Risk attitudes of tax practitioners and firm influence

Abstract

Purpose

The purpose of this paper is to garner a deeper understanding of the site of influence of aspects of risk management for tax practitioners.

Design/methodology/approach

The research design is twofold. Phase one consisted of a wide-scale international survey with 1,061 tax experts across 59 jurisdictions. In phase two, the authors followed up with 68 semi-structured interviews with tax practitioners working in 11 different countries.

Findings

The findings recognise the importance of the firm as a significant “site of influence” for tax practitioners in shaping their risk appetite in their tax work. The firm eclipses other influences of risk such as professional body oversight, public interest and demographic markers such as gender and career stage. The authors show that firm is significant, irrespective of size of firm.

Practical implications

This work has practical implications as the findings highlight the importance of oversight of professional service firms by both the professional accountancy bodies and revenue authorities. The findings may have impact on the ethical training and guidance for trainee accountants in terms of an increased awareness on the employing firm as a site of influence for tax practitioners.

Originality/value

This research is important as it adds to the significant body of work on firm socialisation and highlights the important role that the firm holds in moderating (or exacerbating) the risk appetite of tax practitioners, which has significant implications in terms of pushing the boundaries of tax aggressive behaviours. The work aims to recognise the important role that tax practitioners can have in moderating aggressive tax practice, and, thus, reducing tax inequalities and shaping a better world of “Reduced Inequalities” (SDG10).

Citation

Lynch, R. and McCullagh, O. (2024), "Risk attitudes of tax practitioners and firm influence", Meditari Accountancy Research, Vol. 32 No. 7, pp. 65-87. https://doi.org/10.1108/MEDAR-06-2023-2050

Introduction

Tax practitioners act in an intermediary role between the taxpayer and the relevant tax authorities. As such the risk perspective of the tax practitioner is worth exploring to understand how this risk attitude of the tax practitioner may impact on tax compliance of the taxpayer, and, thus, increase the public coffers (Saad, 2014). Tax practitioners in their daily work have many potential risks that may arise from their tax decision-making such as threat of litigation (Alm et al., 2012); power of the state to impose penalties (Alm et al., 2022; Goldswain, 2001); threat of audit by taxing authorities (Mendoza et al., 2017); threat of audit or penalty by the professional body (Carter and Mahallati, 2019; Killian et al., 2022); and personal reputational loss (and corresponding impact on career development) (Ghosh and Tang, 2015; Kadous et al., 2008; Krishnan and Krishnan, 1997; Reynolds and Francis, 2000).

We use insights from Power (2007) on risk decision-making in highly regularised environments where “value judgements which inevitably enter the decision process are widely shared” (p. 14). This draws upon prior work by Short (1992), who argues that “risk-related decisions often are embedded in organizational and institutional self-interest” (p.8). Whereas the role of tax practitioners may vary, often they are the mediators of client risk and regulatory requirements. We aim to explore the significance of firm socialisation as a “site of influence” for tax practitioners in how they mediate risk in their day-to-day tax work. Through the combination of a large set of surveys and semi-structured interviews with tax practitioners, we aim to gain insights into their lived experience and interactional expertise and thereby contribute to the growing literature that views tax as a social practice (Gracia and Oats, 2015; Makovicky and Smith, 2020).

Attitudes to risk, risk preferences or risk appetite are terms used interchangeably to convey an individual’s risk tolerance levels. An individual’s risk attitude significantly influences risk decision-making, particularly where there is scope for personal judgement. Interest in the measurement of risk attitudes, its influences and its impact, has spawned a rich literature spanning diverse domains such as engineering (Van Bossuyt et al., 2013), economics (Concina, 2014), sociology (Umamaheswar and Tan, 2020) and psychology (Xu and Cheng, 2021). In this research, we focus on the risk attitudes of tax practitioners, and how that attitude may influence their day-to-day tax work. Drawing on data from a wide-scale international survey (n = 1061), and a series of semi-structured interviews (n = 68) with tax practitioners, we garner an in-depth understanding of tax practitioners’ risk attitudes as it pertains to their tax work.

The remainder of the paper is set out as follows; in the next section, we explore the literature in this area. In the following section, we discuss our research methodology. This is then followed by a presentation of our findings. And in the final section we discuss our findings and conclude.

Literature review

Gracia and Oats (2012) refer to tax work as a “complex and fuzzy boundary between acceptable and unacceptable tax practice” (p. 304). The work of tax practitioners has an imperative role in navigating these “boundaries” on behalf of the taxpayer (Fogarty and Jones, 2014). How tax practitioners navigate these complexities has a very significant role on the level of taxpayer compliance (Frecknall-Hughes et al., 2023), which may be influenced by a myriad of factors such as practitioners’ personal ethics (Fatemi et al., 2020), client preferences (Carnes et al., 1996; Tan, 2011; Blanthorne et al., 2014) and practice risk (Kadous and Magro, 2001; Kadous et al., 2008).

Tax practitioners commonly act as an intermediary between the taxpaying public and the relevant tax authority, resulting in a distinctive role as it creates an opportunity for the tax practitioner to have a significant influence on the encouragement of good tax practices by the taxpayer, which has the resulting consequence of increasing availability of public funds to the exchequer. “Non-compliance alters the distribution of income in arbitrary, unpredictable, and unfair ways” (Alm et al., 1992, p. 56). Considerable work has been done in the literature on taxation focusing on tax compliance, avoidance and evasion (Kirchler et al., 2008; Sikka and Willmott, 2010; Cooper et al., 2013; Pickhardt and Prinz, 2014; Kanagaretnam et al., 2018) and on the role of the profession in facilitating tax avoidance (Sikka and Hampton, 2005; Doyle et al., 2013). Although the ultimate obligation for tax compliance falls to the taxpayer, tax practitioners can be seen as mediators of tax compliance “assisting taxpayers in the determination of taxpayer liabilities” (Alm et al., 2012, p. 34).

