Trickle-Down,Trickle-Out,Trickle-Up,Trickle-In,and Trickle-Around Effects: An Integrative Perspective on Indirect Social Influence Phenomena

Leadership

Leadership has received the most attention with respect to trickle-down effects. Eleven papers address the trickle-down effects of one form of leadership or another. For example, research demonstrates ethical leadership trickled down from upper managers to supervisors’ ethical leadership and, in turn, to the ethical leadership of lower-level followers (Schaubroeck et al., 2012) and group-level subordinates’ organizational citizenship behavior (OCB) and deviance behavior (Mayer et al., 2009). Similarly, transformational leadership trickled down from managers to employees through supervisors (Bass, Waldman, Avolio, & Bebb, 1987) and influenced employees’ job performance (Dvir, Eden, Avolio, & Shamir, 2002; Yang, Zhang, & Tsui, 2010). Empowering leadership trickled down from second-level leaders to first-level leaders, which then influenced subordinates’ task performance, OCB, and social loafing (Byun, 2016) and voice behavior (J. S. Park, 2017). Finally, authentic leadership trickled down to influence employees’ helping behavior through supervisors’ authentic leadership (Hirst, Walumbwa, Aryee, Butarbutar, & Chen, 2016).

Researchers have also examined trickle-down effects of “dark” leadership behavior. For example, managers’ abusive behavior trickled down through supervisors’ abusive behavior to affect employees’ deviance behavior (Mawritz et al., 2012) and team members’ creativity (Liu et al., 2012). Managers’ authoritarian leadership trickled down to influence employees’ voice behavior through supervisors’ authoritarian leadership (Li & Sun, 2015).

Justice

Seven articles examine trickle-down effects of organizational justice perceptions. Supervisors’ perceptions of procedural justice from their managers trickled down to affect their subordinates’ OCB (Tepper & Taylor, 2003) and subordinates’ perceptions of their supervisors’ abusive supervision (Tepper, Duffy, Henle, & Lambert, 2006). Supervisors’ interactional and informational justice perceptions of their managers trickled down to influence subordinates’ interactional and informational justice perceptions of their supervisors (Wo et al., 2015). Interactional justice from managers trickled down to influence subordinates’ organizational OCB, individual OCB, and affective organizational commitment through subordinates’ perceptions of their supervisors’ interactional justice and abusive supervision (Aryee et al., 2007). Similarly, managers’ interactional injustice treatment trickled down to influence subordinates’ psychological distress and insomnia through supervisors’ abusive behavior (Rafferty, Restubog, & Jimmieson, 2010) and influenced subordinates’ negative affect toward others through supervisors’ negative affect and supervisors’ abusive behavior (Hoobler & Hu, 2013). Fair interactional treatment supervisors received from their managers also trickled down to influence subordinates’ perceptions of interpersonal treatment from supervisors at an aggregated level, which affected group-level employee citizenship and deviance behavior (Ambrose et al., 2013).

Other constructs

The literature has revealed trickle-down effects involve a variety of other constructs as well. For example, managers’ service performance trickled down through supervisors’ service performance to influence employees’ service performance (Liao & Chuang, 2004). Research also demonstrated supervisors’ perceptions of psychological contract breach with their managers trickled down to affect subordinates’ perceptions of abusive supervision of their supervisors (Hoobler & Brass, 2006) and subordinates’ citizenship behavior toward customers.

Shanock and Eisenberger (2006) demonstrated supervisors’ POS trickled down to influence subordinates’ POS through subordinates’ perceived supervisor support (PSS) and was positively related to their in-role performance and extrarole performance. Relatedly, Wu, Lee, Hu, and Yang (2014) found managers’ nonwork support from the organization and their bosses trickled down to influence supervisors’ nonwork support toward their subordinates’ and employees’ individual OCB and organizational OCB.

