Distrust: A critical review exploring a universal distrust sequence

ABSTRACT

In this article, we review the distrust literature and explore whether the universal sequence for trust as outlined by Dietz (2011) and Dietz and Den Hartog (2006) is also true for distrust. Compared to trust, there is much less research on distrust, although the field has been rapidly developing in recent years. We argue that it is time to explore a universal sequence for distrust to take stock of current knowledge and to focus the empirical and conceptual research. There is sufficient evidence to suggest that such a universal sequence is a valuable framework for distrust research. This analytical exercise also forces us to identify tacit assumptions that frame and guide much of the current distrust (and trust) research. In so doing, we identify two main areas that require more attention: 1) the definition of distrust and its relation to trust and 2) the universal sequences and their dynamics. These findings lead to avenues for further research.

KEYWORDS:

• Distrust
• distrust process
• relation trust and distrust
• trust
• universal sequence

Introduction

Dietz (2011), building on earlier work by Dietz and Den Hartog (2006), argued that there is a universal dynamic sequence for trust: an assessment of trust beliefs, a trust decision with possibly (but not necessarily) a risk-taking action, and finally feedback on the outcome of the trusting action back to the input factors for the trust beliefs. Depending on the context within which the trust encounter takes place, different elements within the steps of the sequence have more weight or are more relevant, but the sequence itself is universal.

With the rise of the perspective that distrust and trust are not simply two ends of the same continuum (Lewicki et al., 1998; Sitkin & Bijlsma-Frankema, 2018), the question emerges whether there is a universal sequence for distrust. Several models for distrust development have been put forward, but these have always appeared in specific contexts or domains, for example e-commerce (McKnight & Chervany, 2001a), strategic interorganisational business relationships (Guo et al., 2017), natural resource management (Emborg et al., 2020), regulatory policy networks (Lee & Dodge, 2019) or specific organisations such as courts (Bijlsma-Frankema et al., 2015). No attempt at generalisation has been made to date.

Research into distrust has been much slower than research into trust, but recently there has been a clear increase in interest in distrust. Nonetheless, gaps remain in our knowledge on distrust compared to trust and on the relationship between trust and distrust processes. The aim of this article is to critically review the distrust literature using the universal sequence for trust that Dietz (2011) proposed. Our exploration suggests that there is also a universal distrust sequence that is structurally very similar to the universal trust sequence. Based on the same types of inputs, an actor forms either trustworthiness or untrustworthiness beliefs. These beliefs influence a decision to (dis)trust and, possibly, (dis)trust-informed actions.

The conceptual exercise we undertake in this study makes three contributions to the (dis)trust literature. First, we formulate a distrust sequence in line with the structure of Dietz’ trust sequence and, in so doing, review current knowledge about distrust. Second, we critically reflect on existing distrust research and on the relationship between the trust and distrust sequences. Third, we critically reflect on the dynamics of the sequences and how these processes may be related. What outstanding issues need to be researched?

We begin by presenting the universal sequence for trust and distrust, in which we briefly describe trust and more in-depth review the distrust literature. Then, we identify areas that require more attention, either because there is insufficient knowledge thus far or because we critically question some of the (implicit) assumptions previous research made. Finally, we also identify avenues for further research into how Dietz’s trust sequence and the proposed distrust sequence may be linked, which also addresses the question of how trust and distrust are related to one another.

The universal sequences of trust and distrust

In response to an article by Reinhard Bachmann (2011), who claimed that there are fundamental differences between sources and hence types of trust, Graham Dietz argued that there is a universal dynamic sequence for trust:

[T]here is always an assessment (however thorough) of the other party’s trustworthiness which informs a preparedness to be vulnerable that, in genuine cases of trust, leads to a risk-taking act. The results of the trusting encounter update the foundational assessment. In other words, there is only one essential ‘trust experience’, and this sequence is it. (Dietz, 2011, p. 215).

This universal trust sequence is depicted in (the top half of) Figure 1. Depending on the substantive content of the inputs, the trustor forms beliefs about the other actor’s (i.e. the trustee’s) trustworthiness. This belief will influence whether the trustor decides to trust or not; when they do, it will lead to trust-informed actions, and the response to these actions will provide feedback to the inputs. In the bottom half of the figure, the distrust sequence is shown, which will serve as the framework for our review of the distrust literature. We start with the core of the process – beliefs, decision and then action – before looking at the inputs and the feedback loop. In each subsection, we first briefly present what the trust literature says for this step of the sequence and then review the distrust literature.

Figure 1. The universal sequence for trust and distrust based on and adapted from Dietz (2011).

Beliefs

Trustworthiness refers to beliefs, evaluations or expectations of another actor that form the basis for trust decisions. On the basis of trustworthiness assessments this phase involves forming expectations about the other actor’s behaviour. The most cited model of trust proposes that the trustworthiness construct includes three dimensions: ability, benevolence, and integrity, also referred to as ABI (Mayer et al., 1995). Another model distinguishes two dimensions of trustworthiness: competence, which concerns expectations of the abilities of the trustee, and goodwill, which relates to expectations of integrity and non-harmful behaviour and merges benevolence and integrity (e.g. Sako, 1992). Trustworthiness beliefs will likely lead to a decision of confident positive expectations about the future behaviour of the other actor (Mayer et al., 1995; Rousseau et al., 1998).

Untrustworthiness beliefs

Untrustworthiness beliefs are manifested in ‘confident negative expectations regarding another’s conduct’ (Lewicki et al., 1998, p. 439). Recent contributions show that the approach of studying distrust through untrustworthiness assessments, similar to studying trust through trustworthiness assessments, may be a promising avenue (PytlikZillig et al., 2016). But studies of untrustworthiness are still scarce.

