Chapter 6 Survey; Method and Results
6.2 Survey: Approach and Application
6.2.2 What is the survey about?
Widener and Selto (1999) surveyed 600 randomly selected Compustat firms with more than 500 employees in the United States. They hypothesized that the Transaction Cost Economic factors Asset Specificity, Uncertainty, and Frequency83 would impact the ‘Make or Buy’
decision of internal audit in organizations. Their study claimed that firms with high Asset Specificity scores for Internal Audit activities indeed internalize Internal Audit sooner than others. As to Uncertainty, their hypothesis was not supported meaning that Uncertainty was not an explanatory factor for the ‘Make or Buy’ decision. Frequency on the other hand, had some merits explaining the Make or Buy issue. However, the qualitative part of their study showed that Asset Specificity was by far the most contributing reason for having an IAF internally. Also (Behavioural) Uncertainty and strategic needs were determinants of outsourcing. Frequency on the other hand was then rarely cited as the basis for Internal Audit sourcing. Frequency was found to be related to archival measures of size and how often Internal Audit Services were demanded. Widener and Selto (1999, p 66) concluded that: “It is plausible that Asset Specificity is a major driver of the decision to outsource Internal Audit, both from out-of-pocket and opportunity cost perspectives.” Frequency, especially when interacting with high Asset Specificity also seemed to be a plausible driver. Their regression model explained 53% (adjusted R2=0.46) of the variation in outsourced internal auditing.
There were a few limitations to their research. First, their survey was geared towards measuring Asset Specificity of the Internal Audit activities predominantly. This seems a major limitation because their survey showed that indeed Internal Audit activities had a high level of Asset Specificity but that in itself doesn’t tell the whole story. It is my conjecture that because the organization has high levels of Asset Specificity, Internal Audit activities will consequently show high levels of Asset Specificity as well. Furthermore, it may very well be that Internal Audit activities do have an inherent high level of Asset Specificity due to the fact that specialized training and knowledge is required, the use of proprietary information (as shown in Widener and Selto (1999)), and therefore already contain a high degree of what is called ‘human Asset Specificity’. It is the well-known chicken-and-the-egg problem.
Measuring Asset Specificity of Internal Audit activities may be less contributing to the explanation of the ‘Make or Buy’ issue than measuring the Asset Specificity and Information Asymmetry of the organization it is part of. In this survey we aimed at measuring Asset Specificity of the organization as a whole. Section 6.2.3 discusses this issue.
A second limitation of the Widener and Selto (1999) survey was the quality of their measures.
They also doubted the reliability of the Uncertainty measures. The third limitation is the relatively small sample size. A response rate of 14% (83 out of 600) may not have been convincing. They could not find any major industry effect with the exception of Wholesalers but there were only 3 represented in their usable sample. And last but not least, their study was limited in geography because it was based in the United States and therefore might not be applicable in other territories. In this context note that the United States has a prevailing one-tier system as opposed to the two-one-tier environment in The Netherlands. Last but not least,
83 See Chapter 4 for an explanation of these three concepts.
their data originated in 1996 and by the time of my research some 8 years old. During this period, significant changes occurred in the Corporate Governance landscape as shown in Chapter 3, which of course might have had their effect.
In the present study only parts of the Widener and Selto (1999) study were used to investigate whether their findings are corroborated by the Dutch Internal Audit environment and are still valid some 8 years later. Several adjustments were made. First, as shown in Chapter 4, Asset Specificity has indeed been shown to be the most powerful indicator of the Make or Buy decision. Widener and Selto contributed to that conclusion on the part of Internal Audit.