The firm has increasingly garnered attention in terms of its significance as a site of influence on the socialisation of accountants and tax practitioners (Anderson-Gough et al., 2002; Cooper and Robson, 2006). It is significant if firms influence tax practitioners to such an extent, as prior literature points to the “antisocial practices” of major accountancy firms that remain “addicted to avoidance schemes even though they have been fined and some of their personnel have received prison sentences” (Sikka, 2016, p. 259). Gebreiter (2020) reports on the firm’s goal to attract and construct “the ideal recruit”. This points to the value placed on tax practitioners to have an overall “firm outlook” in their daily work. The importance of the socialisation process of tax practitioners to firms is also highlighted by Killian et al. (2022) as significant in “set[ting] the tone for career-long ethical framing” (p. 1).

Interestingly, Radcliffe et al. (2018) examine the work of tax practitioners in the context of changing moral boundaries. Anchoring their position to the conception of tax as a social practice, they engage directly with tax professionals to better understand the influences of institutions, society and politics on tax practices. They find that changing moral incentives contrasts commercial “logic” and leads to different cohorts of tax professionals repositioning themselves in response. The risk of negative public perception is evidenced in their interviews with practitioners as a potential consequence of failing to guide moral tax behaviour. The strategic necessity to visibly address public perceptions is argued to direct behaviour rather than the adoption of a socially-focused moral position (Suddaby 2010; Radcliffe et al., 2018). Furthermore, Spence and Carter (2014) find that the technical–professional logics of acting in the public interest, ethics and integrity are reframed in the more business-friendly language of risk management by partners in Big 4 accounting firms. Thus, these issues are managed through technical and legal expertise whilst avoiding the need for a moral stance. The expectation by partners, as identified by Spence and Carter (2014), for these risks to be managed at subordinate levels of the firm highlights the need for further examination of the firm as a site of influence on the risk attitude and behaviour of tax practitioners.

Tax is a highly codified practice (Marchant and Robinson, 1999), with a strong emphasis on legal conformity and compliance. The codification of knowledge in professional service firms has been examined in the literature (Morris and Empson, 1998), however, the codification of risk is under-explored. Mulligan and Oats (2009) explore tax aggressiveness as a manifestation of low risk aversion and find that, whereas a formal document expressing the risk strategy or policy of the firm may not exist, tax practitioners within the firm have a clear understanding of the firm’s risk appetite. Power (2009) laments the “impoverished conception of “risk appetite”” (p. 850) to an auditable organisational process. Moreover, Power (2009) argues that the concept of risk appetite invokes “the neoliberal postulation of organizations as “enterprising” selves” (p. 850) rather than having a precautionary meaning. Interestingly, Power (2007) discusses the repositioning of accounting in the 1990s to business risk analysts to advise the entrepreneurial risk-taker. In a similar vein, the aspirational efforts of management accountants to position themselves as business partners are examined by Goretzki and Messner (2019). Similarly, tax advisors advise their clients in a manner to reduce their tax exposure, thus, enabling entrepreneurial activity. This may appear to disaccord with the notion of tax preparers as gatekeepers of tax compliance and ethics (Frazer et al., 2018), yet with incentives from the two counter positions, tax practitioners aim to satisfy both exigencies. Navigation is corralled by the risk associated with perceived failure on either count. Failing to meet expectations (regulatory or monetary) is the language of risk tolerance or conversely risk appetite with respect to the client, the firm and the individual tax practitioner. In a highly regularised environment, Power (2007) argues that these risk judgements become homogeneous. Furthermore, Short (1992) finds that risk decisions are shaped by organisational preferences.

Attitudes to risk are latent and can only be measured indirectly, typically through self-reporting or constructed preferences. The most popular method in economics is the Holt and Laury (2002) mechanism, which asks participants to make risk decisions under various scenarios (Charness et al., 2018). The assumption under this method is that participants will be consistent in their attitude to risk across the scenarios. This assumption is common to Von Neumann-Morgenstern’s utility theory, a seminal consumer choice model. In utility theory, the axioms of rational behaviour include transitivity, whereby the individual’s preferences are consistent across the options. However, human nature is not always rational or specifically, consistent. Prospect theory addresses some of the shortfalls of utility theory, capturing individuals’ loss aversion and the certainty effect. It is described by Barberis (2013) as “widely viewed as the best available description of how people evaluate risk in experimental settings” (p. 173), however, they acknowledge the difficulty in applying it in a real-world setting. Work by Tversky and Kahneman (1992) shows that risk framing has a significant impact on risk preferences and decision-making. In a series of laboratory experiments, Tversky and Kahneman (1974) find that individual decisions under uncertainty are much better described by heuristics, than objective, rational risk-based decisions. Similarly, Simon (1955) argues that economic market participants prefer to make their decisions based on “rules of thumb” than computing optimal decisions. As mediators between the taxpayer and the relevant authority, tax practitioners are at the forefront of these risk decisions.

This brings us to the role of the professional advisor and the significance of their risk attitudes on their practice. There has been some limited work in looking at the risk attitudes within the accounting profession (Martin and Previts (1982), for example Helliar et al. (2002) compare the risk attitudes of Scottish chartered accountants with UK managers. Nofsinger and Varma (2007) surveyed 100 financial planners in the USA to assess their risk aversion and preferences. They categorised the group into intuitive and analytical thinkers and found the former more risk averse than the latter and that the risk preferences of intuitive thinkers were consistent with prospect theory (loss aversion and certainty bias). They found that more experienced professionals tend to fall into the intuitive group. While intuition based on implicit knowledge and experience can be hugely valuable in complex and strategic decision-making, Hodgkinson and Healey (2011) find that it is necessary to recognise when to rely on intuition as a form of cognitive bias. Kahneman and Klein (2009) find that “reliably skilled intuitions are likely to develop when the individual operates in a high-validity environment and has an opportunity to learn the rules of that environment” (p. 521). They emphasise the significance of the regularity and predictability of the environment and warn about over-reach into decision areas where the individual has no expertise. As cognitive biases such as overconfidence, anchoring and loss aversion can be systemic rather than random (Bellé et al., 2018), it is important to examine the related demographics.