Some authors proposed three dimensions of untrustworthiness holding opposite valence of the trust subconstructs: incompetence, malevolence and deceit, or IMD (McKnight et al., 2004; Moody et al., 2014; Rusk, 2018). Incompetence is defined as the individual’s belief that the other actor lacks the ability to perform what they want them to perform. It refers to the lack of knowledge, resources or expertise to accomplish what the distrustor needs. Malevolence is defined as the belief in the harmful intentions and ill will of the other actor and assesses how far the other actor will go out of their way for the distrustor’s interests. Deceit is defined as the belief in the dishonesty and duplicity of the other actor, and it refers to the distrustor’s wariness that the other will lie, cheat or steal to gain an upper hand (Moody et al., 2017; Rusk, 2018).

But there are other studies suggesting that conceptualising untrustworthiness as a mirror-image of trustworthiness may not cover the whole range of possible distrust beliefs. They show that distrust also arises in situations of value incongruence, when one believes that the other actor does not accept one’s core values or adheres to incompatible values (Bijlsma-Frankema et al., 2015; Chambers & Melnyk, 2006; Sitkin & Roth, 1993; Sitkin & Stickel, 1996). Such value incongruence does not need to be a result of deceitful intentions. It can result from an open and truthful confrontation about values resulting in a shared understanding that both parties are guided by principles unacceptable to the other. A value-mismatch may also emerge if two actors draw the line of what is ethically acceptable differently in various domains (Chen et al., 2011; Williams et al., 2020). For example, where environmental protection is concerned, partners may be in full agreement, but they may disagree about whether corruption and bribery are acceptable if these constitute the only way to enter a market. Furthermore, expectations regarding values may vary across cultures. In some cultures, displaying unacceptable moral principles in personal conduct may seriously damage business or political careers, while in other cultures an actor’s private actions do not affect their public career (Chen et al., 2011). Finally, there may be cases where actors agree on values but one party is simply not committed enough to actually walk the talk and act on the declared values. Such situations are also likely to be ground for untrustworthiness beliefs.

Decision

The trust decision has the following form: A trusts B to do X in context C. This trust decision involves the willingness to be vulnerable towards the other party’s actions (Mayer et al., 1995). It involves the willingness to be dependent on the other actor to embrace the (irreducible) vulnerability in the relationship (Möllering, 2006). As a result of the trust decision, an actor decides to take a leap of faith (Möllering, 2006): to place trust and to act on it (Li, 2007).

Distrust decision

An untrustworthiness belief will likely lead to a distrust decision (Lewicki et al., 1998; Reimann et al., 2017) in the following form: A distrusts B to do X in context C. Similar to trust, there is no certainty and there is vulnerability. However, in distrust there is no willingness to take the leap of faith as if that vulnerability and uncertainty do not exist (Möllering, 2006). Several authors include in their definitions that the distrustee has the intention to deceive (e.g. Moody et al., 2014; Schul et al., 2008).

Actions

Action lies at the heart of the trust concept (Schoorman et al., 2007). This is precisely why trust is so valued in practice: it enables doing what otherwise would not be possible (Li, 2007; Messick & Kramer, 2001; Möllering, 2006). It focuses on risk-taking behaviour (i.e. the trust-informed action) and the reciprocal process that follows (Korsgaard, 2018). Trust requires continuous work by both actors (Gustafsson et al., 2021; Luhmann, 2017; Möllering, 2006). Vulnerability is suspended and the trustor acts ‘as if certain rationally possible futures will not occur’ (Lewis & Weigert, 1985, p. 969). To flesh out this aspect of suspending vulnerability we add to Dietz’ trust sequence model that risk-taking acts are based on the leap of faith (Möllering, 2006).

Distrust-informed actions

Similarly to trust, the functionality of distrust lies in the actions taken to deal with the uncertainty and vulnerability (Luhmann, 2017). The literature suggests that distrust, similar to trust, is an active state that is manifested through specific actions (Bijlsma-Frankema et al., 2015; Cook et al., 2005; Schul et al., 2008). Here, however, the trust and distrust processes start to diverge. When trusting, actors are willing to take a leap of faith and engage in interaction with the other actor, accepting the vulnerability and uncertainty because they have positive expectations about the other actor’s future behaviour (Möllering, 2006). Conversely, when distrusting, actors are not willing to take that leap of faith and will generally try not to engage in an interaction with the other actor or will make safeguarding efforts aimed at protecting themselves from the expected harmful behaviour (Bijlsma-Frankema et al., 2015; Kramer, 1994, 1996, 1999; Lewicki et al., 1998; McKnight et al., 2004; Saunders et al., 2014; Sitkin & Roth, 1993; Sitkin & Stickel, 1996). They may also refuse to provide information or may resist influence (McKnight & Chervany, 2001b). Distrust-informed actions may include involving institutions or third parties or changing the nature of the relationship (for all three, see the below section on inputs). However, alternatives may also sometimes have a more pathological flavour, leading, for instance, to over-vigilance or litigiousness (Sztompka, 1999). Despite this divergence between trust and distrust, the basic structure of the universal trust sequence is also followed in the distrust sequence.

