the patient’s card from their records and give to him/her. If the patient is visiting the hospital for the first time, the personnel at the records first creates patient’s profile in the Hospital Information Database system before the patient is handed his or her new card. From records, the checking of vital signs before going to the consulting room is done. Patient then visit the revenue section and then to the dispensary. These procedures were disclosed by the patients of both hospitals and as a result of these paths it always resulted in long queues within the OPD units which is also consistent with the work of Stafford and Aggarwal (1979) that offering homogeneous service at stations aggregated clinic queuing. This OPD queuing can be reduced or totally eliminated by the introduction of Q-matic system. Q- matic software is one of the systems that can be used to reduce the burden of control of manual processes and resources by creating a completely automated system (Thomas, 2009). Q-matic software had proved capable of reducing waiting times drastically to a tune of 70% of the waiting time. Another important function of the Q-matic system that was discovered in this research is its ability to track patients throughout their visit to the Hospital. With the Q-matic solution, a system administrator sitting in his office can monitor the patient's movement through the OPD clinic. It basically tracks patients from the moment they enter till the time they leave using a single system and a single ticket. Again, the system is able to generate live management information which will enable the Hospital Administrators to take immediate measures if necessary. Administrators can see patient volumes at all times, real and calculated future waiting times for OPD, and consulting rooms. In addition, Administrators can see, in real time, any appointment, delay times as and when they happen. As soon as waiting times exceed the set limits, Administrators are alerted by the system and can act accordingly (Thomas 2009).

CONCLUSION

This study attempted to analyze actual operations of the two hospitals, Nkawie Government Hospital and Aniwaa Medical Centre which is fully private owned. From the study it was found out that Nkawie Government Hospital has longer waiting times compared to Aniwaa Medical Center. Even though the figures are the same for the waiting time but, based on the average number of OPD cases, Nkawie recorded 5040 cases whiles Aniwaa Medical Center recorded 8991. Again, Nkawie Government Hospital’s traffic intensity as well as its utilization factor are the same as Aniwaa medical Center’s which recorded 28.2% higher OPD cases. Despite both hospitals having their waiting times growing at an increasing rate that of Nkawie hospital is more severe than the Aniwaa Medical Center. This the researchers discovered was as a result of inefficiencies in the supervision of some of the services provided by the staff of Nkawie hospital. This confirms the perceived poor monitory and supervision of government institutions in Ghana (Lievens et. al., 2011).

RECOMMENDATIONS

Increasing the service rate of the OPD staff by way of introducing electronic based systems like Q- matic system for pre-registration, re-registration, and document reproduction functions of the patients coming to the Hospital will reduce cost and waiting times drastically in the both hospitals. The use of Q- matic system will help to speed up the patient flow as well as cutting down the cost of operation at the OPD unit. Increasing the number of personnel and rescheduling the work times of the staff of the OPD unit can help in reducing patients waiting in line drastically at the hospital. Introducing an appointment management system in both hospitals will also help in cutting down patients waiting in line.

REFERENCES

Badri MA, and Hollingsworth J. (1993) A simulation model for scheduling in the emergency room. International Journal of Operations & Production Management;13:13±24.

Journal of Applied Business and Economics vol. 16(3) 2014 187 Baurerle N., (2001), Markov Models. Optimization and Operations Research – Vol. IV. Cohen MA, et. al. (1980), Analysis of capacity decisions for progressive patient care hospital facilities. Health Services Research;15:145±60. Coté (1999) Patient flow and resource utilization in an outpatient clinic. Socio-Economic Planning Sciences 33 (1999) 231 245 0038-0121/99/$ - see front matter # 1999 Elsevier Science. Gravelle H, et. al.(2002), The demand for elective surgery in a public system: time and money prices in the UK National Health Service. Journal of Health Economics;21:423–49. Gross (1998), Fundamentals of queuing theory. Chichester: Wiley; [Accessed 2 May 2012]. Harrison (2000), Access to elective care: what really should be done about waiting lists. London: Kings Fund; 2000. Hornsby AS, and Crouther J.(2000) Oxford advanced learners’ dictionary. Oxford University Press, Oxford. 1042. Keller TF, and Laughhunn DJ (1973). An application of queuing theory to a congestion problem in an outpatient clinic. Decision Sciences;4:379±94. Koole and Mandelbaum (2002). Queuing models of call centres: an introduction. Annals of Operations Research;113:41–59. Lievens et. al., (2011). Creating Incentives to Work in Ghana, the International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW Washington, DC 20433 www.worldbank.org/hnppublications Ministry of Health (2010), Ghana Healthcare Industry. Report-2010. Morton A. and Bevan G (2008) Contrasting management science and economic perspectives on waiting for emergency care. Elsevier Health Policy 85 pages 207–217. Nketiah-Amponsah, E. and Hiemenz U. (2009), Determinants of Consumer Satisfaction of Health Care in Ghana: Does Choice of Health Care Provider Matter?.Global Journal of Health Science, 1(2): 50- 61. Ortola P (1993). User satisfaction in primary health care: the results of a home survey. Atencion Primaria; 12: 578-585. Rising EJ.(1977). In: Ambulatory care systems: design for improved patient flow, vol. 1. Lexington:Lexington. Santillan D (2000). Uses of satisfaction data: report on improving patient care. Soc Sci Med; 12: 24 – 26. Stafford and Aggarwal (1979) Managerial analysis and decision-making in outpatient health clinics. Journal of the Operational Research Society; 30:905 15. Thomas C. (2009), An introduction to customers flow management. White paper No 20, pg 23.

188 Journal of Applied Business and Economics vol. 16(3) 2014