Uncertainty and Frequency were excluded for two reasons. One of them is that Uncertainty did not deliver any support in the Widener and Selto study and they were not sure about the reliability of their measures. Furthermore, the case studies of Chapter 5 showed more promising numbers for Asset Specificity then they did for Uncertainty and Frequency. The decision to exclude Uncertainty and Frequency were backed up after drafting and execution of the survey by three other studies. One study was done by David and Han (2004), one by Carter and Hodgson (2006), and the last one by Boerner and Macher (2006). Their specific findings will be discussed in section 6.2.3. The study done by Van Elten (2005) using the Widener and Selto questionnaire also concluded that Uncertainty did not provide support for their hypothesis that Uncertainty explained the make or buy issue. Frequency however – though less convincing – provided some support. Therefore, it was decided to focus on Asset Specificity solely and to research whether the Widener and Selto findings hold 8 years later and in a different territory and culture and not the least also in a two-tier environment as opposed to the one-tier environment in the United States. Finally I intended to take a look at industry effects as well. Since Widener and Selto were not able to find any it would be nice to be able to further our knowledge if we could get some clue as to the existence of these effects.
A high degree of Asset Specificity requires Internal Audit to acquire deep knowledge of the organization and therefore to in-source an IAF. This knowledge is best acquired via learning-by-doing and being part of the organization. Furthermore, an external provider would need to bear significant costs for training and sustaining firm specific auditor knowledge which would only be acceptable if the contract has a longer term horizon to enable to recover the investments. Not only would any organization be reluctant to underwrite such a contract because it would be difficult to specify the terms of the services rendered and foremost, it would also be difficult to ascertain whether those investments were made. Who is going to exercise oversight on behalf of the organization? One needs to be a professional to be able to assess the proper execution of the contract anyway. If such a contract would be ‘incomplete’
by nature the possibilities to be subject to opportunistic behaviour would be manifold. On the part of the outside provider such a contract would lead to significant opportunity losses if it was ended before initial costs would have been recouped.
Hypothesis 1 therefore is as follows:
Hypothesis 1:
If Asset Specificity of an organization is high, this will be positively associated with an in-sourced IAF and negatively with outsourcing.
Secondly, since the Widener and Selto (1999) study – and others (see below) - already indicated that Asset Specificity explained the Make or Buy decision, it was decided to study whether their conclusion held in the Netherlands in some more detail. To that end we investigated whether there is a relation between the degree of Asset Specificity of the
organization and the size of the IAF and the breadth of the Scope of Services. Therefore, we needed to get as much participating organizations in this survey that do have an IAF rather than sending it to a sample of all organizations. The next sections elaborate on how this goal was achieved.
As shown in the previous Chapters, there are no conclusive clues as how large the IAF should be. There is some understanding that the size of the organization will impact the numbers of the IAF. The same is supposedly true for the industry; Financial Services companies do have large IAF’s. But, would Asset Specificity also tell us how large the IAF would be? Would the degree of Asset Specificity be associated to the staffing of the IAF and thereby provide some kind of metrics that would allow for organizations to ‘calculate’ how large the to-be internalized IAF should be or at least benchmark the size of their existing IAF? Based upon the earlier findings of this research study it was assumed that increasing degrees of Asset Specificity will also lead to increasing staff numbers for the IAF. Due to idiosyncrasy, organizations that do have high degrees of Asset Specificity will most likely need not only more specialized skilled auditors and this knowledge will need to be contained not only within the IAF but also within the organization, but foremost this idiosyncrasy might have an increasing effect on staff numbers as well. More specialized and skilled auditors maybe less easily brought in for other audits for two reasons; first of all their expertise needs to be exploited to their full potential and therefore need to be geared towards their main area of expertise; secondly, they might be rather useless for other purposes due to the lack of specialized knowledge in other areas. These factors might drive staff numbers. Therefore:
Hypothesis 2:
If Asset Specificity of an organization increases, staff numbers of the IAF increases as well.