Simons, Friedman, Liu, and Parks (2007) found supervisors’ perceptions of their managers’ behavioral integrity trickled down to influence subordinates’ trust in supervisors, their interpersonal justice perceptions, and their job satisfaction, commitment, and intention to stay through subordinates’ behavioral integrity perceptions of supervisors. Gratton and Erickson (2007) found senior executives’ cooperation trickled down through supervisors’ cooperation to affect employees’ cooperative behavior.

Detert and Treviño (2010) found how senior managers treated voice behavior trickled down via how the supervisors interacted with their employees to affect employees’ speaking-up behavior. Biron, Farndale, and Paauwe (2011) showed senior executives’ promotion of certain performance management practices trickled down to influence employees’ performance management effectiveness through their impact on the performance management effectiveness of middle-level managers.

Trickle-down effects also have been demonstrated to occur from the highest to lowest levels of the organization. Tucker, Ogunfowora, and Ehr (2016), for example, showed that the safety priorities of CEOs trickle down to influence frontline employee injuries, indirectly, through their influence on top management team and supervisory perceptions of safety climate and supervisory support for safety.

Finally, Kolk et al. (2016) performed a comparative case study analysis of trickle effects involved in the formation of corporate social responsibility (CSR) partnerships. They found evidence that higher-level management’s engagement in and support for CSR partnerships trickled down to influence lower-level employee participation in these partnerships via their impact on supervisors’ ability and willingness to facilitate and support that participation.

In all, research provides strong evidence for the existence of trickle-down effects within organizations. These effects emerge across a variety of actors, constructs, and organizational levels.

Trickle-out effects

Although not as extensive as trickle-down effects, trickle-out effects have also been documented. Trickle-out effects refer to how the perceptions, feelings, attitudes, and behaviors of a source in the organization (e.g., a supervisor) influence the perceptions, feelings, attitudes, and behaviors of a transmitter in the organization (e.g., an employee) and then pass through the organizational boundary to influence the perceptions, feelings, attitudes, and behaviors of an external recipient. This work focuses primarily on effects on customers or family members. Our review identified six papers studying trickle-out effects.

Masterson (2001) is broadly recognized as the first to utilize the term “trickle-down effects” in an organizational research setting, although her paper actually describes a trickle-out process. Masterson found that employees’ (instructors’) perceptions of distributive justice and procedural justice from their supervisors trickled out to influence their customers’ (students’) perceptions of distributive justice and procedural justice from their instructor. Also examining the trickle-out effect on customers, Lichtenstein et al. (2010) found store managers’ organizational identity trickled out through their subordinates’ organizational identity to affect customers’ identification with the organization, ultimately leading to higher financial performance of the organization. Bordia et al. (2010) showed subordinates’ perceptions of their supervisors’ psychological breach trickled out to affect customers’ satisfaction.

Examining trickle effects on families, Hoobler and Brass (2006) found that supervisory perceptions of psychological contract violations trickled out to affect subordinates’ family member perceptions of being undermined via their impact on subordinate perceptions of abusive supervision. Restubog, Scott, and Zagenczyk (2011) found that abusive supervision trickled out to affect family member perceptions of being undermined via its impact on subordinate psychological distress. Hoobler and Hu (2013) showed supervisors’ interactional justice experiences trickled out to subordinates’ family member perceptions of work-family conflict via their impact on subordinate perceptions of abusive supervision and negative affect.

Trickle-up, trickle-in, and trickle-around effects

Trickle effects emerge in other directions as well. Trickle-up effects refer to indirect social influence that is transmitted upward in the organizational hierarchy. We found one paper examining trickle-up effects. A comparative case study by Kolk et al. (2016) suggested employees’ commitment to CSR partnerships trickled up to influence their supervisors’ engagement in the partnership, which, in turn, affects managers’ active participation.

Trickle-in effects refer to the process by which an outside source’s (e.g., a customer or a family member) influence on a transmitter (e.g., an employee) passes through the organizational boundary to affect how the transmitter influences a recipient (e.g., a supervisor or a coworker) in the organization. Although we found no published studies explicitly examining trickle-in effects, a dissertation by Wo (2015) found employees’ perceptions of customer deviance trickled in to influence employee deviance directed at coworkers and toward the organization.