Age, gender, nationality and employment are among the key demographic factors examined for their relationship with risk decision-making (Weber and Hsee, 1998; Picazo-Tadeo and Wall, 2011; Eckel et al., 2012; Nga and Ken Yien, 2013; Rieger et al., 2015). Whereas gender differences in risk preferences are an assumed stylised fact in economics and psychology, there are studies that concur (Eckel and Grossman, 2008; Ebert and Wiesen, 2014) and those that dispute this assumption (Harrison et al., 2007; Nga and Ken Yien, 2013; Filippin and Crosetto, 2016). March (1988) find variation in risk attitudes by age and seniority. Eberhardt et al. (2019) found that older adults’ wealth of experience and associated lower levels of negative emotion, led to better financial decision-making. Interestingly, Harrison et al. (2007) find that risk aversion declines over the age of 40.

Studies in professionalism, have examined the tendency for tax professionals to have attitudes to risk homologous to their organisation (Killian et al., 2023). Similarly, studies have suggested that recruitment of tax professionals is biased towards individuals conveying similar attitudes to risk (Gebreiter, 2020). Stage of career is important. Gilad and Kliger (2008) find that professionals (investment bank executives and accountants) are more influenced by priming (situational manipulation) than economics students who tended to rely more on analytics rather than intuition.

It is important that we do not view accountants or tax practitioners as a homogeneous group. The particular role of the tax practitioner (e.g. compliance, planning or consultancy) and the nature of the firm bring different incentives. A study by Martin and Previts (1982) of Certified Public Accountants (CPAs) in the USA found the size of the firm to have little relevance for an individual CPA’s risk attitude. Depending on the context of the accounting or tax practitioner role, they may be subject to different commercial, legal and ethical incentives. The potential risks faced by tax practitioners in their daily work include threat of litigation, potential penalties, threat of audit and reputational loss (Alm et al., 2022; Carter and Mahallati, 2019; Goldswain, 2001; Killian et al., 2022; Mendoza et al., 2017). Interestingly, a further element of the study of financial planners (Nofsinger and Varma, 2007) looked at customer retention. They found that financial planners willing to take risks in the gain domain showed a higher propensity to retain or gain clients. However, it is clear from these studies that navigating risks is a key component of the tax practitioner role.

Bringing together the literature threads of gender and financial services professionalism, an interesting study by Bollen and Posavac (2018) compare asset allocation recommendations by gender for finance professionals versus graduate business students. Their original hypothesis of differences in risk allocation by gender of the financial advisor used literature evidencing that men exhibit higher risk tolerance than women (Byrnes et al., 1999; Barber and Odean, 2001; Dwyer et al., 2002; Agnew et al., 2008). However, they found no difference in the riskiness of the allocation by gender for the finance professionals. Interestingly, they found that male students selected a riskier asset allocation. This suggests a moderating effect of the professional context. A further study by Sapienza et al. (2009) finds that students entering the finance industry have a higher risk tolerance than their peers. Note, however, that they specify a “risky career in finance (e.g. investment banking or trading)” (p. 15268). Differences in risk attitude by gender in accounting and taxation roles pre- and post-employment are under-explored.

Risk decision-making is influenced by many factors and contexts within financial services. While we have explored interesting research in relation to accountants and other finance professionals, the risk-attitude of tax practitioners is under-explored. The important role of tax practitioners in moderating aggressive tax behaviours as a means of reducing societal inequalities (sustainable development goal – SDG 10) warrants further exploration of the risk attitudes of this cohort.

Research methodology

In response to the call by Radcliffe et al. (2018) for further empirical work examining the experience of tax practitioners, we focus on practitioners’ risk attitude and their perception of their firm’s risk attitude. We adopt an approach to exploring tax practitioner risk through a prism of critical realism, whereby we understand that “human agency lives its life making sense of reality, to varying degrees and with differing consequences”, and furthermore individuals, and in our case tax practitioners, “bring with them an accumulated understanding of [their reality] that necessarily shapes all subsequent interactions with it and that organises or structures these interactions and thereby the accumulated experience carried forward” (Roslender, 2016, p. 81). This critical realism approach underpins our methodological design to understand the factors that tax practitioners perceive as influencing their day-to-day work decisions. We have elected to follow a mixed methods approach to obtain “a fuller picture and deeper understanding of [this] phenomenon” (Johnson et al., 2007, p. 119).

Our primary data source is a survey of tax experts. The survey was constructed and hosted on Qualtrics, which is GDPRcompliant, and ethical approval for the research was granted through our affiliated university’s Research Ethics Committee. The survey was designed to assure participants of their anonymity but requested some basic demographic information including age category, gender, country of employment and level of experience. The survey was piloted in early 2017, and the final survey was disseminated during March to November 2017. During this time, 1,351 responses were collated. In the preliminary analysis of the survey data, partial responses were eliminated, those surveys completed in an unrealistic time frame (Curran, 2016), and through manual scanning, the final data set was reduced to 1,061 usable responses. In the final 1,061 data set, 59 individual countries were represented. There was a split between professional practice (n = 662); corporate sector (n = 278), public sector (n = 94) and other (n = 27).

To garner the self-perception of tax experts to a range of factors around their day-to-day decision-making in their tax work, we asked a series of questions, using Likert-scale attitudinal statements. Table 4 details the specific Likert-scale questions asked in the survey, along with the individual attitudinal statements. The Likert attitudinal statements were informed inter alia by the literature ’relating to accounting/tax professionals and media (Carnegie and Napier, 2010; Sikka, 2010), ethics (Tsakumis et al., 2007; Doyle et al., 2013), regulation (Canning and O’Dwyer, 2016; Loft et al., 2006), firm culture (Taylor and Richardson, 2012; Hasan et al., 2016), managerial influence (Bamber et al., 2010) and gender (Anderson-Gough et al., 2005).

