Lancaster university management school

Lancaster University Management School Working Paper 2004/046 Diagnosis and Improvement of Service Quality in the Insurance Industries of Greece and Kenya Rand, Graham K The Department of Management Science Lancaster University Management School Lancaster LA1 4YX UK ©Rand, Graham K All rights reserved. Short sections of text, not to exceed two paragraphs, may be quoted without explicit permission, provided that full acknowledgement is given. The LUMS Working Papers series can be accessed at http://www. lums. co. uk/publications/ LUMS home page: http://www. lums. lancs. ac. uk/ Diagnosis and Improvement of Service Quality in the Insurance Industries of Greece and Kenya (mr )Evangelos Tsoukatosa, b, Simmy Marwaa, Graham K. Randa a Department of Management Science, Lancaster University Management School, Lancaster LA1 4YX, UK. b School of Management and Economy, TEI of Crete, Estavromenos, P. O. Box 1939, 71004, Heraklion, Crete, Greece. Abstract There is widespread customer dissatisfaction in the insurance industry, stemming from insurers’ failure to satisfy customers’ needs. Therefore, further research to improve the industry’s understanding of service quality is required. Using data from the Greek and Kenyan insurance industries, service quality is measured using the SERVQUAL methodology to identify quality determinants and existing quality gaps in the industries. Quality improvement strategies are recommended in each case. Some observations are made on the efficacy of the SERVQUAL diagnostic in assessing service quality in the insurance industry. 1. Introduction Previous studies, notably those of Wells & Stafford (1995), the Quality Insurance Congress (QIC) and the Risk and Insurance Management Society (RIMS) (Friedman, 2001a, 2001b), and the Chartered Property Casualty Underwriters (CPCU) longitudinal studies (Cooper & Frank, 2001), have confirmed widespread customer dissatisfaction in the insurance industry, stemming from poor service design and delivery. Ignorance of customers’ insurance needs (the inability to match customers perceptions with expectations), and inferior quality of services largely account for this. The American Customer Satisfaction Index shows that, between 1994 and 2002, the average customer satisfaction had gone down by 2. 5% for life insurance and 6. 1% for personal property insurance respectively (www. theacsi. org). In Greece, for example, 48% of consumers consider that the industry as a whole is characterized by lack of professionalism. Furthermore, 34% believe that insurers find various pretexts to avoid promised compensations (www. icap. gr). This is a legacy the industry has cultivated, sparking a host of controversies, denials and counter denials which unfortunately have not helped to 1 bolster its image worldwide. Several causes of poor service quality have been suggested: some with general application in the service industry and some specific to the insurance industry. It is therefore not surprising that measurement of service quality has generated, and continues to generate, a lot of interest in the industry (Wells & Stafford, 1995). Several metrics have been used to gauge service quality. In the United States, for example, the industry and state regulators have used ” complaint ratios” in this respect (www. ins. state. ny. us). The “ Quality Score Card”, developed by QIC and RIMS, has also been used. However, both the complaints ratios and the quality scorecards have been found to be deficient in measuring service quality and so a more robust metric is needed. Therefore, further research to improve the industry’s understanding of service quality is imperative. Using data from the Greek and Kenyan insurance industries, diagnostics have been constructed and service quality measured with a view to identifying quality determinants and existing quality gaps in the industries. Quality improvement strategies are recommended in each case. In the rest of this introductory section, background to the insurance industries of Greece and Kenya is given. In the second section the SERVQUAL metric is introduced. The following two sections describe how SERVQUAL has been applied to data from Greece and Kenya, before a comparison is made in Section 5. A final section draws some conclusions on the efficacy of the SERVQUAL diagnostic in assessing service quality in the insurance industry. 1. 1 Greece The origin of insurance in modern Greece is closely associated with commercial naval activities. Migrant Greek businessmen were involved in naval insurance activities from the closing decades of the 18th Century (Simitsek, 1997), relocating their activities to national soil immediately after liberation from the Turks. The Greek government tried to establish industry supervision guidelines as early as 1909 though the rules for the operation of insurance enterprises were, for the first time, introduced in 1917 after law 1023/1917 on ” private insurance enterprises” was passed by Parliament (see http://www. gge. gr/4/organ. asp? 209, in Greek). The delay in the industrialisation of the country, the distorted model of growth of Greek capitalism, bureaucracy and the stifling embrace of every enterprising effort by the state, forced 2 Greek insurers to remain, for a number of years, inactive and under funded (Simitsek, 1997). Until the beginning of the 1970s, business was mostly restricted to the transport and general property sectors. The country is now fully participating in the EMU and its economy is open, with no artificial restrictions. A total of 102 private insurers are active in the market, 20 of which exclusively transact life assurances, 69 non life insurances and 13 composite (Association of Insurance Companies – Greece, 2002). 