The trickle-around effect refers to an indirect effect transmitting horizontally (at the same organizational hierarchical level). Thus, the perceptions, feelings, attitudes, or behaviors of Employee A influence the perceptions, feelings, attitudes, or behaviors of Employee B, which, in turn, influence the perceptions, feelings, attitudes, or behaviors of Employee C. Our review identified one paper that examined the trickle-around effect. In a negotiation setting, Foulk et al. (2016) found Partner 1’s rudeness toward Partner 2 influences Partner 2’s rudeness toward Partner 3 in subsequent negotiations.

Methodology

Researchers have utilized a variety of methods for studying trickle effects. However, the cross-sectional survey approach is by far the dominant method used in studying trickle effects. Given our proposal that trickle effects can take place in different directions, establishing causality rather than mere correlation between independent and dependent variables becomes especially important. For example, the average correlation of the trickling construct is .32. However, when we observe that managers’ perceptions, feelings, attitudes, and behaviors correlate with subordinates’ perceptions, feelings, attitudes, and behaviors, it is important to investigate whether these indirect social influence effects represent trickle-down or trickle-up processes. Correlational studies do not allow for strong conclusions regarding this point.

Cross-sectional survey studies are not the optimal approach to study trickle effects, given their limited usefulness in determining causality (Antonakis, Bendahan, Jacquart, & Lalive, 2010). Longitudinal survey studies (utilized in nine of the trickle studies) alleviate the causality concern to some extent, as these studies temporally separate the measurement of independent and dependent variables and, hence, help in establishing causality (Antonakis et al., 2010). However, an experimental approach provides the most stringent test of causal relationships (Antonakis et al., 2010; Cook, Shadish, & Wong, 2008). Only one trickle study utilized an experimental approach.

One reason for the lack of experimental work may be the difficulties involved in executing a true experimental design in a field setting. However, the focus of the existing research suggests another reason. Researchers have been more interested in demonstrating trickle effects than explaining how and why they occur. Establishing the phenomenon exists is important early in the development of a research area. However, understanding the process by which trickle effects occur is important as the field moves forward. The multiple mediation model utilized by Wo and his colleagues (Wo, 2015; Wo et al., 2015) is one approach. However, carefully designed experiments that complement field surveys would also be useful. Such studies could create the opportunity for indirect social influence to flow through an experimenter-created hierarchy or across coworkers or from customers into the organization. Indeed, experimental work in the persistence of organizational cultures by scholars such as Zucker (1977) and Caldwell and Millen (2009) provides an example of how trickle effects might be examined. Additionally, experiments could be designed to tease apart the relative impact of different theoretical explanations, such as distinguishing social exchange motives from social learning. We also revisit this issue when we discuss research avenues that stem from the new framework we propose.

Another related issue in methodology of current research involves the role of time in understanding trickle effects. This is an important issue, in that the literature recognizes at least conceptually that trickle effects unfold over time as the influence flows from the source through the transmitter to the recipient (Ambrose et al., 2013). For example, in exploring trickle-down effects in performance management systems, Biron et al. (2011) convey how important time is in the trickling down of supervisory feedback. In the Biron et al. paper, an employee recalls how that unfolds over time:

[Today] I sit down with my boss, we talk about my performance for the year . . . and we set objectives for the forthcoming year. And at that point in time my boss will decide . . . where he would place me in terms of the spectrum profile. . . . Eventually, I’ll be able to see that in my [personal development plan]. (1304, emphases added).

Mayer et al. also note the role of time in the social learning processes they theorize underlie the trickle down of ethical leadership (i.e., “When subordinates learn over time [emphasis added] that positive behaviors are valued and rewarded. . .”; 2009: 3). Additionally, researchers note that some phenomena, like abusive supervision and ethical leadership, take time to emerge, which may influence how rapidly they trickle down (Hoobler & Brass, 2006; Schaubroeck et al., 2012).