For the purpose of this study, we are primarily interested in the variables associated with Risk attitude. This included two key variables that asked survey participants to self-evaluate their personal and firm attitudes to risk using a five-point Likert scale:

1. Q11. In your day-to-day tax work, how influential are the following factors?

   • Personal attitude to risk

2. Q12. To what extent do the following descriptions apply to the leadership/ethos of your organisation?

   • Risk averse

In the second phase of data collection, we conducted semi-structured interviews with 68 individual tax practitioners, working across 11 different countries. Following from Everett et al. (2015), we used the “quantitative indicators [identified in Phase One (the Survey)] to describe the status quo and provide a starting point for a more in-depth qualitative analysis” (p. 41) through a series of interviews. The interviews were explorative in nature, and we used open-ended questioning across a range of factors that may influence tax practitioners’ daily work, inter alia questions to examine the risk appetite of the tax practitioner. Following O’Dwyer (2011), all interviewees were sent a list of the range of topics that would be covered ahead of the interview (p. 1242), and during the interview process the interview participants were encouraged to speak openly across the range of interview topics identified at the outset.

Using semi-structured interview technique, we were able to gain “privileged access” to tax experts and to garner a deeper understanding of their “basic experience of the lived world”(Kvale and Brinkmann, 2009, p. 32). The interviews were conducted during the period May to November 2018. These practitioners worked in a range of employing organisations, from small firms/sole traders to “Big 4” international firms. We had practitioners at varying levels of career stage, from trainees to managing partners. The interviews were conducted face-to-face primarily (n = 58), with the remainder conducted online and on the telephone. The average duration of the interviews was 57 min in length, with a total of over 64 h of recorded material. See Table 1 for full detail of the demographics of interviewees. The interviews were recorded, professionally transcribed and analysed using Nvivo^© software. The interviews were coded inductively through iterative listening (O’Dwyer, 2004), and thematic analysis was used to analyse the interview data (O’Dwyer, 2011).

In the next section, we present our research findings.

Findings

Survey findings

Personal and firm risk — demographics.

In this section, we present the results of various statistical tests carried out on the two key risk attitude variables; Personal Attitude to Risk and Firm Risk Aversion, and their relationship with the demographical information provided. Tables 2 and 3 show the breakdown of the results for the two key risk attitude variables.

We examine how these variables behave when cross-referenced with participant demographic information. We further examine how these variables correlate with other personal and firm-specific characteristics and influencing factors using questions 11 and 12 from our survey (see Table 4).

Firstly, we examine the relationship of the risk attitude variables with gender. We found a very small deviation in the mean Likert score for “Personal Attitude to Risk” by gender: 4.02 Male, 4.00 female. For “Firm Risk Aversion”, there was no gender difference in the average Likert score. We use the Mann–Whitney U test to examine if “Personal Attitude to Risk” varies by gender. We find that the hypothesis that the distribution of Likert scores for “Personal Attitude to Risk” are the same across gender categories is retained, Z score of 0.139 and a two-tailed p-value of 0.889. Secondly, we examine the relevance of age categories. The age categories are defined as follows: less than 30, 30–39, 40–49, 50–50, 60–69 and 70 plus. When we compare the mean Likert score for “Personal Attitude to Risk”, we find significant deviations across the age categories. The younger age categories tend to have a lower mean Likert score than the higher age categories, reflecting a lower significance attached to risk in the younger cohort. Table 5 shows the average Likert score by age category for “Personal Attitude to Risk”. We use the Kruskal–Wallis test to examine if “Personal Attitude to Risk” is affected by age category. We find that the distribution of Likert scores for “Personal Attitude to Risk” are the same across each age category, χ²(4) = 8.632, p = 0.071.

Interestingly, performing the same analysis for the variable “Firm Risk Aversion”, showed less variation in mean value by age category. Table 6 shows the results.

We deploy the Mann–Whitney U test on combined categories “Less than 30” and “30–39” (Collectively, “Under 40”) compared to combined categories “40–49”, “50–59” and “60–69” (Collectively, “Over 40”). The Mann–Whitney U is used to compare differences between two independent groups, with the null hypothesis that the two groups have the same distribution. The Mann–Whitney U test rejects the hypothesis that the distribution of “Personal Attitude to Risk” is the same for participants “Under 40” as for those “Over 40”. Whereas, for “Firm Risk Aversion”, the Mann–Whitney U test retains the null hypothesis that the Likert scores follow the same distribution for participants “Under 40” as for those “Over 40”. These tests are carried out at a significance level of 0.05. Furthermore, when we condition by age, the partial correlation between “Personal Attitude to Risk” and “Firm Risk Aversion” is equivalent to the zero-order correlation (0.258**).

Next, we look at Career Stage. We group our participants into “Early Career” and “Established”. The Mann–Whitney U test retains the null hypothesis of the same distribution for both “Personal Attitude to Risk” and “Firm Risk Aversion” regardless of career stage. It is interesting that career stage was not found to be an influencing factor when we have seen that age is an influencing factor for “Personal Attitude to Risk”.

In keeping with our study’s interest in the “site of influence”, we examine the impact of the variable Firm Size. As described above, Martin and Previts (1982) find no impact of firm size on CPA’s risk attitude. Our survey results show the mean “Firm Risk Aversion” value is higher for larger firms (Medium and Large) (see Table 7).

We grouped (micro and small) and (medium and large) and tested the distributions using the Mann–Whitney U test (0.05 significance). The results retain the null hypothesis of the same distribution across the two groups. Furthermore, conditioning by firm size results in a partial correlation only marginally different from the unconditional correlation between “Personal Attitude to Risk” and “Firm Risk Aversion”(conditioned correlation: 0.261**, zero-order correlation: 0.259**).

Our final set of demographic variables include employment sector (corporate, public, advisor or “other”), discipline (accounting, tax or law) and whether the firm has international business. The mean “Personal Risk Attitude” and “Firm Risk Aversion” is the same for each employment sector and the Mann–Whitney U does not find any distributional differences across sectors. Again, the Mann–Whitney U retains the null hypothesis of the same distribution (Personal or Firm Risk Aversion) regardless of discipline type. There is a small difference in mean for Firm Risk Aversion for one-country versus multinational (see Table 8).

However, the Mann–Whitney U test retained the null hypothesis of the same distribution. Personal risk attitude increased incrementally for multinational firms versus one-country firms, but the Mann–Whitney U test found they had the same distribution.