Inputs

Inputs are crucial in the sequence, especially since most of the variety in trust and distrust processes is based on variation in the inputs. While the distrust literature is not as far advanced as the trust literature, very similar inputs for distrust have been proposed (Guo et al., 2017; McKnight & Chervany, 2001b; Patent, 2022; Sitkin & Bijlsma-Frankema, 2018). We identify five types of inputs (slightly diverging from Dietz), starting with (1) the trustor and (2) the trustee and then zooming out to (3) the nature of the relationship and (4) domain-specific concerns (i.e. A trusts B to do X). This dyadic relationship takes place in (5) a context that can facilitate or thwart trust (i.e. A trusts B to do X in context C).¹

(Dis)trustor’s predispositions

Predispositions are persistent views that people hold irrespective of the other people involved and across contexts. A trustor’s predisposition or propensity to trust others is part of their individual character (Mayer et al., 1995); both nature (biological factors) and nurture (socialisation, education and previous trust-related experiences) affect it. A predisposition to trust impacts the decision to trust: two people with different tendencies to trust can make two different decisions in what is otherwise the same situation. Hatzakis (2009) has linked predispositions to attribution styles. People with a high propensity to trust are more likely to initiate new relationships, so they have more chances to test the assumption of trust.

Distrustor’s predisposition to distrust

A predisposition to distrust refers to the extent to which one displays a consistent tendency to be unwilling to depend on general others across a broad spectrum of situations and persons (Guo et al., 2017; McKnight & Chervany, 2001b). It includes the belief that people are not honest, benevolent, competent and predictable. Elements that contribute to the predisposition to distrust are, for example, personality characteristics of high Machiavellianism (Dahling, Whitaker, & Levy, 2009), paranoid thought patterns (Kramer, 1994), or strong motivations to attend to harmful stimuli (Locascio & Snyder, 1975; Marr, Thau, Aquino, & Barclay, 2012). As in trust, actors’ predisposition to distrust reflects the general attitudes that they have towards other humans and the world around them.

(Dis)trustee’s characteristics

Characteristics of the other actor (B) are used to assess their trustworthiness, as perceived by the trustor. Because the trustor uses this information as input for their trust beliefs, it is their perception of these characteristics that counts and not any objective assessment. By examining information about the trustee’s past behaviour, the trustor extracts clues about characteristics of the trustee that guide their future behaviour and are thus relevant for actor A’s belief about actor B’s trustworthiness.

Characteristics of distrustee

Characteristics of the distrustee are similarly used to assess their untrustworthiness. Several specific actions have been identified as possibly leading to perceptions that the actor may be incompetent, malevolent or deceitful, thus forming foundations for untrustworthiness beliefs. Among such actions, are, for example incivility (Darley, 2004; Gill & Sypher, 2010) or contract violation and contract breach (Rousseau & McLean Parks, 1993). These actions may lead to perceived falsehoods (Simons, 2002) and perceived value incongruence with the other party (Bijlsma-Frankema et al., 2015; Chambers & Melnyk, 2006; Sitkin & Roth, 1993) which are grounds for assessing the other party as untrustworthy.

Nature of relationship

The nature of the relationship refers to the levels and referents of trust and the degree of (inter)dependencies (Fulmer, 2018). Relationships vary in the levels and referents of trust: they can be between individuals, organisations or other collectives or cut across these dimensions (Fulmer, 2018). Whether the subject of the study is an isolated dyadic relationship or a relationship embedded in a wider network is relevant, as trust relations unfold differently in dyads compared to networks (Svare et al., 2020). Furthermore, the degree of interdependence and intensity of the relationship are relevant, as these determine whether relationships are ‘deep’ or ‘shallow’ (Sheppard & Sherman, 1998). Existing dependencies between the parties may foster trust building (Lazzarini et al., 2008; Zollo et al., 2002).

Nature of distrust relationship

Similar to the nature of the trust relationship, characteristics like the levels and referents of distrust and the degree of (inter)dependencies are also relevant to the distrust relationship. However, we have found no distrust research that explicitly addresses the nature of the relationship in general. There are, however, several professions whose nature dictates suspicion and possibly even some degree of distrust without it becoming pathological, such as controllers, auditors, state inspectors or secret service organisations (Siebert & Czarniawska, 2020; Six & Verhoest, 2017; Sztompka, 1999). Often there are formal dependencies involved, such as legal obligations to provide information to controllers, auditors or inspectors, while at the same time these professions are dependent on the other party to honestly provide them with all the information they need (Six & Verhoest, 2017).

Domain-specific concerns

Domain-specific concerns refer to the X in the formula A trusts B to do X in Context C. Trust is domain-specific, which means that A may trust B to do X but not to do Y (Mayer et al., 1995; Schoorman et al., 2007; Zand, 1972). Domain specificity has different boundaries and meanings for different trustworthiness dimensions. It is most explicit and clear in relation to ability, when A trusts B to be competent to complete certain tasks, but not others (Levin & Cross, 2004; Mayer et al., 1995; Schoorman et al., 2007).

Domain-specific concerns for distrust

The issue of domain-specific concerns is especially poignant for distrust. Some researchers propose that distrust is always pervasive (Bijlsma-Frankema et al., 2015; Sitkin & Roth, 1993; Sitkin & Stickel, 1996). Pervasiveness implies that negative perceptions in one domain of the relationship have spread to other domains. For these authors, distrust, by definition, involves pervasiveness of negative perceptions and negative expectations towards a distrusted party in all situations. Empirical observations, however, indicate that distrust may not necessarily always be pervasive. If it were, then what would we call a situation wherein an actor distrusts someone in one domain but trusts them in another? We develop this issue further in the discussion section.