We agree that how long it takes for a trickle effect to take place likely depends on the construct that is being transmitted and the process that underlies the transmission. For example, the forming of justice perceptions and leadership styles takes time, so the same will be true of their trickling down. In contrast, constructs such as rudeness may pass quickly through a trickle chain. Similarly, the transmission of affectively based processes that take little time to process may occur more quickly than transmission via cognitively based processes that require greater thought. To date, however, no research addresses these questions.

Although the role of time has played into several authors’ thinking about how trickle effects unfold and persist, few have formally integrated it into their research designs. Several authors have employed time-separated measures (ranging from 1 week to a few months), but we know of no trickle study designed to shed light on the specific questions of how long it takes trickle effects to play out or how long they persist. (In a post hoc analysis, Foulk et al., 2016, found that how much time had elapsed—up to 7 days—did not affect the trickle effects of rudeness.)

This suggests a significant research opportunity, in that several trickle scholars address—at least implicitly—the important role that time may play in the emergence of trickle effects. One distinction that may be useful in thinking about time and trickle processes is provided by George and Jones (2000). George and Jones distinguish between incremental and discontinuous change processes in organizations. Incremental processes are those in which the relationship between two variables builds up slowly. This describes processes such as social exchange, which typically unfold slowly over time. We might expect time lags of several weeks to be necessary for trickle effects to emerge. Discontinuous processes, on the other hand, are those that are triggered in a much quicker fashion, perhaps even in a single event. This describes processes such as displaced aggression, which often unfolds in a relatively short time window. We might therefore expect shorter time lags, perhaps the same day, to be sufficient for trickle effects to emerge, depending on the specific constructs being focused on. Thus, researchers should be thoughtful about their constructs of interest and the time span in which they are likely to unfold.

Moderators

It is not surprising that most of the moderators identified in the trickle-effect literature focus on characteristics of the transmitter (i.e., the person who is influenced by the source and then influences the recipient). Because the transmitter is the middle link in the chain effect of indirect social influence, his or her disposition to be influenced by the source plays an important role in affecting the magnitude of the trickle effect. However, as both the source and the recipient are also critical parties in the trickle-effect process, attention should be given to their personal characteristics and contextual variables that affect them as well. We discuss potential moderators in more detail as part of our discussion of a framework for examining trickle effects in the final section of this paper.

The AELM and the organization of the trickle-effects literature

The AELM provides a framework for organizing the trickle-effects literature. First, the AELM can provide structure by organizing the literature according to the principal theoretical explanations provided in the literature for the emergence of indirect social influence effects. More specifically, the three dominant theoretical perspectives that emerged from our review—social learning theory, social exchange theory, and displaced aggression—differ in the amount of elaboration that the processes they propose require.

Social exchange theory adopts a cognitive approach. It suggests that trickle effects are triggered when employees are motivated to follow the norm of reciprocity, which requires considering the relationship and the obligations owed and a complicated mix of interdependent observations and responses. This process requires thinking (i.e., elaboration) about the relationship, benefits received, and appropriate means of reciprocation. It therefore reflects central route processing. In contrast, displaced aggression suggests employees develop negative emotions such as frustration or anger after receiving negative treatment from others, which increase their need to vent these emotions toward third parties. Displaced aggression is cathartic and reactive. It requires little thinking and therefore reflects peripheral route processing.

Interestingly, whereas the third dominant theoretical perspective from our review—social learning—might appear to elicit elaboration as well (learning appears to require thought), Bandura (1977) suggests individuals may imitate the behavior of role models without extensive thought or deliberation. Social learning theory can be considered a bridge between behaviorist and cognitive learning theories because it acknowledges the role of attention, memory, and motivation but also relies on principles of (vicarious) reinforcement. It recognizes that cognition plays a role in learning (to a greater extent than classical or operant conditioning), but it does not require extensive elaboration. Social learning requires that individuals attend to the behavior of the model, remember the behavior, are able to reproduce it, and are motivated to reproduce it, but none of these components requires extensive elaboration. Therefore, social learning represents a middle ground between central route processing (extensive elaboration) and peripheral route processing (little or no elaboration).