Firm and personal variables.

The survey nests the question of the level of influence on Personal Attitude to Risk within a series of other potentially influential factors at a personal level (Q11). We examine the level of correlation of the Likert scores for “Personal Attitude to Risk” with each of these personal influence factors. We use the non-parametric Pearson correlation for our analysis. See Table 9 for results. Note that the highest correlation for “Personal Attitude to Risk” is observed with Personal Reputation at 0.469 (significant at the 0.01 level). See Table 10 for a full table of results for firm risk correlated with personal attitude to risk.

We then examine the relationship between individuals’ self-evaluation of the influence of their Personal Attitude to Risk and their evaluations of the applicability of leadership/ethos characteristics to their firm (Firm Ethos Q12). Perhaps unsurprisingly, we find the strongest Pearson correlation with the firm characteristic Risk Aversion at 0.258 (note this is a moderate correlation, but is strong relative to other correlation values). The characteristics Ethical, Reputation Conscious and Pragmatic had the next highest correlation values at 0.246, 0.239 and 0.235, respectively (see Table 11). When we analyse the ranking of each of the variables under day-to-day influencing factors (Q11), firm ethos (Q12), we find the ranked Personal Reputation (a variable in Q11) has the strongest correlation with the ranked “Personal Attitude to Risk” at 0.283 (0.01 significance level). This exceeds the correlation with ranked “Firm Risk Aversion” at 0.143.

We perform a similar analysis for the “Firm Risk Aversion” variable (see Table 12). Under day-to-day influencing factors (Q11), aside from the relationship with “Personal Attitude to Risk” (as above 0.258), the next highest Pearson correlation is 0.199 for Risk of Sanction by Taxing Body. Furthermore, the relationships between “Firm Risk Aversion” and other Firm leadership/ethos characteristics (Q12) show low levels of correlation. Note that the variables “Ethos of organisation” and “Workplace ethos” are designed to measure the same construct and are used as an equivalent-forms reliability measure (Correlation 0.601**).

We acknowledge the limitations of univariate analysis which is not designed to identify causality and may not provide a complete understanding of the complex interdependencies between variables. However, it does provide useful insights into relationships within the data that can then form a basis for further research.

Interview findings

Using the findings from the survey as a guide, we examine the emergent themes in the interviews to develop a better understanding of the tax practitioners’ lived experiences around risk (Everett et al., 2015). The initial thematic analysis through the inductive coding (O’Dwyer, 2004, 2011) of risk considerations identified can be seen in Figure 1. Certain themes were classified both as risk identified to the tax practitioners personally, and as a consideration for the firm (namely, reputation, legislation/governance, litigation and client relationships).

Theme 1: Personal risk.

Risk is in an inherent aspect of tax practitioners work (Neuman et al., 2020). Our findings support existing research, whereby our interview participants acknowledge that risk is a bona fide part of their work, with specific reference to reputation risk (Graham et al., 2014; Kadous et al., 2008):

I would be concerned about reputation but as far as the risk of losing an audit if we recommended something is always possible that the client can come back and say, these are my finances involved and hit us with a malpractice suit and that has it”s own ramifications that quite frankly I’m terrified of [IP63 – Partner].

Extant research explores the considerations for tax practitioners in managing their client relationships (Bobek et al., 2010; Fogarty and Jones, 2014). The burden of potential damage to their client relationships was a forefront consideration of the practitioners interviewed in this study, irrespective of their career stage:

[…] like if you send something out and you’re like, oh God I hope that’s OK. You know, obviously it’s always reviewed and everything. But in the back of your mind you’re kind of like, and especially I suppose from a relationship point of view, you’re always worried you wouldn’t want to affect a client relationship or anything and just kind of, in that respect I suppose [IP41 – Tax Junior].

I suppose you can go so far with trying to please them but you’d listen to what they want but you really highlight and stress there is definitely a risk here, there’s risk risk risk [IP51 – Tax Senior].

These practitioners also speak about the threat of potential litigation risk, in line with existing research on the threat of litigation risk for tax practitioners (Beasley et al., 2006; Kadous et al., 2008; Doyle et al., 2009); Donelson et al., 2022):

[…] as a profession I see [litigation] is an increasing risk because [the work] is becoming very complicated, more and more complicated and also there is also more and more economic pressure, people are not willing to pay any more like also getting less and then they question about quality and then in the end you want to present something which has valuable impact [IP46 – Tax Senior].

There is no other field like tax law where, say tax advisors or lawyers are sued so often that it’s much, it happens much more often than in any other area of law, because it’s, the risk in tax laws always there will be a tax audit on what you have advised previously so everything you do will be checked again later and if you do a contract [IP18 – Tax Partner].

However, the tax practitioners are quick to reallocate the risk burden. From junior practitioners focusing on more senior staff being responsible for final sign off: “someone higher than myself signing it off” [IP58 – Trainee], to more senior staff placing the ultimate responsibility on management:

If it was an advice 100%, yeah, I get it looked at by somebody either like I would bounce it off somebody else or I’d give signed off by somebody higher than me [IP51 – Tax Senior].

Like if I was giving advice I wouldn’t just be sending it out, there is a four eye review, here everything looked at, I just can’t send off an email with what I think it is, it’s discussed and it’s like I will always go to the director or the partners if I need to. [IP55 – Tax Senior].

Interestingly at more senior levels, the practitioners are quick to state that it is the client baring the ultimate risk at end of the day:

We push back on the client then we are like look we have caveat here that this is a very grey area you should reach out to revenue to confirm this and at the end of the day it falls back on the client in that regard so I think that our clients respect that. [IP44 – Tax Manager].

[…] it does come down to the risk appetite of the client at the end of the day [IP47 – Tax Manager].

[…] the client know the risks of what they are doing and what they want to do. It’s not something that I would really like to be worried about in day-to-day job [IP55 – Tax Senior].

The practitioners interviewed spoke about mitigating their risk through various methods from the use of caveats:

I would say that most clients just trust your advice, but then some will push the boat out a little bit more and then it just comes down to a very highly caveated memo [IP47 – Tax Manager].

