Impact of Communicated Company Accounts During Various Crises on Consumer Emotions, Attitudes and Behavioural Intentions

Abstract

Despite the fact that major crises cost companies many millions of dollars in lost sales, damaged reputation, and reduced market share annually, there is little research on consumers' reactions to company crises. In fact, no research has investigated - from a consumer's perspective - the range of thoughts, feelings and actions consumers have in response to a product or service crisis. This creates a problem for firms, both in their understanding of how and why consumers react to a crisis, and in their management of the crisis to reduce negative reactions from current and prospective consumers, and therefore minimise negative outcomes for the company. Following a crisis event, the only tool under organisational control that may influence the crisis outcome by reducing negative reactions from current or prospective consumers of the product or service is the company's communicated 'account' of its role in the crisis. Accounts commonly used by companies during a crisis are 'no comment', 'denial', 'justification', 'excuse' and 'confession'. However, there are currently few guidelines and empirical investigations indicating the optimum company account to reduce negative consumer reactions to crises. In fact, no crisis researcher has examined the impact of all these company accounts on consumer behaviour. One crisis researcher (Jorgensen, 1996) using Weiner's attributional (1986) theory (WAT) found that company accounts (denial, confession) and two different crisis causes impacted consumer anger and behaviour (purchase intention, investment intention). However, although anger was an immediate precursor of negative consumer behaviour, the focus in attribution theory is on attributions, not emotions as drivers of behaviour and attitude. Additionally, WAT limited emotions to anger and sympathy. Therefore, no research has examined either the array of emotions and behaviours consumers of a product or service exhibit during a crisis, or whether different emotions drive particular consumer behaviours and attitude to the company. As well, while Jorgensen (1994, 1996) examined the impact of different crisis causes, or crisis types (internal/controllable and external/uncontrollable), on consumer reactions, and Jorgensen (1994) also investigated a 'mixed' or ambiguous crisis, how other crisis types impact consumers has not been examined. Also not investigated is whether each of these crises, that is Locus crisis (internal/external) or Controllability crisis (controllable/uncontrollable) have a differential impact on consumers. In addition, while WAT refers to personal relevance, it was not included in Weiner's (1986) model. However, one theory that includes personal relevance, attributions and emotions to explain behavioural and attitudinal reactions to an event is Weiss and Cropanzano's (1996) Affective Events Theory (AET). Using AET, this thesis investigated the impact of five commonly used company accounts in different crisis types on an array of consumers' emotions and behavioural intentions and on attitude, as well as on attributions (including foreseeability and intentionality), involvement (i.e., personal relevance), responsibility, and accountability. Further, this thesis examined which emotions acted as drivers of different behavioural intentions. The research for this thesis comprised three studies using mixed methods in a triangulated design. The first study used focus groups to investigate consumers' recalled thoughts, feelings and behaviours, and any other factors consumers considered important in various salient company crises. The results showed that consumers recalled a full range of both positive and negative emotions - anger, fear, sadness, joy, surprise and love. Reported behaviours included boycott, switching, avoidance and loyalty. As well as variables that were identified as being congruent with those described in the literature, emergent variables included attitudes to government, media and companies. The second study pilot tested the operationalisation of the independent variables, Accounts (no comment, denial, excuse, justification, confession) and Crisis types of Locus and Controllability (internal/external, controllable/ambiguous/uncontrollable). Crisis type was operationalised using an airplane crash scenario in order to remain as neutral as possible across various demographic variables. This between-subject 5 x 2 x 3 factorial design used a student sample (n = 316) randomly allocated to 30 treatments. Results were analysed using MANOVA and showed a main effect for Crisis Type, but no main effect for Accounts and no interaction effect. Follow-up mini pilots tested possible causes of account failure e.g., faulty operationalisation, use of a crash photograph. An extra independent variable, Harm level (high with 33 dead/low with 33 slightly injured), was added and the ambiguous crisis type was removed. A number of dependent variables were tested: emotions, behaviour, attitude as well as involvement, attributions, foreseeability, intentionality, responsibility and accountability. Demographic details were also collected. Scales were tested and adjusted or pared down, and some variables were removed. The third study used a between-subject factorial 5 x 2 x 2 x 2 design testing the impact of Accounts (no comment, denial, excuse, justification, confession), Crisis types of Locus (internal/external) and Controllability (controllable/uncontrollable), and Harm level (high/low) on consumers' emotions, behaviour and attitude, as well as involvement, attributions, responsibility and accountability. Demographic details were also collected. A random electoral roll sample (n = 907) was randomly assigned to 40 treatments. Results were analysed using MANOVA and other tests (e.g., ANOVA, multiple regression, confirmatory factor analysis). Results showed a main effect for Accounts and Crisis Types but not for Harm levels, and no interaction between independent variables. For Crisis type, internal crises resulted in significantly higher anger, fear, surprise, disloyalty, responsibility and accountability. External crises resulted in significantly higher joy, loyalty and attitude. Controllable crises resulted in significantly higher anger, fear, complaining, disloyalty, accountability and responsibility. Uncontrollable crises resulted in significantly higher joy, loyalty and attitude. The account of 'confession' showed best consumer outcome, followed by 'no comment', while 'excuse', 'justification' and 'denial' received similarly poor results. 'Confession' and 'no comment' resulted in lower anger and disloyalty and higher sympathy, loyalty and attitude.? Additionally, emotions of fear, anger, and low attitude predicted disloyalty. Anger, attributions and fear predicted complaining. Joy, low fear and low anger levels predicted loyalty. This thesis made a number of theoretical contributions. To the researcher's knowledge, this was the first investigation of multiple emotions and behaviours in different crisis types, finding that crisis types impacted a range of consumer emotions and behaviour, as well as attitude. It was also the first study to find that different emotions in crises predicted different behaviours and attitude. It was the first application of constructs of involvement and accountability to company crises. It was also the first crisis study to empirically investigate the impact of demographic variables in a crisis, finding that culture and age impact emotions and behaviour.