Thus, the three main theoretical perspectives associated with trickle effects fall across the elaboration continuum. Social exchange theory suggests the greatest amount of elaboration, social learning a modest amount of elaboration, and displaced aggression the least elaboration.

The second way in which the AELM provides a framework for organizing the trickle-effects literature is that it allows us to categorize the constructs that have been examined in trickle-effects research. Research on the ELM indicates personal relevance is the most important factor affecting individuals’ motivation to process information in a more deeply elaborative way (Petty & Wegener, 1998). Personal relevance reflects individuals’ beliefs that the issue/event will have significant consequences for their own lives. According to Petty and Cacioppo (1986: 145), it can be judged based on a variety of factors, including number of personal consequences, magnitude of consequences, and duration of consequences. Although all workplace interactions are to some extent personally relevant, the level of relevance varies and may be useful in categorizing trickle effects. Indeed, categorizing constructs by their personal relevance highlights differences between constructs that on the surface appear to be similar, as well as similarities between constructs that on the surface appear to be different. Table 1 categorizes trickle constructs by personal relevance into high, moderate, and low categories. (Although these categorizations are subjective, we strove to differentiate the constructs on how personally relevant the perception, attitude, feeling, or behavior was likely to be to the receiver.)

OPEN IN VIEWER

Table 1

Categorization of Constructs by Personal Relevance

Low Personal Relevance	Moderate Personal Relevance	High Personal Relevance
Performance management effectiveness Authentic leadership Support for corporate social responsibility partnerships Service training	Organizational identity Transformational leadership Rudeness Cooperation Authoritarian leadership Distributive justice Procedural justice Informational justice Ethical leadership Aggressive norms Ethical culture Perceived organizational support Behavioral integrity CEO safety priorities Perceived supervisor nonwork support Organizational deviance behavior	Interpersonal deviance behavior Interactional justice Interpersonal justice Voice Psychological contract breach Abusive supervision Abusive behavior Empowering leadership

Some interesting observations arise from categorizing trickle-effect constructs based on personal relevance. For example, the categorization suggests not all types of justice or all types of leadership should necessarily trickle the same way. Rather, different facets of justice and different types of leadership may differ in their personal relevance. Therefore, constructs from the same domains may trickle in very different ways. In the leadership area, for example, empowering leadership is targeted directly at the follower, placing it in our categorization of high personal relevance as it is designed to shape directly the followers’ environment, expectations, and behavior. In contrast, authentic and ethical leadership are focused primarily on the characteristics and behavior of the leader. Although it is expected these leadership types will exert some impact on the work unit and beyond, they are categorized as less personally relevant than empowering leadership. Similarly, different facets of justice display different degrees of personal relevance. Procedural and distributive justice are less personal, and more organizational, than interpersonal justice. Given these differences in personal relevance, the AELM suggests individuals would be motivated to engage in greater elaboration for empowering leadership and interactional justice than for authentic leadership and procedural justice. This highlights the observation that constructs that appear to be similar, based on their core construct foundations (e.g., empowering and authentic leadership; distributive and interpersonal justice), may actually behave very differently in trickle situations. Likewise, constructs that appear to be different, due to their core construct foundations (e.g., empowering leadership and interactional justice; authentic leadership and procedural justice), may behave similarly in trickle situations.