[…] that’s what [the clients] wanted to do, so, we just had to have a caveat ourselves to say that we do not agree with this approach and they had to agree to and then we could file the return for them [IP55 – Tax Senior].

To keeping up to date with legislative changes:

[…] keeping on top of the current state of the law, reading the industry publications, periodicals, the daily tax reports that discuss cases that have come out, changes to the regulations, what’s on the horizon what other firms have done in their practice, what they are seen firsthand [IP63 – Partner].

Theme 2: Firm risk.

However, an-overarching theme in the conversations with tax practitioners was the significance of the risk appetite within the firm in terms of the approach the practitioners adopt in their decision-making:

Personally I kind of have that core values anyway but it’s certainly reinforced by the firm and the training [IP48 – Tax Manager].

There is definitely risk involved [in tax work] and we [as a firm] are very aware of that and I’m sensitive to it [IP63 – Partner].

It’s firm level but, you know, the firm would make you well aware of all kind of lot of risks would be out there for us [IP26 – Tax Manager].

We as a law firm, we usually, when we make a plan or tax planning then we also, we are involved also in the implementation of that and that’s why we prefer kind of simpler and more straight forward kind of things and also tax planning wise we don’t usually propose the most aggressive tax schemes for the clients [IP5 – Partner].

We have very strong firm values and ethics which again personally for me it fits my own very well so they are like identical [IP36 – Partner].

Doyle et al. (2009) note that ethics in tax is perceived differently by tax practitioners depending on the firm size, with a recognition that “ethical issues faced by large international firms and smaller, locally based tax practices are different and are typically dealt with differently” (p.637). The practitioners interviewed in our study feel that perhaps big firms are more risk adverse than smaller firms:

[Big 4 firms] are very risk averse I suppose and they are much more conservative because you know, I suppose they have a lot more International clients […] […] they are much more regulated internationally into what they can do. So I suppose they are more risk-averse I think [IP6 – Tax Senior].

This practitioner describes her understanding of why firm size might contribute to the difference by highlighting that litigation and reputation may have higher impact on bigger firms rather than small firms:

[the small firm] don’t have the same kind of legal issues, engagement letters, they weren’t concerned about risk to the firm’s reputation, which you would have in the big [firms]. I would say in the small firm they certainly, thinking about it now, they would push the boundaries a lot more than the likes of [Big 4 firm A] or [Big 4 firm B] [IP48 – Tax Manager].

Our findings on risk and firm size, echo the accounting literature which shows that Big Four firms tend to display more ethical sensitivity than practitioners working in smaller firms (Doyle et al., 2014; DeAngelo, 1981; Cruz et al., 2000; Yetmar and Eastman, 2000).

In the next section, we discuss our findings and present our conclusions.

Discussion and conclusion

Although ostensibly technical, there is a growing recognition of accounting and tax practices as social and moral practices (Hopwood and Miller, 1994; Carnegie et al., 2021). In this context, the neutrality of the practices cannot be assumed (O’Connell et al., 2015). Value judgements and risk-related decisions are often influenced by the organisational setting (Power, 2007). The tax practitioner navigates the legal, compliance and ethical context, while also ensuring the tax preparation meets the needs of the client. The risk appetite or tolerance of the client, the tax practitioner and the firm influences the approach taken. This intermediary role of the tax practitioner warrants the exploration of their risk attitude and the significance of firm socialisation on its moderation or exacerbation.

Our study brought together a wide-scale international survey with 1,061 tax experts across 59 jurisdictions and 68 semi-structured interviews with tax practitioners working in 11 different countries. The context of the survey tax experts was split between professional practice (n = 662); corporate sector (n = 278), public sector (n = 94) and other (n = 27). Their self-perception of their personal risk attitude and that of the firm were evaluated using Likert-scale attitudinal statements. Using the statistical results from the survey as a prompt, we explored the lived experiences of the tax practitioners as captured by the semi-structured interviews. This helps us to understand better how risk is experienced by the tax practitioner.

Firstly, it is interesting to note that over 74 % of those surveyed rated “Personal Attitude to Risk” as having a “high or “very high” influence on their tax work. For “Firm Risk Aversion”, 63.8 % rated it as having a “high” or “very high” influence on the leadership/ethos of the firm. An interesting finding from the study was that we found no significant gender difference in personal risk attitude or the perception of the firm risk aversion. This is a little surprising at first as many studies have shown significant disparity in risk attitude by gender (Eckel and Grossman, 2008; Ebert and Wiesen, 2014). However, other studies have found little or no gender disparity with the regularisation of the context a significant factor (Filippin and Crosetto, 2016).

We examined the influence of age on both personal risk attitude and perception of the firm risk aversion and found that younger age categories had lower mean Likert scores than older age categories for personal risk attitude. This is in keeping with previous research on risk attitude and age (Eberhardt et al., 2019). However, there was very little variation in perception of “Firm Risk Aversion” when conditioned by age category. This was much more homogeneous which suggests that “Firm Risk Aversion” is clearly understood by practitioners across the age categories. When we condition the correlation between “Personal Attitude to Risk” and “Firm Risk Aversion” and age category, we find this has no impact on the correlation. This is reinforced by the findings in the interviews that the firm makes its employees aware of the prevailing risk appetite.

Our analysis of the survey results showed no significant difference in either “Personal Risk Attitude” or “Firm Risk Aversion” by firm size, career stage or discipline (accounting, tax or law). The lack of difference with regard to career stage was supported in our interview data, with practitioners irrespective of career stage being cognisant of the potential risk in their working relationships with clients (and thus potential impact on future fee income); “there’s risk risk risk” [IP51 – Tax Senior]. There was a small increase in mean “Firm Risk Aversion” for multinational operations versus those operating in one country only. This is also supported by interview participants who indicate more risk aversion from Big 4 firms. This suggests that the latent “Firm Risk Aversion” is clearly communicated and understood by practitioners.