As Widener and Selto (1999) pointed out that the Scope of Services of the IAF might be related to the:
1. Highly specialized knowledge (and therefore Asset Specificity) needed to perform internal audits; and
2. Strategic demand for their services.
The idiosyncratic use of Internal Audit is also demonstrated quite clearly in Chapter 5. The abundant number of different audit types might also be explained by the degree of Asset Specificity of the organization they are part of. The almost perfect split between Internal Audit and the external auditor in 3 out of the 4 case study companies, whereby the external auditor focused on the financial audit and the IAF on other audits, predominantly operational and IT audits, is exemplary in this respect. The, at least supposedly, commodity financial audit type is done by the outsider and the rest by the insider.
This shift during the last decade may have had an impact on the possibility of the IAF to gain stature and deepen their knowledge of the business and thereby becoming increasingly involved in the strategic realm of the organization thus supporting the conjecture of Widener and Selto (1999, p 67) ‘….that if Internal Audit personnel wants to stay within the firm they should become more central to the management and day-to-day control of strategic assets and less concerned with more generalized services that can be outsourced easily’. That argument was already made by Ratliff and Beckstead (1994). This development might have fostered the need for their services, not only increase their numbers, but foremost impacting the number of different audit types and thereby the breadth and reach of the IAF services. In other words, might the degree of Asset Specificity also be related to the Scope of Services of the IAF?
The previous paragraph elaborates on the need to have specific knowledge of the organization, creating more specialized auditors that need to be exploited to their full potential and might be – perhaps in some way – be rather useless for other types of audits. This effect might reiterate in the way described in Chapter 2 where we applied the ‘unbundling’ model (Matthyssens et al, 1998). This model was then used to offer some explanation why internal audit sought to expand its services beyond standardized audits such as financial audits. This model might also offer some clues to the breadth of services.
If Asset Specificity is high and deep knowledge and expertise is called for, specialized auditors will be recruited and predominantly in-sourced. Outsiders are not only more expensive but their knowledge will be lost as soon as they no longer render their service to the organization. Furthermore, the availability of their services will allow the ‘buyer/user’ to become experienced and – taking that the services are rendered to the satisfaction of the Management and the Supervisory Board – create a buy impulse which may also foster the need to explore other areas as well. If this ‘pull’ wouldn’t be spontaneously, a specialized and skilled audit force may very well be able to apply a ‘push’ strategy (see Chapter 4). This situation might explain what Power (1996, p 312) once called the ‘reproducing’ of all kinds of audits.
Therefore:
Hypothesis 3:
If Asset Specificity of an organization increases, the Scope of Services of the IAF will be broader.
Until now we have build solely on Transaction Cost Economics and the dominance of Asset Specificity. Now let us take another look at Agency Theory. As shown in Chapter 4, the dominant idea behind Agency Theory is Information Asymmetry. Information Asymmetry founded either on ‘hidden information’ or ‘hidden action’. In any case, principals are in need of a monitor to exercise some kind of oversight over their agents. In this respect, there might be some similarity with the Asset Specificity case made above that if Information Asymmetry is high or increased similar effects might occur.
Higher levels of Information Asymmetry might induce more anxiety on the part of the principal and mutatis mutandis lead to an increased need to have an Internal Auditor at hand to close the information gap. Now, the question would be does that need to be an in- or an outsourced IAF? Clearly, there might not be an easy answer. It is my conjecture that if Information Asymmetry is high, there will be a need on the part of the principals to be informed on a more frequent basis to close the gap. The higher the level of Information Asymmetry the higher the frequency of using an auditor will be. Although we did not include Frequency as a contributing factor of the Transaction Cost Economic perspective, Frequency in relation to Information Asymmetry (and thus Agency Theory) might still be of importance.
As discussed above, Widener and Selto (1999) showed that Frequency did have an effect on the outsourcing issue, especially in conjunction with Asset Specificity. The very same could be true for Information Asymmetry in conjunction with Frequency. In this study there is some belief that there is a relation between Information Asymmetry and Frequency. The higher Information Asymmetry, the more likely it will be that there is a more frequent need for audit services. And, if there is a more frequent need due to Information Asymmetry, the more likely