Recognizing the importance of personal relevance can change how scholars interpret existing trickle-effects research. However, it can also change how they approach examining and theorizing about the trickle effects of constructs for which trickle effects have yet to be examined. Consider incivility, which has not yet been examined in the trickle-effects literature. One might expect the trickle-down effect of incivility to mirror those of a related construct, abusive supervision. Both are forms of interpersonal mistreatment and are often considered to vary primarily on the magnitude of the mistreatment (Reich & Hershcovis, 2015). However, the AELM suggests the two constructs differ in a way that may influence how they trickle down. Abusive supervision is subordinates’ perceptions of the extent to which their supervisors engage in the sustained display of hostile verbal and nonverbal behaviors (Tepper, 2000). That is, abusive supervision is sustained and it is targeted directly at the recipient. It is perpetrated by an individual (the supervisor) who is central in the recipient’s work life. In contrast, incivility is low-intensity deviant behavior with ambiguous intent to harm the target; it reflects a general lack of regard for others (M. Andersson & Pearson, 1999). Viewed through an AELM lens, therefore, abusive supervision, with its centrality of the perpetrator to the recipient, its ongoing nature, and its perceived targeted hostility, makes it more personally relevant than incivility. Therefore, rather than assuming the two constructs would trickle down via the same process because of their similar affective tone, the AELM suggests abusive supervision would elicit greater elaboration (central route processing) and incivility would elicit less elaboration (more peripheral route processing); thus, each may emerge via different trickle mechanisms.

The AELM also allows us to see commonalities across constructs that might not otherwise be linked. For example, the literature does not connect trickle-in effects of customer interpersonal deviance (a negative behavior) with trickle-down effects of empowering leadership (a positive behavior). But the AELM perspective would suggest both are high on personal relevance and would therefore best be categorized together.

New research directions

The AELM also suggests new avenues of research not currently explored in the trickle-effects literature. For example, on the basis of the AELM, the processing route (central or peripheral) utilized by the transmitter and the recipient should have implications for the strength and stability of the trickle-effect relationship. When utilizing the central route, individuals not only consider the information provided by the source or event but also generate and reflect on other potentially relevant information as well. They utilize information from memory. They seek information from other sources. They consider how new information fits with other experiences. They integrate these multiple pieces of information to form a perception, derive a judgment, form an attitude, or make a behavioral choice. That is, when individuals (in this case, transmitters and recipients) engage in greater elaboration, they consider the information from the source, but they supplement this with information from other experiences and other sources (Petty & Cacciopo, 1986), thereby diluting the impact of information received from the source. Thus, the relationship between the source’s original perception, attitude, feelings, or behavior and the recipient’s perception, attitude, feelings, or behavior should be weaker than if peripheral processing is used because with central processing, the source’s information is only part of the information considered. In peripheral processing, on the other hand, little other information is considered. Therefore, the influence that stems directly from the source should experience less distortion or dilution as it passes through the transmitter to the recipient. Thus, the relationship between the source’s perception, attitude, feelings, or behavior and the recipient’s perception, attitude, feelings, or behavior should be stronger when peripheral processing is employed.

As an illustrative example, let us consider how abusive supervision might trickle down through the central route versus the peripheral route. When supervisors work for an abusive manager and engage in central route processing, they may also consider their experiences with other managers in previous jobs, they may examine the fit of the abusive style with the organizational culture, and they may reflect on their own beliefs about the appropriateness and effectiveness of abusive behavior for individuals in a leadership position. All of this information then informs their decision as to whether to engage in abusive behavior toward their subordinates. To the extent that the additional sources of information do not support abusive behavior, these transmitters are less likely to engage in abusive supervision behavior themselves, and abusive supervision is less likely to trickle down. In contrast, if abusive supervision flows through the peripheral route, little additional information beyond the manager’s abuse is considered by the supervisor. Rather, they may enact a “monkey see, monkey do” response to the behavior—simply mimicking their manager—or they might mindlessly vent their frustration on their subordinates. In either case, there would be a stronger relationship between the manager’s behavior and the trickle-down behavior of the supervisor.

To examine whether the processing route employed (central vs. peripheral) influences the strength of trickle effects as suggested by the AELM, we looked to the categorizations of personal relevance of the trickle constructs provided in Table 1. As noted above, personal relevance elicits elaboration (Petty & Brinol, 2012), so we therefore expected that higher elaboration constructs would display weaker trickle effects than lower elaboration constructs.