Institutional context factors

Institutional factors comprise the context (C) in which the dyadic relationship is embedded. They play a key role in building and maintaining institution-based trust, first identified by Zucker (1986) and defined by Bachmann (2018, p. 220) as ‘trust between two actors whose interactions are constitutively and deeply embedded into the institutional arrangements that surround them’. In the trust literature, different forms of institutions have been distinguished. They can be either formal or informal (Bachmann, 2001), and they can be rules or actors acting as third parties (Nooteboom, 1999). Laws and organisational procedures are examples of formal rules, and cultural or social norms are examples of informal rules (Bachmann, 2001, 2018). Inspectorates and accountants are examples of formal third parties (Six & Verhoest, 2017). Informal third parties can include anyone who steps in to help other actors (re)build trust (Nooteboom, 1999).

Whether formal rules actually build trust depends on how trust and control are related. Empirically, two relations are observed. Trust and formal rules are substitutes when formal rules ‘crowd out’ the intrinsic motivation to behave cooperatively, which is an important component of our trustworthiness judgements (Malhotra & Murnighan, 2002; Molm et al., 2000). For example, Sitkin and Roth (1993) observed that contracts foster compliance but do not create any commitment on the part of the trustee toward the trustor. On the other hand, trust and controls, such as formal rules, may complement and strengthen each other (Bentzen, 2019; Vallentin & Thygesen, 2017). Weibel and Six have theoretically identified which conditions need to be met for trust, control and intrinsic motivation to strengthen each other (Bentzen et al., forthcoming; Six, 2018; Weibel, 2007; Weibel & Six, 2013).

Informal rules may contribute to trust building by inducing the ‘logic of appropriateness’ (Farrell, 2009; March & Olsen, 1989; Sitkin & George, 2005). Informal rules are more durable than formal rules (Sztompka, 1999). They are also more ambiguous than formal rules and thus create more room for choices that can signal trustworthiness and commitment to the relationship (Farrell, 2009).

Third parties also provide opportunities to develop (Nooteboom, 1999) and repair (Gillespie & Siebert, 2018) trust. They achieve this by embedding relationships and thus ‘vouching’ for the trustworthiness of the other actor. In the language of our formula, this would be stated as follows: if A trusts D and D trusts B, then A is more likely to also trust B. In such a triangle, D serves as a third party to the trust between A and B. Key in this role are mechanisms of transference (Bachmann et al., 2015) and reputation spill-over (Gillespie & Siebert, 2018). Informal third parties are relatively rare at the organisational level or societal level, but they do exist. At the interpersonal level in a work context, when people are dependent on each other to achieve their work goals but may experience trouble and suspicion causing them to wonder whether they can continue to trust, they may seek out third-party individuals informally to help resolve the issue.

Institutional context factors for distrust

In distrust processes, the role of institutional context factors is to reduce distrust and the vulnerability that distrustors face. Conceptual research has argued that institutional context factors that build trust may be different from those that reduce distrust (Cook et al., 2005; Guo et al., 2017; Lumineau, 2017). Empirical research, however, appears not to have made that distinction yet. For example, while similar to trust types of institutions may be distinguished for distrust, their effects on reducing distrust are as yet unclear since research so far has been conceptual rather than empirical. Exceptions are studies that show that institutional factors actually created or exacerbated distrust (e.g. Bijlsma-Frankema et al., 2015). Formal rules, the first type of institution identified, may reduce or eliminate distrust by protecting actors against exploitation and making the cost of defaulting more costly or even prohibitive (Cook et al., 2005; Lumineau, 2017). They aim to reduce the vulnerability and uncertainty to such an extent that the first actor becomes willing to take a leap of faith and engage in trust-informed actions (Lindenberg, 2000). Formal rules do this by altering the incentive structure of the other actor so that their behaviour becomes (sufficiently) trustworthy. Contracts are also examples of formal rules and work in the same way (Lumineau 2017). In the public governance literature, formal institutions, often rules and regulations, are sometimes labelled as institutionalising distrust (e.g. Van de Walle, 2010), and they are indeed often experienced as creating and amplifying distrust rather than reducing trust as they should (e.g. Six, 2018).

There is as yet no explicit empirical research into how informal rules (i.e. organisational or cultural norms) interact with distrust. Could informal rules reduce distrust similar to how they can help build trust, or are there different mechanisms involved? The logic of appropriateness mentioned above with trust can also support distrust; the outcome depends on what is deemed appropriate in the specific circumstance.

Moreover, can third parties take away distrust? In the language of our formula, this can be stated as follows: if A distrusts B but trusts E, and B distrust A but trusts E, E may step in as intermediary to reduce distrust between A and B. In such a triangle, E serves as a third party between A and B. This role can only be performed effectively if the third party is trusted by both parties in the relationship. Examples of formal third parties that aim to reduce distrust are some professions, mentioned in the section on the nature of the relationship, such as controllers, auditors or state inspectors (Siebert & Czarniawska, 2020; Six & Verhoest, 2017; Sztompka, 1999). There is no research into how informal third parties help reduce distrust.

Feedback loop

In the original universal sequence model (Dietz, 2011) the feedback is not problematised apart from suggesting that it informs the inputs which then again start the trust process. We elaborate the feedback for trust insofar as it is necessary for our analysis. Feedback elements for trust are very particular to the nature and context of the relationship, and the specific domain and update the trustee’s characteristics (which are input factors). Two bodies of literature have contributed to our understanding of how to increase the level of trust; one on trust-building (Korsgaard, 2018; Nooteboom & Six, 2003; Zand, 1972) and the other on trust repair (Ferrin et al., 2007; Kramer & Lewicki, 2010; Tomlinson & Mayer, 2009). A nascent third body is trust preservation (Gustafsson et al., 2021). For the feedback loop to reinforce trust, it is necessary that the trustee responds favourably and interprets the feedback loop positively, as intended by the trustor. This requires that both actors are willing to be open to each other’s influence. Also, ‘[i]t is necessary not only to trust others before acting cooperatively, but also to believe that one is trusted by others’ (Gambetta, 1988a, p. 216; italics in original).

