Client influence on valuation: perceptual analysis of the driving factors.

ABSTRACT. Recent behavioural real estate research effort on valuation bias judgement suggests that client influence is an important source of such bias. This paper, therefore, considers the theoretical potentials for client influence to bias valuations, and empirically analyse the factors impacting on the extent of influence client may exert on estate surveyors and valuers. To accomplish this aim, questionnaires were randomly administered to samples of estate surveyors and valuers in Lagos Metropolis. Using mean rating point, the survey ranked three factors; integrity of valuer or valuation firm, importance of the valuation outcome to the client and client size, as the most significant clients' influencing factors. The results of the Chi-square test did not, however, provide any statistical relationship between the size of firm, amount of experience and education of estate surveyors and valuers and their perception on the comparative importance of the identified clients' influencing factors.

KEYWORDS: Client influence; Property valuation; Empirical analysis; Driving factors; Estate surveyors and valuers

1. INTRODUCTION

In the light of the subjectivity nature of property valuations, an important issue that has engaged the attention of real estate researchers over the last decade is the disparity between property transaction prices and market values. To date, the consensus, except for the study of Lai and Wang (1998), is that valuations are generally not accurate estimates of market values. The evidence to that effect, according to Levy and Schuck (1998, 1999), falls in three areas. First anecdotal evidence clearly indicates that practitioners believe valuations are lagged estimates of market value (i.e. Webb, 1994). Second, empirical studies that analyze the relationship between transaction prices and contemporaneous valuations also conclude frequently that incongruence between these variables is evidence of error (see for example Zotzour, 1988a, 1988b; Diaz, 1997; Diaz and Wolverton, 1998). Third, empirical studies that investigate the statistical properties of time series of valuation based returns conclude that serial correlation is evidence of bias (e.g. Ibbotson and Siegel, 1984; Cole, 1988; Geltner, Graff and Young, 1994; Geltner and Geotzmann, 1998; Geltner, 1993, 1998; Young and Graff, 1995; Graff, Harrington and Young, 1997).

Valuers' error in price predictions arises from random variation in observed prices of comparable sales. For instance, any individual property at a particular point in time has, possibly, different prices due to different circumstances of sale, deferring buyer preferences, different buyer information sets or other factors which the valuer is unable to effectively account for (Aluko, Ajayi and Amidu, 2004). A number of factors, however, contribute to the existence of random error; including the generally poor quality and quantity of information available to valuers, the shortcoming in valuation methodologies and a disparity between the definitions of 'market value' and the figure that a valuer seeks to estimate. Valuations are believed not only to suffer from random error but also from measurement and misspecification errors, also known as "interpretative errors" (Aluko, 2000), in the price differences models, which in turn increases as more comparables are added (Aluko, 2000; Kummerrow and Galfary, 2002; French and Gabrielli, 2004).

Recently, the quest to explain the error in valuations has tended to centre on behavioural attitudes of the parties to the production of valuations. This, according to Diaz (2002) is premised on the simple argument that before valuation improvement can be engineered, valuation behaviour must be understood. Behavioural real estate research, however, was initially focused on understanding the actual valuer problem solving in terms of departure from normative models (Diaz, 1987, 1990; Diaz et al., 2002), bias in the process of selecting comparable sales (Gallimore and Wolverton, 1997), heuristic problem solving behaviour (Gallimore, 1994, 1996; Diaz, 1997) and anchoring behaviour (Harvard, 1999, 2001; Cypher and Hansz, 2003; Diaz, 1997; Diaz and Hansz, 1997, 2001; Hansz, 2004a; 2004b; Hansz and Diaz, 2001; Diaz and Wolverton, 1998). All these studies tend to establish the fact that an expert usually solves problems in a manner that is inconsistent from the way in which he/she learned to solve them as a novice, thus leading to valuation judgement bias.