To investigate this question, we explored the strength of the overall trickle effect by examining the correlation between the source’s perceptions, feelings, attitudes, or behaviors and the recipient’s perceptions, feelings, attitudes, or behaviors. As predicted by the AELM, the average correlation is strongest for the low personal relevance constructs (peripheral processing; r = .38) and weakest for the high personal relevance constructs (central processing; r = .19). The moderate personal relevance constructs’ average correlation falls between the two (r = .27).

Beyond its impact on the strength of the trickle effect, the amount of elaboration should also have an impact on the stability of the trickle effect. Because of the greater thought given to information when the central route is utilized, the perceptions, attitudes, judgments, and behaviors that result from this elaboration are more stable and enduring than attitudes, judgments, and behaviors that flow from peripheral processing (Petty & Brinol, 2012).

Taken together, the strength and stability propositions present an interesting picture. Although elaboration may weaken the trickle effect, when the effect does trickle, it is more stable. In contrast, constructs may flow more easily through the peripheral route, but there may be less long-term impact of the trickle effect, as peripheral trickle effects are more easily changed. These propositions highlight the importance of additional longitudinal work in this area. The current cross-sectional approaches employed in most studies do not provide a strong foundation on which to evaluate the permanence of trickle effects. Additionally, these propositions highlight the complexity of this work, as even constructs that trickle via the peripheral route will endure until other information or events supplant them. Perceptions, feelings, attitudes, and behaviors that emerge from peripheral processing will be more easily changed, but they still require a “nudge” to make that happen.

The AELM also suggests a new set of moderators that may influence trickle effects. More than half of the studies in our review were silent with respect to moderators of the trickle-effect process. The AELM suggests that variables that affect individuals’ motivation and ability to process information should influence the flow of trickle effects. Factors such as need for cognition, the quality of information, accountability, and consistency with prior beliefs influence individuals’ motivation to engage in elaboration. Factors such as repetition, distractions, and cognitive load influence individuals’ ability to engage in elaboration. Considering these factors highlights moderators that should be examined in trickle research.

For example, individuals (transmitters and recipients) higher in need for cognition are more likely to engage in elaboration and, thus, utilize central route processing. Factors that influence cognitive load and distraction, such as work overload and time pressure, should encourage more peripheral processing. In addition, the originating perceptions, feelings, attitudes, or behaviors of the source may play a role. When these are consistent with previous perceptions, feelings, attitudes, and behaviors of the target (the transmitter and then the recipient), they are more likely processed by a peripheral route. Yet when the source’s perceptions, feelings, attitudes, and behaviors are unusual, they are more likely to stimulate the transmitter’s interest in understanding them, leading to processing by the central route.

The quality of information also matters; higher quality information leads to greater inclination to elaborate. Consequently, if the source is widely viewed as highly competent, credible, or expert in an area, there is more likely to be greater elaboration on the part of the transmitter because the information coming from the source would be viewed as high quality and worthy of attention and reflection (Petty & Brinol, 2012).

The factors that influence motivation and ability can be extrapolated to support other moderators as well. For example, accountability increases the motivation for elaboration. It may be that if the source (i.e., manager) is an empowering leader, the transmitter (i.e., the supervisor) experiences greater accountability and, therefore, is more likely to engage in elaboration. Similarly, PSS might influence elaboration likelihood. If individuals (transmitter or recipient) believe their supervisors care about their welfare on the job, they may also be experiencing fewer distractions and be less likely to be operating under a burdensome cognitive load, both of which would facilitate elaborative processing.

In all, the AELM provides a road map for thinking about moderators that directly influence motivation and ability to engage in elaborative thinking, as well as those that indirectly influence motivation and ability. As we note above, in general, the trickle relationship should be weaker when there is greater elaboration. This suggests factors that encourage greater elaboration might be useful in buffering the flow of negative trickle effects but might also weaken the flow of positive trickle effects.