Distrust feedback loop

Research into the distrust process and thus the distrust feedback loop is limited. When the distrustor is able to exit the relationship, there is no feedback loop and the distrust process ends. However, when the relationship continues, distrust creates a much more vicious and emotional cycle than trust (Lewicki & Bunker, 1996; McKnight & Chervany, 2001c; Six, 2005; Zand, 1972). The viciousness of the distrust cycle has several facets.

First, in distrust feedback, actors are actively trying to protect themselves from harm by the other actor, based on their confident negative expectations about the other actor’s future behaviour. Actors that distrust each other are likely to try to get out of the relationship as much as possible and reduce interaction. Thus, the distrust feedback limits opportunities for learning. For this reason, ‘as a strategy, trust possesses greater scope’ than distrust (Luhmann, 1995, p. 128).

Second, distrusting behaviour often evokes a circle of negative reciprocity (Serva et al., 2005). This happens because in distrust, even otherwise positive signals are often interpreted negatively. The distrusting actor expects that the partner intentionally acts to conceal the truth and mask deception (Kramer, 1994, 1999; Schul et al., 2008). Thus, even signals genuinely meant as trust-building may exacerbate distrust when they are perceived through a distrusting lens.

Third, research in psychology shows that distrust is characterised by different emotions and thought patterns than trust: while trust is characterised by calm, assurance, and security, distrust is generally associated with fear, anger, worry, fear of loss, suspicion, wariness, and doubt (Deutsch, 1958; Dimoka, 2010; McKnight et al., 2004; McKnight & Chervany, 2001c). However, a suspicious and doubtful attitude does not automatically imply distrust. A suspicious actor actively considers multiple hypotheses regarding the intentions of another’s behaviour and cannot yet adopt either a positive or a negative attitude (Kramer, 1999). This lack of decisiveness increases the motivation to acquire more information through a more thorough decision process (Hilton et al., 1993). Both trust and distrust are possible with suspicion. This situation changes when suspicion morphs into paranoia (Kramer, 1998). Paranoid cognitions are ‘delusions and false beliefs whose propositional content clusters around ideas of being harassed, threatened, harmed, subjugated, persecuted, accused, mistreated, wronged, tormented, disparaged, vilified, and so on, by malevolent others’ (Colby, 1981, p. 518). Individuals who exhibit paranoid cognitions are likely to be in (pervasive) distrust.

Perspectives on the trust–distrust relation

Having reviewed the distrust literature using the universal sequence framework, we conclude that there is also a universal sequence for distrust and that both sequences may be visualised together as shown in Figure 1. This raises the issue how trust and distrust are related. There are three main perspectives in the literature on the relation between trust and distrust. In the first, trust and distrust are the two ends of the same continuum. This view of distrust as merely the opposite of trust is implicit in the term itself, with the prefix dis- indicating the opposite of something (Sitkin & Bijlsma-Frankema, 2018). This perspective is predominantly based on assumptions about the functional similarities of both trust and distrust, as they both help people deal with complexity and uncertainty (Cook et al., 2005; Luhmann, 2017). For many decades, this was the dominant perspective, and we still see it in some large-scale studies (e.g. Edelman’s Trust Barometer).

The second perspective recognises that lack of trust is not the same as distrust and introduces the concepts of active trust and active distrust, with a neutral ground in the middle. In this neutral ground, there is insufficient information for either active trust or active distrust. This neutral ground is thus situated in an in-between space of neither distrust nor trust (Cook et al., 2005; Ullman-Margalit, 2004).

The third perspective claims that trust and distrust are distinct yet related concepts (Guo et al., 2017; Lee & Dodge, 2019; Lewicki et al., 1998; Lindenberg, 2000; Oomsels et al., 2019; Sitkin & Bijlsma-Frankema, 2018; Sitkin & Roth, 1993; Van De Walle & Six, 2014). Trust and distrust have different characteristics, relate to different relational aspects (Lewicki et al., 1998), and result from different antecedents (Reimann et al., 2017; Sitkin & Roth, 1993).

There is an on-going debate on which of these perspectives is the most appropriate. There seems to be an emerging consensus that the first perspective - which for many years was dominant - is indeed outdated and distrust needs to be conceptualised separately from trust. However, there is no consensus yet regarding which of the two remaining perspectives is most adequate. Both conceptual and empirical advances are needed to be able to meaningfully address this conundrum. Our review constitutes a step in this direction because it indicates gaps in our knowledge about the relation between trust and distrust. Addressing these gaps is necessary to improve our understanding of distrust. We discuss the most pressing of these unresolved issues in the following section.

Discussion

Our review of the distrust literature and our exploration to construct a universal sequence for distrust leads us to propose that the basic sequence for distrust indeed has a similar structure to that of trust, despite several important differences. In this discussion, we first reflect on our critical review of the distrust literature: how have distrust and untrustworthiness been conceptualised; and how does this relate to conceptualisations of trust? To advance the conceptualizations of trust and distrust, we critically reflect on two terms that in our view are key to advance the literature on trust and distrust: pervasiveness and suspicion. The second issue we address is our proposal that there are universal sequences for both trust and distrust. We reflect on the dynamics of the sequences, identifying several asymmetries between trust and distrust processes, and we address the question of how to move from trust to distrust sequences and vice versa. Considering the possibility that trust and distrust can be present at the same time, we also address the potential implications of a combined trust-distrust sequence.