74 of these companies are registered in Greece, while 28 are foreign subsidiaries, of which 28 are subsidiaries of EU companies and 6 non-EU companies. In the life sector the 5 largest companies write roughly 70% of total premium while in non-life insurance the 5 largest write 47% of premium (www. eaee. gr). According to a recent study by ICAP (www. icap. gr), the total volume of the private Greek Insurance market is approximately €2. 9 billions annually (2002) although its potential is estimated at €15—16. 5 billions. In the life sector, significant market extension through the Professional Insurance Funds, the Pension Funds and the private insurance programs is expected. In non-life insurance, the incorporation of EU laws into the Greek legal system, and the application of obligatory civil liability insurance to more professions, is expected to stimulate growth in the market (Tsoukatos, 2003). 1. 2 Kenya The concept of insurance and particularly the “ social insurance programme” dealing with socio-economic problems has been around Africa for a long time (Kenyatta, 1962). Members of a community pooled together resources to create a “ social insurance fund”. The “ premiums” ranged from material to moral support or other payments in kind. From the fund, “ drawings were made out” to support the few unfortunate members exposed to perils (Azevedo, 1993). However, the history of the development of commercial insurance in Kenya is closely related to the historical emancipation of Kenya as a nation (Throup, 1988). With the conquest of Kenya as a British colony complete, settlers initiated various economic activities, particularly farming, and extraction of agricultural products (Huxley, 1990). These substantial investments needed some form of protection against various risk exposures. British insurers saw an opportunity in this, and established agency offices to service the colony’s 3 insurance needs. Prosperity in the colony soon justified expansion of these agencies to branch networks with more autonomy, and expertise to service the growing insurance needs. By independence in 1963, most branches had been transformed to fully-fledged insurance companies (Maxon, 1993). In the forty years since independence, Kenya’s insurance industry has flourished, and by 2002 had 41 registered insurers, 15 transacting general insurance business, 2 transacting life business, while 24 were composite insurers — transacting both life and general insurances. Kenya’s insurance industry leads within the East Africa Community (a trading block of Kenya, Uganda and Tanzania), and is a key player in the COMESA region, (Common Market for Eastern and Southern Africa). The industry employs over 10, 000 people, underwrites well over €300m premiums, and pays over €120m per annum in claims. The largest 10 insurers handle over 70% of the motor business with a similar number handling well over 90% of the property business in the market. 1. 3 Comparison of the two industries The main characteristics of the two industries are summarized in Table 1. Greece Kenya – – – – – EU member Participates in EMU and Euro Zone – Open Economy Leading economy within East Africa Community (EAC) Key player in the Common Market for Eastern and Southern Africa (COMESA) Open economy Characteristics of Insurance Industry: Greece Kenya – – – – – – – – 9, 500 people employed Underwrites € 2. 9 billion per annum 102 Insurers Leading 5 insurers handle 70% of life business. In non-life insurance leading 5 handle 47% of business 10, 000 people employed Underwrites €300 million per annum 41 Established Insurers Leading 10 insurers handle 70% of the motor business and over 90% of the property business Table 1: Comparison between the Greek and Kenyan Private Insurance Industries 4 2. Quality Measurement – The SERVQUAL metric Various alternative instruments have been used to assess service quality, notable among these being the SERVQUAL diagnostic presented in 1988 and refined in 1991 by Parasuraman, Zeithaml and Berry, abbreviated as PZB. They conceptualized service quality (Q) as the difference between customers’ perceptions (P) of services of a specific firm and their expectations (E) of services in this particular industry. The negative P-E difference was characterized as a “ gap” or quality flaw. The following dimensions were used to construct the 22-item SERVQUAL scale (Zeithaml et al., 1990). – Tangibles — The appearance of physical facilities, equipment, personnel and communication materials. – Reliability – The ability to perform the promised service dependably and accurately. – Responsiveness – The willingness to help