When we examine the relationship between “Personal Risk Attitude” and other personal or workplace variables, we find the highest correlation with personal reputation, followed by “Ethos of organisation”. This confirms the belief that the tax practitioner is impacted by the risk of an improper filing, thus their navigation through the legal, ethical and commercial is fraught with personal risk. This is further evidenced in the interview findings; “hit us with a malpractice suit and that has it”s own ramifications that quite frankly I’m terrified of” [IP63 – Partner]. For the tax practitioner, risk is not an abstract concept but is part of their everyday decisions. Similarly, we examined firm and personal values and their relationship with the variable “Firm Risk Aversion”. After “Personal attitude to risk”, we found the strongest relationship was with “Risk of sanction by taxing body”. This is a layer of exposure specific to the practitioner and is a further reputational risk. This confirms that the tax practitioner is personally exposed in the risk decisions undertaken.

Our research findings with regard to the importance of the firm as a significant “site of influence” for tax practitioners in shaping their risk appetite in their tax work warrants further exploration. Our findings reflect that the firm eclipses other influences of risk such as professional body oversight, public interest and demographic markers such as gender and career stage. Power (2007) suggest that in highly regularised environments risk judgments become homogenous. This suggests that firms are effective in conveying a risk culture. Further studies could explore scenarios and contexts in which risk judgments may vary using experimental research methodologies.

From a policy perspective, this has implications for the importance of ethics policies and training to aid the tax practitioner in navigating complex risk-decisions. The absence of differences in “Personal Risk Attitude” or perceived “Firm Risk Aversion” by gender is interesting. Females are noted to have augmented abilities in ascertaining the risk appetite of others therefore we would expect to see this in the response data. The absence of evidence of a gender difference in this respect is consistent with the lack of consensus in the literature. Together with other demographics for which we saw little or no impact on “Personal Risk Attitude” or “Firm Risk Aversion” this leads to the suggestion that there is a homogenisation of risk attitudes upon assimilation into the firm. This homogenisation of risk appetite potentially dampens differences in attitude to risk, superseding it with the risk appetite of the firm. This may be a benign issue if the risk attitude of the firm is not supportive of tax aggressive behaviour, but the converse is also true.

Figures

Figure 1.

Thematic analysis of interview data

Table 1.

Details of interviews conducted

Interview participant (IP)	Duration (min)	Country	Gender	Position	Firm size
IP1	31	Italy	Female	Partner	Sole trader
IP2	55	Italy	Male	Partner	Large firm
IP3	51	Finland	Male	Partner	Big 4
IP4	59	Finland	Male	Partner	Big 4
IP5	59	Finland	Male	Partner	Large firm
IP6	51	Ireland	Female	Senior/Management	Small/Medium firm
IP7	73	Ireland	Male	Partner	Small/Medium firm
IP8	46	Ireland	Female	Senior/Management	Small/Medium firm
IP9	31	Ireland	Female	Senior/Management	Small/Medium firm
IP10	62	Ireland	Male	Partner	Small/Medium firm
IP11	68	Ireland	Male	Partner	Small/Medium firm
IP12	52	Netherlands	Male	Senior/Management	Large firm
IP13	70	Netherlands	Female	Senior/Management	Large firm
IP14	49	Netherlands	Female	Partner	Big 4
IP15	35	Netherlands	Female	Trainee	Big 4
IP16	42	Germany	Male	Senior/Management	Large firm
IP17	70	Germany	Female	Partner	Large firm
IP18	71	Germany	Male	Partner	Large firm
IP19	48	Germany	Female	Partner	Large firm
IP20	46	Germany	Male	Senior/Management	Large firm
IP21	62	Luxembourg	Female	Senior/Management	Large firm
IP22	34	Luxembourg	Female	Partner	Large firm
IP23	41	Luxembourg	Female	Partner	Sole trader
IP24	72	Luxembourg	Male	Partner	Large firm
IP25	72	Germany	Female	Partner	Sole trader
IP26	63	Ireland	Female	Senior/Management	Big 4
IP27	43	Ireland	Female	Senior/Management	Large firm
IP28	62	UK	Male	Partner	Sole trader
IP29	55	Germany	Male	Partner	Large firm
IP30	49	UK	Female	Partner	Large firm
IP31	53	Ireland	Male	Partner	Small regional
IP32	58	USA	Male	Partner	Sole trader
IP33	40	USA	Male	Partner	Sole trader
IP34	67	Denmark	Male	Senior/Management	Large firm
IP35	60	Denmark	Male	Partner	Big 4
IP36	69	Denmark	Male	Partner	Big 4
IP37	60	Denmark	Female	Senior/Management	Large firm
IP38	82	Denmark	Male	Partner	Large firm
IP39	85	Denmark	Female	Senior/Management	Small/Medium firm
IP40	85	Denmark	Male	Partner	Big 4
IP41	63	Ireland	Female	Recently qualified	Big 4
IP42	60	Ireland	Male	Recently qualified	Big 4
IP43	60	Ireland	Male	Trainee	Big 4
IP44	57	Ireland	Female	Senior/Management	Big 4
IP45	37	Germany	Male	Partner	Large firm
IP46	57	Germany	Male	Senior/Management	Big 4
IP47	60	Ireland	Female	Senior/Management	Big 4
IP48	63	Ireland	Female	Senior/Management	Big 4
IP49	60	Ireland	Female	Senior/Management	Big 4
IP50	39	Ireland	Female	Trainee	Big 4
IP51	31	Ireland	Female	Senior/Management	Big 4
IP52	40	Ireland	Female	Senior/Management	Big 4
IP53	60	Ireland	Female	Senior/Management	Large firm
IP54	42	Ireland	Female	Senior/Management	Large firm
IP55	33	Ireland	Female	Senior/Management	Large firm
IP56	93	Ireland	Male	Partner	Big 4
IP57	58	Ireland	Male	Senior/Management	Large firm
IP58	46	Ireland	Male	Trainee	Large firm
IP59	50	UK	Female	Senior/Management	Big 4
IP60	104	Switzerland	Female	Senior/Management	Large Organisation
IP61	65	UK	Female	Partner	Sole trader
IP62	101	UK	Female	Senior/Management	Small/Medium firm
IP63	64	USA	Female	Partner	Small/Medium firm
IP64	42	UK	Male	Senior/Management	Large Organisation
IP65	53	UK	Female	Senior/Management	Big 4
IP66	59	Germany	Male	Partner	Small/Medium firm
IP67	50	Sweden	Female	Senior/Management	Large Organisation
IP68	50	Sweden	Male	Senior/Management	Large Organisation

Source: Table by authors

Table 2.