Distinguishing distrust from related concepts

The key issue emerging from our distrust literature review is the need for more conceptual clarity regarding distrust and related concepts like suspicion, mistrust and trust. This has important implications for different types of distrust. Based on our analysis of distrust literature we propose that there is enough evidence on asymmetries in the stages of trust and distrust sequences to break with the first perspective on the trust-distrust relation -trust and distrust as two ends on one continuum. Conceptualising distrust as a negation of trust may have been popular because of convenient operationalisation (low trust scores are interpreted as distrust), but arguments from existing literature show that it may be too simplistic (cf. Sitkin & Bijlsma-Frankema, 2018). The second and third perspective, however, are still at the beginning to theorise distrust. In both of them there remain important unresolved issues that seriously hinder progress in the study of distrust. An important issue concerns the reasonably widespread agreement that the distrust decision involves confident negative expectations about the behaviour of the other actor (Lewicki et al., 1998; Luhmann, 2017), making the distrustor unprepared to accept the vulnerability and uncertainty that come with taking the leap of faith to trust (Möllering, 2006). Such confident negative expectations in distrust involve either withdrawing from cooperation altogether or, if this is not possible, taking preventive and protective actions against the other actor. If the trust decision is about having confident positive expectations about the other’s behaviour and a willingness to take a leap of faith, do these two scenarios cover all possible situations? What if there is no confidence in expectations in either direction? In the second perspective on the relation between trust and distrust, the lack of either confident positive or confident negative expectations is called the space in between, and in the third perspective it can be conceptualised as the position of both low trust and low distrust. Also, the size of this ‘space of lacking confident expectations’ depends on our definitions of trust and distrust, respectively. To explore this issue, we look at two concepts used in the literature on distrust: pervasiveness and suspicion.

Pervasiveness

A more in-depth exploration of pervasiveness may help to resolve questions about the relation between trust and distrust. Specifically, if trust and distrust are distinct yet related concepts, they cannot be necessarily pervasive, because then it would not be possible to both trust and distrust the same actor at the same time in the same relationship. However, the reflection of what pervasiveness entails and how it is related to trust and distrust respectively is only beginning (Sitkin & Bijlsma-Frankema, 2018). To most people, pervasiveness implies that something is present everywhere within something, thus crossing domains within a particular relationship between actors. Some authors claim that distrust is always pervasive, implying that pervasiveness is a necessary condition for distrust and that the same situation without pervasiveness would instead be categorised as low trust. In this view, pervasiveness distinguishes trust (which is domain-specific) from distrust (e.g. Bijlsma-Frankema et al., 2015; Guo et al., 2017; Sitkin & Bijlsma-Frankema, 2018; Sitkin & Roth, 1993; Sitkin & Stickel, 1996). However, Sitkin and Bijlsma-Frankema (2018, p. 56) also raise the question, ‘When pervasiveness is part of our theories, what exactly do we mean by pervasiveness in terms of what domains are being subsumed?’, suggesting that distrust may be domain-specific after all within a given distrustor-distrustee relationship. So, while not questioning the empirical reality that pervasive distrust exists in some relationships, it seems that pervasiveness is not necessary condition of distrust. Distrust can also be domain-specific (i.e. non-pervasive). An actor can have confident negative expectations about another actor’s competence in one domain and confident positive expectations in another domain of their relationship. Or they can have confident negative expectations about how the other will act where corruption is concerned and confident positive expectations where environmental protection is concerned.

Interestingly, there is an intuition in the literature that trust may be pervasive as well. Some authors argue that trust has a necessary quality of pervasiveness and that when that condition is not met, there is distrust (e.g. Krouwel & Abts, 2007; Rosanvallon, 2008). Trust in this perspective is all-encompassing; it involves a routine-based, unquestioned leap of faith (Krouwel & Abts, 2007). This is close to blind or naïve trust, based on routine only and not on reasons or reflexivity (Möllering, 2006). These authors label any monitoring activity or control put in place as distrust. However, for example, while most researchers would agree that citizens who exhibit some scepticism and vigilance (Van De Walle & Six, 2014) are desirable for a vibrant community, the question arises whether it is correct to call them distrusting. In sociology, such citizens would rather be seen as reflexive trustors (Kroeger, 2019; Luhmann, 2017; Möllering, 2006). This argument relates to the different perspectives on the trust-control relationship briefly mentioned above. So, while there is certainly the empirical presence of pervasive, so-called ‘blind’ trust, not all trust is necessary pervasive. It is domain-specific.

Thus, our critical review has given us both pervasive and domain-specific distrust and trust. For each of these types actors have confident expectations about the behaviour of the other actor. However, these types do not cover the whole range of situations. There is also a space wherein there is not enough information for confident expectations. Depending on the perspective of the relationship between trust and distrust, this can be called the space ‘in between’ trust and distrust. In Figure 2a, we visualise what this would look like in the perspective of a linear trust-distrust relation with a neutral space in between. In the perspective that trust and distrust are distinct yet related concepts, this space of lacking confident expectations refers to the position in Lewicki et al.’s (1998) matrix where trust and distrust are low (see Figure 2b). This raises the question of how we may conceptualise this space of lacking confident expectations. We use the concept of suspicion to further reflect on this.

Figure 2. a. Relation between trust and distrust – second perspective b. Relation between trust and distrust – third perspective.

