سرمایه انسانی حسابرسی و بقای شرکت های حسابرسی: صنعت ممیزی هلندی در 1930-1992

This paper studies the relationship between auditor human capital and audit firm survival. Specifically, the effects are investigated of the human capital of auditors on the survival chances of newly established audit firms. Human capital is analyzed both at the time of entry of a new audit firm and during the lifetime of an audit firm. The data set contains 1693 firms that entered into the Dutch audit market in the period 1930–1992. To analyze the data, the technique of event history analysis is applied. There are two key results of the study. First, a higher level of education of the firm's auditors, both at founding and during the lifetime of an audit firm, generally increases audit firm performance. Second, the effects of experience at founding and experience during the lifetime of an audit firm vary considerably. Higher levels of experience at founding have a positive influence on audit firm performance. However, higher levels of experience during the lifetime of an audit firm, i.e., the aging of the firm's auditors, have a negative effect on firm performance.

The current structures of audit markets are the result of long-term processes of firm entry and exit, internal growth of firms, and inter-firm mergers and acquisitions. In these long-term processes, some audit firms appear to be more successful than others. Of the numerous small audit firms established in the past decades, most audit firms have exited the market while a few others survived and have become part of the Big Five.1 This raises a twofold question: why are some audit firms more successful than other audit firms, and what makes a new audit firm a likely candidate for survival? As an audit firm provides a professional service, human capital can be expected to play a key role in the explanation of audit firm success and failure (Morris & Empson, 1998). In fact, many changes in the structure of the audit market as a result of entry, exit, internal growth, and merger and acquisition imply the regrouping of individual auditors without any substantial change in tangible assets. Therefore, this paper will analyze the effects of two important components of human capital on audit firm performance. These two components are education (general human capital) and experience (specific human capital). More specifically, this paper attempts to assess whether audit firms with founders with high levels of education and experience have different survival and exit rates compared to their less educated and experienced counterparts. In addition, the paper attempts to assess whether hiring more highly educated and more experienced auditors over time affects the survival and exit rates of audit firms. The motivation for this paper is fourfold. First, this paper extends existing audit market research by investigating the effects of audit firm characteristics on audit firm survival. Determinants of audit firm survival, which can be considered to be an important measure of audit firm performance, have seldom been analyzed explicitly in previous audit market research. Second, the paper uses an alternative measure of the effects of auditor education and auditor experience. While there is a vast body of research on auditor education and auditor experience, most studies pertain to the consequences for decision making in specific auditing tasks. This paper analyzes the effects of education and experience for a wider group of decisions, i.e., decisions about building up and managing an audit firm. Third, the paper analyzes a part of the audit market neglected in previous audit market research, i.e., the small firm segment. Studying the process of entry, survival and exit of small firms adds to the understanding of the structures and dynamics of audit markets as a whole. Finally, from an organization sciences perspective, this paper integrates organizational ecology and industrial organization by including variables that are used in both streams of research. For the empirical analyses the Dutch audit market was selected. Many characteristics of the Dutch market are similar to those of the Anglo-Saxon audit markets covered in previous research (Buijink, Maijoor, & Meuwissen, 1998). For example, the international audit firm networks are all represented, dominating the Dutch market, and Dutch auditing and financial accounting regulations mirror international standards. The sample of the study consists of 1693 firms that entered the Dutch audit market in the period 1930–1992. This paper is organized as follows. In the following section relevant previous research is discussed covering both audit market studies and organizational survival studies. In the second section the hypotheses of this study are formulated using human capital theory. The third section describes the characteristics of the data set and the variables used in the empirical tests. The fourth section discusses the specific characteristics of the event history model at hand. The fifth section reports the results of the study. Finally, the sixth section provides a discussion and appraisal of the study.