Percentage breakdown of responses for Q11 influence of personal attitude to risk on day-to-day tax work

Variable	1	2	3	4	5
Personal attitude to risk (%)	3.3	4.8	17.6	36.1	38.2

Notes:

The Likert scales for Q11: 1 is very low influence; 5 is very high. Note that 1,026 participants completed this question

Source: Table by authors

Table 3.

Percentage breakdown of responses for Q12 firm risk aversion

Variable	1	2	3	4	5
Firm risk aversion (%)	1.8	5.9	28.6	38.2	25.6

Notes:

Likert scale for Q12: 1 is not at all (applicable); 5 is a great deal (applicable). Note that 1,017 participants completed this question

Source: Table by authors

Table 4.

Survey Likert-scale questions

Survey question	Likert scale attitudinal statements
Q11. In your day-to-day tax work, how influential are the following factors?	Personal ethical values; political landscape; media; the public interest; workplace ethos; your professional body; knowledge of regulations; risk of sanction by taxing body; influence of the client; ethos of your organisation; personal attitude to risk; personal reputation; spiritual or religious values; trends in tax regulation; litigation risk
Q12. To what extent do the following descriptions apply to the leadership/ethos of your organisation?	Risk averse; technically strong; ethical; pragmatic; mentoring; cost conscious; innovative; reputation conscious; client-focused; controlling; ambitious; profit focused; gender balanced

Source: Table by authors

Table 5.

The mean Likert scores for personal attitude to risk for each age category

Age category	Mean	%	SD
Less than 30	3.85	16.8	1.108
30–39	3.98	36.1	1.015
40–49	4.09	23.7	0.932
50–59	4.05	17.3	1.057
60–69	4.19	6.2	1.052

Notes:

There were no participants in the 70 plus age category. Note, the number of participants who completed this question is 1,026

Source: Table by authors

Table 6.

The mean Likert scores for firm risk aversion for each age category

Age category	Mean	%	SD
Less than 30	3.80	16.5	0.974
30–39	3.80	36.0	0.967
40–49	3.72	23.9	0.916
50–59	3.88	17.3	0.909
60–69	3.88	6.3	0.984

Notes:

There were no participants in the 70 plus age category. Note, the number of participants who completed this question is 1,017

Source: Table by authors

Table 7.

The mean Likert scores for firm risk aversion for each firm size category

Firm size	Mean “firm risk averse” value	No. of firms	SD
Micro	3.75	137	1.056
Small	3.75	185	0.941
Medium	3.82	90	0.919
Large	3.82	604	0.929
Total	3.8	1,016	0.947

Source: Table by authors

Table 8.

The mean Likert scores for firm risk aversion for firms that operate in one country or are multinational

Operation	Mean	No. of firms	SD
One-country	3.76	443	0.973
Multinational	3.83	574	0.927

Source: Table by authors

Table 9.

The Pearson correlations significant at the 0.01 level (two tailed) between the personal influencing factors and personal attitude to risk

Personal values	Correlation with personal attitude to risk
Personal ethical values	0.324**
Political landscape	0.041 (sig 0.2)
Media	0.078*
The public interest	0.185**
Workplace ethos	0.279**
Your professional body	0.197**
Knowledge of regulations	0.205**
Risk of sanction by taxing body	0.245**
Influence of the client	0.199**
Ethos of your organisation	0.311**
Personal reputation	0.469**
Spiritual or religious values	0.127**
Trends in tax regulation	0.160**
Litigation risk	0.287**

Notes:

**Correlation is significant to 0.01 (two tailed); *correlation is significant to 0.05 (two tailed)

Source: Table by authors

Table 10.

Correlation between personal attitude to risk and firm ethos values

Firm ethos	Correlation with personal attitude to risk
Risk averse	0.258**
Technically strong	0.166**
Ethical	0.246**
Pragmatic	0.235**
Mentoring	0.205**
Cost conscious	0.199**
Innovative	0.159**
Reputation conscious	0.239**
Client focused	0.191**
Controlling	0.11**
Ambitious	0.18**
Profit focused	0.139**
Gender balanced	0.136**

Notes:

**Correlation is significant to 0.01 (two tailed); *correlation is significant to 0.05 (two tailed)

Source: Table by authors

Table 11.

The Pearson correlations significant at the 0.01 level (two tailed) between the firm ethos factors and firm risk aversion

Firm ethos factors	Correlation with firm risk aversion
Technically strong	0.199**
Ethical	0.249**
Pragmatic	0.130**
Mentoring	0.116**
Cost conscious	0.145**
Innovative	0.067*
Reputation conscious	0.201**
Client focused	0.122**
Controlling	0.139**
Ambitious	0.053*
Profit focused	0.074*
Gender balanced	0.117**

Notes: **Correlation is significant to 0.01 (two tailed); *correlation is significant to 0.05 (two tailed)

Source: Table by authors

Table 12.

The Pearson correlations significant at the 0.01 level (two tailed) between the influencing factors and firm risk aversion

Influencing factors	Correlation with firm risk aversion
Personal ethical values	0.196**
Political landscape	0.072*
Media	0.142**
The public interest	0.183**
Workplace ethos	0.155**
Your professional body	0.192**
Knowledge of regulations	0.142**
Risk of sanction by taxing body	0.199**
Influence of the client	0.095**
Ethos of your organisation	0.187**
Personal attitude to risk	0.258**
Personal reputation	0.181**
Spiritual or religious values	0.021**
Trends in tax regulation	0.089**
Litigation risk	0.194**

Notes: **Correlation is significant to 0.01 (two tailed); *correlation is significant to 0.05 (two tailed)

Source: Table by authors