customers and provide prompt service. – Assurance — The knowledge, competence, and courtesy of service employees and their ability to convey trust and confidence. – Empathy – The caring individualized attention provided to customers. The SERVQUAL metric has been adapted to measure service quality in a variety of settings: numerous health care applications (Babakus and Mangold, 1992; Bowers et al., 1994), acute care hospital (Carman, 1990), independent dental offices (McAlexander et al., 1994), AIDS service agencies (Fusilier and Simpson, 1995), with physicians (Brown and Swartz, 1989; Walbridge and Delene, 1993), in large retail chains (Teas, 1993), and in banking and fast food restaurants (Cronin & Taylor, 1992). In addition, there have been several studies involving the insurance industry (Stafford et al., 1998; Leste and Wanderley (1997); Westbrook and Peterson (1998); Mehta et al. (2002)). Most of these studies brought about disagreements on two major issues: the dimensions of service quality, and the linkage between satisfaction and quality. Disagreement concerning the proposed linkage between quality and satisfaction has led to a division over causality, with one group supporting the proposition that quality leads to satisfaction (Woodside et al., 1989) and another supporting the proposition that satisfaction leads to quality (Bitner and Hubbert, 1994). Others suggest that quality and satisfaction are determined by the same attributes (Bowers et al., 1994). 5 Joseph et al. (1999) report that the Gaps/Disconfirmation model of SERVQUAL has been the object of some major criticisms, including ambiguity in the definition of expectations and its applicability to a variety of industries (Teas, 1993, 1994; Cronin & Taylor, 1992). The satisfaction approach to measuring quality runs into difficulty when complex services are evaluated as customers may not know what to expect, even after the service is delivered, as they may not know with certainty how good the service was (Lovelock, 1999). Furthermore, the model may be appropriate for large service organizations, but represents inaccurately service quality in small firms (Haksever et al., 2000). Another criticism is that for the model to function correctly expectations must remain constant, though Carman (1990) maintains that expectations change with familiarity to the service. Despite these limitations, the Gaps model provides valuable insight into understanding challenges of delivering quality service and sheds light into the various quality gaps (Zeithaml, 1988). Parasuraman et al. (1993), in response to a critique by Brown et al. (1993) of SERVQUAL’S difference score conceptualization, argue that the expectations component of SERVQUAL is a general measure and pertains to customers’ normative standards (i. e. the service level customers believe excellent companies should deliver). This serves as a yardstick against the services of a particular service provider (that the customers have experienced); so as to ascertain the latter’s service quality. As such, there is no conceptual reason for customers’ general evaluation standards to be correlated with their assessment for a specific company. PZB further argue that the SERVQUAL metric represents the core evaluation criteria that transcend specific companies and industries. Its items are the basic “ skeleton” underlying service quality and can be supplemented with context-specific items when necessary. The research studies presented here extend previous research by utilizing past findings to develop customized SERVQUAL metrics which are then used to diagnose service quality in the insurance industries of Greece and Kenya. The paper also examines the suitability of SERVQUAL’s application in the insurance industry. 3. Applying SERVQUAL to the Greek Insurance Industry. 3. 1 Adapting the SERVQUAL Instrument GIQUAL, the SERVQUAL type instrument developed for the measurement of service quality in the Greek Insurance 6 Industry, initially included 26 items, 22 from the revised SERVQUAL scale (Parasuraman et al., 1991) and 4 from extensive consultation with a group of 10 Area and Branch Managers of three leading Greek Insurers. The group concluded that although the five quality dimensions can indeed accommodate the various aspects of insurance quality in Greece, four additional items should be added to the SERVQUAL scale to evaluate the effect of price, product quality, ambiguity of insurance contracts terms and delays in claims settlement, on customers’ perceptions of service quality in the Greek insurance industry. Price was considered as a tangible item and product quality, ambiguity of terms and settlement delays as reliability items. The group further suggested that GIQUAL would better be used in the context of personal interviews. This was confirmed in the instrument’s pretesting phase, with a group of 50 experienced insurance customers, as in many cases extensive explanations on the meaning of certain items were necessary. GIQUAL was initially applied to a sample of 168 insurance consumers over 25 years old, having some contact with their insurance company in the last three months. For each consumer the difference scores Qi = Pi — Ei for the 26 items were computed and the Cronbach’s α reliability coefficients were calculated for each of the 5 quality dimensions. The removal or redeployment of items between dimensions was based on the “ increase of α if item deleted” criterion (Pallant, 2001). During this process only one item (Q7 – price) had to be removed, leaving GIQUAL with the following 25 items1. – – Tangibles (four items) — modern equipment and technology (Q1), visually appealing physical facilities (Q2), neat appearing employees and agents (Q3), visually appealing materials associated with service (Q4) Reliability (eight items) — keeping promises when promise to do something by a certain time (Q5), offering products and services of utmost quality (Q6), issuing contracts with clear, transparent and non ambiguous terms (Q8), settling customers’ claims with no unnecessary delays (Q9), showing sincere interest in solving customers’ problems (Q10), offering services right the first time without unnecessarily discomforting customers (Q11), providing services within the specified contract time 1 α values varied from . 78 to . 93 between dimensions. 7 limits (Q12), issuing error free bills, statements, receipts, contracts, claims and other documents (Q13). – – – Responsiveness (four items) — telling customers exactly when the services will be performed (Q14), doing their best to give prompt service to customers (Q15), always willing to help customers (Q16), never being too busy to respond to customers’ requests (Q17). Assurance (four items) — customers feeling safe in their transactions (Q18), behaviour instilling confidence in customers (Q19), being consistently courteous with customers (Q20), having employees and agents with the necessary knowledge to give professional services to customers (Q21) Empathy (five items) — giving customers individual services (Q22), operating hours convenient to all customers (Q23), giving customers personal attention (Q24), having the customers’ best interests at heart (Q25), understanding the specific needs of customers (Q26). The difference scores for the instrument’s 25 items, the Qs, were factor analyzed to examine the dimensionality of the scale. Before the analysis the data set was screened for errors of omission, tested for normality, outliers, sampling adequacy, factorability of the correlation matrix (R) and, because the determinant of R was very close to 0 (5. 88E-10), examined for singularity and multicollinearity. The results of all these tests revealed no problem for the analysis (Belsley et al., 1980; Comrey and Lee, 1992; Tabachnick and Fidell, 1996; Hutcheson and Sofroniou, 1999; Pallant, 2001). Although the Kaiser criterion (Eigenvalues> 1) suggested a 4-factor solution and the Screeplot criterion (retention of factors above the elbow) suggested a 2-factor solution, in an attempt to verify the 5 quality dimensions of SERVQUAL, the analysis was initially constrained a-priori to 5 factors, using the Principal Axis Factoring procedure as suggested by Parasuraman et al. (1988). Setting the “ criterion of meaningful factor loading” to 0. 35, the unrotated solution was degraded to a 4 factor one as the loadings of the 5th factor were all less than 0. 35. To allow for factor intercorrelations the initial solution was subjected to oblique rotation, using the SPSS oblimin procedure. After deleting items Q5, Q18, and Q24, as loading on several factors, a 22 items rotated solution resulted. 8 The Tangibles and Reliability dimensions were still there but the Responsiveness and Assurance items merged into a single factor together with 2 Empathy items while the 2 remaining Empathy items formed a fourth factor. The importance of factors with respect to extracted variance was: factor 3 (Responsiveness, Assurance and Empathy items), factor 2 (Reliability), factor 4 (Remaining empathy items) and factor 1 (Tangibles). Although α values were high (with the exception of factor 4 with α = 0. 68), factors 2, 3 and 4 were highly or relatively highly correlated. This suggested that they could possibly form a single factor in a 2-dimensional solution, in line with the Screeplot criterion. A 2-factor solution was next investigated, using once again the principal axis factoring procedure and maintaining 0. 35 as the criterion of meaningful factor loading. After subjecting the initial solution to orthogonal rotation, using the SPSS Varimax procedure, a full 25 items clear-cut solution resulted (Table 2). Both factors were reliable (having high α values), internally consistent and well defined by the variables, as the Square Multiple Correlations for factors from variables were 0. 96 and 0. 80 for Non-Tangibles and Tangibles respectively (Tabachnick & Fidell, 1996). Dimension Reliability Coefficient (α) Items (Factor Loadings) Q5 (00. 59) Q6 (0. 61) Q8 (0. 64) Q9 (0. 8) Q1 (0. 87) Q11 (0. 82) Q12 (0. 79) Q13 (0. 65) Q14 (0. 82) Q15 (0. 81) Q16 (0. 77) Q17 (0. 72) Q18 (0. 89) Q19 (0. 83) 1. Non – Tangibles 0. 96 Q2 (0. 66) Q21 (0. 76) Q22 (0. 56) Q23 (0. 48) Q24 (0. 57) Q25 (0. 79) Q26 (0. 81) 2. Tangibles 0. 78 Q1 (0. 60) Q2 (0. 76) Q3 (0. 73) Q4 (0. 61) Loadings of items on dimensions to which they don’t belong are all