The AELM also allows us to adopt a more nuanced approach to the role of affect in trickle-down processes. Whereas affect generally elicits peripheral processing, research on the ELM demonstrates that affect-laden constructs are not necessarily processed via the peripheral route. Rather, factors that influence the motivation and ability of individuals to process information influence how affect is processed. For example, when an affect-laden construct is high in personal relevance (e.g., abusive supervision), the ELM (and AELM) suggests a more elaborative process will be employed. In this case, the affect is part of the information the individual considers. Consistent with the affect infusion model (Forgas, 1995) and the emotions as social information model (Van Kleef, De Dreu, & Manstead, 2010), when individuals engage in central route processing, affect is information. Thus, while many affect-laden constructs are likely to elicit peripheral processing, this more nuanced approach allows scholars to consider when—and why—affect-laden constructs will be the subject of elaboration.

The AELM sparks some thought about methodological issues as well. One of these concerns the role of time in studying and understanding trickle effects. Beyond the well-established observation that judgments resulting from high elaboration processing are more durable than those resulting from low elaboration peripheral processing in that they persist longer, are more resistant to change, and exert greater influence on other attitudes and behaviors (Krosnick & Petty, 1995), the literature is largely silent with respect to the role of time.

However, ELM research has revealed several fronts on which researchers do need to be cognizant about issues related to timing when executing studies based on ELM thinking. For example, research has shown that time pressures may inhibit the ability of individuals to engage in central processing (Kruglanski & Freund, 1983). Therefore, studies must be designed carefully so as to avoid creating situations in which participants might be motivated to engage in elaboration but are prevented from doing so as a result of protocol-induced procedures. Similarly, the measures themselves that are employed in a study can place unintended restrictions on the type of processing employed by participants. Implicit measures of attitudes, for example, tap evaluations that come to mind quickly, while more deliberative measures allow participants to reflect more fully on the target being assessed (Petty, Fazio, & Brinol, 2009). Thus, measures must be selected carefully to allow researchers to detect the processes they believe to be at work. Likewise, research has highlighted the importance of designing studies capable of addressing the potential gap between the salience of attitudes or judgments at the point in time at which they are measured versus the salience of those same attitudes or judgments at the point in time at which a behavioral decision is measured (Petty & Wegener, 1998). With most dependent variables in trickle studies reflecting behavioral outcomes, reducing that timing gap is critical for establishing the true relationship between trickle-based judgments and outcomes. In all, the role of time represents the especially challenging issue for researchers employing an AELM perspective on trickle effects.

Other methodological issues are worth noting as well. As we noted above, previous research has not focused on the process by which trickle effects occur. However, once the process of transmission becomes of interest, it almost demands examination via an experimental approach paired with research in organizational settings. The laboratory provides an opportunity to test the theory by manipulating moderating variables (e.g., time pressure, personal relevance) and examining the strength and stability of trickle effects. For example, one might compare the trickle effect of abusive behavior aimed specifically at the target (high personal relevance) versus abusive behavior targeted at a group (lower personal relevance). One might examine the trickle effect of interpersonal injustice under conditions of low cognitive load (central route processing) versus high cognitive load (peripheral route processing). The AELM provides a framework for exploring the trickle-effects process and studying these processes in experimental settings.

Finally, organizing extant research using the AELM holds promise for pointing researchers in more productive directions in understanding the theoretical foundations of trickle effects. Using AELM, scholars will be able to craft a theory-driven perspective based on whether the focal construct is of lower personal relevance, moderate personal relevance, or higher personal relevance. The AELM provides a new perspective that helps scholars integrate trickle research from different research domains, thereby promoting an interdisciplinary approach to understanding them. The AELM also can guide researchers in identifying relevant boundary conditions for trickle effects from both transmitters’ and recipients’ perspective. We believe such advancements are useful for trickle research, and such advancements will help the trickle-effects literature to grow, develop, and mature.