Suspicion

To advance our understanding of distrust and its relation to trust, we also need to critically reflect on the concept of suspicion. On the one hand, our review showed that some authors (especially in the field of political science) argue that any form of monitoring implies suspicion and necessarily leads to distrust (Krouwel & Abts, 2007; Rosanvallon, 2008). On the other hand, psychologists have shown that actors under suspicion are unsure whether to trust or distrust; they do not yet have enough information for confident expectations either way (Kroeger, 2019; Luhmann, 2017). This makes them very sensitive to new cues about the other actor’s (un)trustworthiness and relatively open to learning more about the other actor (cf. Six, 2005). Suspicion suggests not having confident expectations (yet). Some (psychological) studies indicate that while people are under suspicion both trust and distrust decisions are possible, but they are not equally likely. This is because people generally tend to avoid risks, and negative attitudes have more weight than positive attitudes (Kahnemann & Tversky, 1979). When evaluating others, people also tend to overestimate the likelihood of hidden motives (Hilton et al., 1993). Both of these effects will tend to lean towards distrust rather than trust. When suspicion turns into paranoid cognitions, then actors will choose to distrust without any further considerations (Kramer, 1994, 1998). This paranoid cognition is part of the distrustor’s disposition to distrust and impacts their distrust decision (Mayer et al., 1995).

The space of lacking confident expectations

Conceptualising suspicion as wondering whether to trust or distrust may be a promising avenue to deepen our understanding of the space of lacking confident expectations. In this perspective, suspicion resides in the space between confident negative expectations (distrust decision) and confident positive expectations (trust).

Recently, political scientists (Jennings et al., 2021) have proposed to call the space of lacking confident expectations mistrust. They conceptualise mistrust for situations where the belief about trust or distrust is not yet settled: ‘It involves a continuous process of feedback and updating, reconciling assessments of trust against trustworthiness, reflecting caution or scepticism concerning the expected actions of B in respect of X. The orientation of a mistruster is to be alert, informed and investigative. Mistrust will thus be manifested in a desire to assess the performance of B relative to expectations’ (Jennings et al., 2021, p. 1178). This aligns with other literature emphasising that being open to learning about the other actor in interactions is key to successful trust building. However, given suspicion’s tendency towards distrust rather than trust, we need further research to understand how to contain these negative tendencies so that trust can be built and maintained where warranted (cf. Six, 2005). Depending on how narrowly or broadly we define trust and distrust, respectively, the space of lacking confidence may be larger or smaller.

Universal sequences

This study explored whether it was possible to develop a distrust sequence similar to the one Dietz developed for trust. Having completed our exercise, we propose, paraphrasing Dietz (2011, p. 215), that there is always an assessment (varying in thoroughness) of the other party’s trustworthiness or untrustworthiness which informs a preparedness to be vulnerable, or not, that may lead to a trust-informed or distrust-informed action. The results of a trusting or distrusting encounter update the foundational assessment. In other words, there are universal sequences for both trust and distrust. This is depicted in Figure 1.

Universality of processes

Some elements (i.e. building blocks) of the two sequences – in particular inputs, actions and feedback loops – can take different forms. However, these differences do not change the sequence of the elements in the process. This means that both trust and distrust decisions are informed by inputs, based on beliefs, resulting in actions and are updated through an on-going feedback process that potentially reshapes the inputs. The inputs, in particular, may vary substantially in importance in different types of trust and distrust relationships. For example, actors that lack sufficient knowledge or experience to make informed judgments about (dis)trustworthiness may form attitudes stemming predominantly from predispositions and not actual interaction. The relative importance of different inputs also changes in various cultural settings, across time and levels of analysis. For instance, it may be easier to favourably resolve doubts about the other actor and take (small) leaps of faith in stable and reliable (vs. unstable and unreliable) institutional contexts.

In the perspective that trust and distrust are distinct yet related concepts, it should be empirically possible to study trust and distrust simultaneously. So far, empirical tests have shown that trust and distrust are indeed different (Benamati et al., 2008; Dimoka, 2010; Ou & Sia, 2010; Saunders et al., 2014). However, we still lack specific methodological tools to study trust and distrust unfolding within the same relationships (Lyon et al., 2015; Searle et al., 2018). Our sequential framework indicates some ways in which this could be possible, notably by simultaneously looking at trust-distrust beliefs (through their subdimensions), then trust-distrust decisions and trust-distrust actions. Some attempts have already been made in this area, exploring the components of beliefs (Engelke et al., 2019; Moody et al., 2014, 2017).

Dynamics of sequence over time

The universal sequence framework shows that the feedback loop updates the assessment of (un)trustworthiness through the revision of the inputs. The intensity and content of feedback loops varies because of differential self-reinforcing processes. While both trust and distrust processes are based on feedback loops reinforcing the initial behaviour, the distrust loop is far more vicious and emotional than the trust loop, thus reducing opportunities for learning about the possible trustworthiness of the other actor, as argued above.

Recent studies show that the assumption that trust is self-amplifying and that its resilience grows over time is likely to be too simplistic. Changes in the inputs sometimes do not lead to an update of the beliefs or decisions, or there is a delay in the updating. This then leads to a ‘trust trap’, where actors can be ‘stuck’ at a certain level of trust that does not cor­respond with recent inputs (Möllering & Sydow, 2019; Patent, 2022; Skinner et al., 2014; Swärd, 2016). There is no research yet that explicitly studies the existence of similar distrust traps, although our discussion of pervasive distrust and paranoid cognitions suggests the existence of distrust traps, which may be more difficult to escape than trust traps. The reason for this may be that trust can be falsified in action, while the same is hardly possible in the case of distrust (Gambetta, 1988b; Luhmann, 2017; Messick & Kramer, 2001; Nooteboom, 2002; Six, 2005). Since people generally refrain from forming relations with those they distrust, there is little opportunity for the other to prove their trustworthiness or experience positive actions that would allow for overcoming the initial distrust and building trust. Thus, actors in a trust trap have more chances to amend their judgements of the other’s trustworthiness compared to those in a distrust trap, who will err on the side of too much distrust.