More recent research effort, apparently looking beyond valuation issues per se, has uncovered evidence of influence (from client, market and valuers) as also a likely cause of biased estimates of market values (see for example, Gallimore, 1994, 1996; Worzala et al., 1998; Kinnard et al., 1997; Roberts and Roberts, 1991; Smolen and Hambleton, 1997; Rushmore, 1993; Wolverton and Gallimore, 1999; Gallimore and Wolverton, 2000; Levy and Schuck, 1998; 1999; Amidu, 2006; Amidu and Aluko, 2007). These studies set out a body of theory and empirical evidence, which suggests that relationship between clients and valuers, would influence the valuation formation process. Except for the study of Levy and Schuck (1999) and Amidu (2006) that has consider the entire issue of client influence holistically, others have concentrated either on ascertaining whether specific type of influence exists or on the reactions of valuers in turn. The study of Levy and Schuck (1999), however, only qualitatively identified factors potentially impacting on the type and amount of influence client may exert on a valuers and ultimately the reported value of a property. This study contributes to knowledge by providing an empirical analysis of the factors from the view point of professionals in valuation practice in Nigeria.

2. DATA DESCRIPTION AND METHODOLOGY

The theoretical expectation about the potential that clients have for influencing valuations as developed in Levy and Schuck (1999) forms the basis upon which an empirical analysis of clients' influencing factors was carried out. According to the authors, such potential depends on several factors that can be categorised under four main headings, as shown in Figure 1.

[FIGURE 1 OMITTED]

These factors determine the extent of influence that may be brought to bear on valuers and the valuation process. However, only the characteristics of individual valuer and valuation firms and the characteristics of the clients were adopted for this study since the rest do not relate to client influence but rather to procedural influences that may not necessarily emanate from clients. Variables under these two factors include:

i. Integrity of the valuer and or ethical culture of the valuation firm.

ii. Earnings of the valuer/valuation firm from a client.

iii. Number of task undertaken for a client.

iv. Valuer's prior involvement with the subject property.

v Size of a valuation firm.

vi. Valuer's involvement in other consultancy work for the client.

vii. Valuer's style of decision-making.

viii. Type of client (sophisticated and unsophisticated).

ix. Client size (big or small).

x. Importance of the valuation outcome to the client.

xi. Client's financial condition.

The estate surveyors and valuers' perception of the significant of these factors as they affect possible client influence (by way of either encouraging client to put pressure or helping the valuer to resists such pressure) was measured on a 5 point likert format from 1 (strongly disagree) to 5 (strongly agree). The Estate Surveying and Valuation Firms or Estate Surveyors and Valuers, by virtue of the Estate Surveyors and Valuers Act No 24 of 1975, are statutorily empowered to value proprietary interest in land for all purposes. According to the Institution directory (2003), approximately 52 percent (228) of firms practicing estate surveying and valuation in Nigeria have their headquarter offices located in Lagos Metropolis, the study area. Consequently, a random sample of 137 estate firms, representing 60% of firms in the study area and 31% of firms in Nigeria, were selected. Questionnaires were then personally administered to the Head of Valuation Department of the selected firms. The valid 88 responses to the measure used in the analysis shows a total response rate of 64%, representing 39% for the study area and 20% in the whole country.

The study employed both descriptive and statistical methods of data analysis. These comprise frequency distribution, cross tabulation and chi square. The frequency distribution shows the basic distributional features of the data on the respondents' estate surveyors and their respective firms and also the data employed in the subsequent statistical analysis. The descriptive statistics like mean was used to rank the variables in addition to frequency and percentage distributions.

Finally, cross tabulation and chi square test was employed to determine whether the differences in the rating of the variables could be based on the size of firm, the amount of education and level of valuation experience of respondents.

3. EMPIRICAL RESULTS

3.1. General information

In order to establish the reliability of the data collected for this study; which is a measure of the suitability of the target population, the study sought information on issues relating to the characteristics of the sampled firms and the responding estate surveyors and valuers. Respondents were, for instance, asked to indicate their firms' areas of professional practice. Their responses are as given in Table 1.

The table indicates that the entire responding estate surveying and valuation firm have more than one areas of professional practice. When the facets of professional practice are considered individually, 87.5% of the 88 firms sampled engage in property valuation, 92% in agency, 60.2% in property development, 88.6% in property management and 79.5% in general real estate services. These results attest to the fact that the sampled firms are very relevant to this research work.