This paper has analyzed the effects of human capital on audit firm survival, which is an important measure of audit firm performance. Four categories of human capital were distinguished: general human capital (education) and specific human capital (experience), human capital at founding and human capital during the lifetime of the audit firm. Additionally, various modes of exit of audit firms were distinguished, including pure exit and exit as a result of M&A. The hypotheses were tested with a longitudinal database of 1693 newly founded Dutch audit firms, and by applying event history analysis techniques that allow including time-varying independents in studying the duration until exit. There are two main results of this paper. First, the findings suggest, indeed, that human capital variables are important determinants of audit firm performance. This holds true for all four categories of human capital identified. Second, the results suggest that, with respect to performance, the various modes of exit should be treated differently. In the case of pure exit, a long period until exit is preferred, while in the case of exit by M&A, a short period points at superior performance. The specific results regarding the influence of the level of education on the longevity of audit firms are the following. A higher education level, either at founding or during the lifetime of the audit firm, has a positive effect on the longevity of the firm until pure exit. This positive effect is absent in the case of exit as a result of M&A. An increasing level of education after the founding of the firm even increases the chances of exit as a result of a take-over by a Big Four firm. As indicated in the section Human capital hypotheses, the positive influence of education on longevity until pure exit from the market provides an indication of a higher performance of these firms relative to audit firms with less educated auditors. The fact that there is no positive influence of education on longevity until exit by M&A, can be explained as follows. While education increases the performance of audit firms, it makes them also more attractive take-over candidates, which decreases their longevity. The results are in line with the suggestion that a higher education level increases the performance of the audit firm: either the firm has a higher chance of survival or the firm has a higher chance of becoming a Big Four target for an acquisition. The main results regarding the influence of previous experience on the longevity of audit firms are the following. On the one hand, experience at founding has a positive influence on the longevity of the audit firm until pure exit, and a negative influence on the longevity until M&A exit. Hence, experience at founding has a positive effect on audit firm performance. On the other hand, higher experience levels during the lifetime of the audit firm, i.e., the aging of the personnel of the firm, has a negative effect on audit firm performance. The aging of personnel during the lifetime of the firm decreases survival chances until pure exit, and reduces chances of becoming a take-over target. Table 8 classifies our key findings by reporting the overall association between the audit firm human capital features on the one hand and the audit firm events on the other hand. For the sake of the argument, we have reversed the results, so that the influence on the survival rate is reported. Table 8. An audit firm exit typology Audit firm feature Audit firm event Total exit Pure exit Small firm merger Big firm acquisition Diaspora resolution Education at founding (+) (+) NS NS NS Education over time NS (+) NS (−) NS Experience at founding NS (+) (−) NS (−) Experience over time (−) (−) (+) NS NS Mean education at founding (in %) 16 26 27 21 Mean experience at founding (in years) 3.46 3.78 4.63 4.79 Mean longevity 13.44 13.39 14.67 11.63 Audit firm type Retirement destiny Experience trap Education acquisition Career move Note: an NS indicates a non-significant, a plus (+) a positive and a minus (−) a negative association with the survival rate. Table options Overall, the pattern of our results suggests a typology of audit firm exit. Before we introduce this typology we would like to make the following two remarks. First, observing the descriptive statistics in Table 8 it appears that despite the significant differences in human capital variables, the differences between the various exit categories are rather modest. For example, the mean education level expressed as the percentage of academics at founding varies from 16 to 27. The variation in longevity is even smaller. Second, what follows is an appraisal that is meant to point the way to interesting conjectures and interpretations that may offer a steppingstone for future research by combining the `hard' statistical results from our hazard rate estimates with some `soft' suggestive descriptives in terms of the absolute levels of a number of key firm features. Together, the `hard' evidence and `soft' descriptives suggest an audit firm exit typology that links firm-level features to individual-level considerations, so offering challenging opportunities to combine labor (career) and firm (industry) issues in an overarching framework of firm and industry-level events. For sure, though, future empirical work is needed to assess the plausibility of the argument pursued below. (1) The retirement destiny. The pure exit event analyses reach significance for all four human capital features: education at founding, education over time and experience at founding are positively associated with prolonged survival, whilst experience over time is negatively correlated with the survival rate. At founding, the mean levels of education (16% academics) and experience (3.46 years) are low. The mean life duration is 13.44 years. All this together seems to indicate that the pure exit case pertains to the marginal audit firms in the market's fringe. That is, the low educated auditor has developed experience over time after establishing a small audit firm until the date of exit. Probably, in many cases exit occurs as a result of retirement. With this interpretation, pure exit is anything but a measure of success. Rather, in many cases, this route is taken for want of anything better. (2) The experience trap. The small firm merger case is associated with non-significant education parameters estimates. Both experience coefficients are significant, however. Experience at founding is negatively associated with the survival rate, whereas a positive linkage is observed for the experience over time proxy. In absolute terms, the mean level of education at founding (26% academics) is at the high end, whereas the mean founding experience level (3.78 years) is at the low end. Mean longevity is 13.39 years. An interpretation of these results may be that the auditors involved are caught into an experience trap: after establishing a rather small audit firm, they were not targeted as an attractive take-over candidate by a Big Four firm, but rather stayed where they were, without producing a substantial growth rate, until they were forced to merge. This route is therefore anything but a measure of success. (3) The education acquisition. The big firm acquisition event reveals only one significant effect: the firm's survival rate is negatively associated with education over time. The absolute mean levels of education (27% academics) and experience (4.63 years) at founding are relatively high. Additionally, the target firms witness the longest mean longevity of 14.67 years. This suggests that the Big Four firm buys the stock of highly educated auditors that the medium-sized firm has developed over a prolonged period of time. From the target firm's perspective, this is a measure of success. By building up an educational asset bundle over time, the medium-sized firm develops into an attractive take-over candidate for the big market leaders. Through this route, the target firm's auditors advance their careers by entering into the career ladder of a big rival. This entry into the internal labor market of a big audit firm is surely regarded as a manifestation of professional success. (4) The career move. The diaspora resolution event is linked to a single significant finding: exit through diaspora is negatively associated with experience at founding. In terms of mean experience at founding, the diaspora firms are ranked first (4.79 years), with the mean education level at founding being at the low end (21% academics). Moreover, mean longevity is the lowest with 11.63 years. Probably, the diaspora history reflects a career move. That is, highly experienced auditors apparently start a new, rather small, audit firm to build up a reputation over time that opens up opportunities to climb up the professional career ladder after a prolonged period of substantial growth. Contrary to the big firm acquisition case, the diaspora firm is not viewed as an attractive take-over target by the Big Four because the diaspora firm is: (1) on average too small, and (2) does not stand out for its education stock. So, by the end of the day, the highly experienced diaspora auditors advance their individual careers by all taking their own road to success. Of course, our typology is tentative only. For one, future research may focus on disentangling the processes that underlie this rough classification, since many of the suggestive interpretations are testable by focusing on auditor-level career data. For example, what happens to the partners of the medium-sized firms that are acquired by a Big Four firm? Additionally, the typology may be brought to perfection by introducing other and/or fine-grained variables. For instance, what is the influence of the partner-associate distinction that is so dominant in the audit industry, and what explains the different growth (and decline) paths of audit firms? The bottom line is that the audit industry is an intriguing mixture of a labor and product market. Much of what happens at the supply side of the product market in terms of firm-level events, is inextricably bound up with career motives at the supply side of the labor market. On the one hand, that is, a key driver of the industry's competitive processes is the human capital development within the industry as a whole and across audit firm types. The dominant production factor of an audit firm is the human capital embodied in the firm's auditor associates and partners. Much of what happens to an audit firm, is a direct consequence of the professional career motives of the firm's auditors. Our key argument is, perhaps, that a large part of the interpretation of the different audit firm events must rest on career arguments. So, this paper's main contribution is threefold: first, we have sought to deepen our understanding of what explains audit firm performance by introducing firm-level human capital variables in an IO-type of industry framework; second, we have illustrated how the hazard rate techniques of event history analysis offer a promising methodology for those interested in the longitudinal study of events at the level of individual auditors, audit firms or audit industries; and third, we have suggested how individual auditor career arguments are inextricably bound up with audit-firm level events. As argued above, this opens up a large array of avenues for future research. In any case, these and other questions indicate the challenging nature of the study of audit industries along the lines suggested in this paper.