The dynamics of the sequence are relevant when the nature of relationships evolves and trust shifts occur over time. At the societal level, Zucker (1986) found that disruption and change may damage trust, but then the nature of the relationship adapts to produce trust again. She distinguished three key bases for trust: process-based trust in relationships that are bounded by exchange logic; characteristic-based trust for relationships bounded by interpersonal connections; and institutional-based trust for relationships resting on formal structures. With the modernisation of society, process-based trust and characteristic-based trust were supplemented, and sometimes even replaced, by institution-based trust. This development impacted the inputs of the universal sequence and thus influenced the rest of the processes.

At the interpersonal level, Lewicki and Bunker (1996) postulated three forms of trust with different capacities for sustaining interpersonal, social relations that tend to follow each other over time, although the third level is often not reached: calculus-based trust (economic exchange), knowledge-based trust (a relationship exists with sufficient information about persons to understand their behaviour), and identification-based trust. Further research could examine the similarities and differences between these phases. For example, Zucker’s process-based trust and Lewicki and Bunker’s knowledge-based trust are both based on knowledge gained by repeated interactions.

No equivalent models have been proposed for distrust development phases to date. There is as yet also very little empirical research into distrust processes (for an exception, see Bijlsma-Frankema et al., 2015). Also, how does a relationship move from suspicion and questions about the other actor’s trustworthiness/untrustworthiness to pervasive distrust? Moreover, and perhaps more interestingly, how can a relationship move from pervasive distrust to mere suspicion and even trust? What are the different bases of such developments? This question leads us to a different kind of dynamic – the dynamic between the two sequences.

Moving between trust and distrust sequences

When does a relationship between actors that is in one sequence move to the other sequence? How does this happen? This line of research will only be able to make meaningful progress if the conceptual issues highlighted above about distrust, trust and their relationship are resolved. However, there is consensus that, although trust becomes more robust and resilient over time, once broken it is often difficult to repair, as reflected in the idiom ‘once bitten, twice shy.’ Trust violation and repair studies have listed an impressive catalogue of types of breaches that may disturb the trust process (for overviews, see Gillespie & Siebert, 2018). However, most studies on trust repair do not explicitly specify whether they deal with a situation in which (a) trust decreased after a violation but still exists in the relationship or (b) trust switched to (active) distrust. In the former position, the efforts for trust repair focus on increasing the trustworthiness of the violator to rebuild trust (Gustafsson et al., 2021). In the latter position, however, increasing trustworthiness may not be a sufficient or appropriate remedy. Here, distrust may first need to be reduced sufficiently so that trust can be (re)built (Lindenberg, 2000).

The above discussion assumes that there is a trust sequence and a distrust sequence and a relationship between actors is in one of the two sequences. However, when taking the perspective that trust and distrust are two distinct yet related concepts possibly coexisting in the same relationship, this is too simple an assumption. Further research is needed to ascertain what this implies for the dynamics of the processes both in time and moving from trust to distrust and vice versa.

Conclusions and next steps

Our conceptual exercise leads us to propose that there are universal sequences for both trust and distrust. This is this study’s first contribution to distrust (and trust) theory. We believe we have also made two further contributions. As we reviewed the distrust literature, we identified several related assumptions regarding the necessity of deceit and pervasiveness in distrust and untrustworthiness. Based on these findings we propose, as our second contribution, that both pervasive and domain-specific forms exist for both trust and distrust. All of these forms are based on confident expectations. Further, there is also an as-yet underexplored space of lacking confident expectations. This conclusion fits in the perspective that a neutral space exists between trust and distrust, as well as in the perspective that trust and distrust are distinct yet related concepts. Our third contribution lies in a critical reflection on the dynamics of the sequences, showing clear asymmetries between trust and distrust and addressing how the trust and distrust processes may be connected to each other.

These contributions raise issues for future research. To start with, the issue of pervasiveness versus domain-specificity as a characteristic of distrust is far from being resolved. Further work is also needed to explore suspicion and how it may relate to distrust and trust processes. Next, we need to better understand the space of lacking confident expectations; depending on how narrowly or broadly we define trust and distrust, respectively, the space of lacking confidence may be larger or smaller. Whether this space should be called mistrust or not is as yet unclear. Also, what does this space of lacking confident expectations imply for related issues such as trust building, trust maintenance, trust repair or distrust reduction? Do we need to revisit all this research and reassign it to the new conceptual categories? And where do the institutional factors that build trust, respectively reduce distrust, fit in the new conceptualisation? While awaiting consensus on how to conceptualise and define trust, distrust, mistrust and their relationship, we call on researchers to be very explicit regarding what they assume in their research and on reviewers to be alert to implicit assumptions that authors make in submitted manuscripts and encourage them to make them explicit. Assumptions are challenging because they are tacit and thus difficult to identify for those making them (Schein, 1992). The last area that particularly needs research concerns the distrust sequence, specifically more insight is needed on the conditions that create (dis)trust traps and how can they be avoided or removed. Correspondingly, the discussion of how to break the distrust cycle still awaits systematic exploration.

To sum up, our framework illustrates that distrust is not a mirror-image of trust but entails a similar variform process (cf., Dietz, 2011). We hope this extension of Graham Dietz’s framework for the trust process to include distrust will facilitate more nuanced and focused research and perhaps also trigger a conversation among trust researchers in the Journal of Trust Research.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This research received support from the National Science Centre, Poland [Project Number 2020/37/B/HS4/02940].

Notes

1 Please note that these categorizations are not meant to be black-and-white distinctions but only to help think through the sequence more carefully.