Trickle-down effect of organizational trust on coworker trust: The moderating role of cultural dissimilarity and relationship length

The COVID-19 pandemic has accelerated trends of globalization and digitalization, making geographically dispersed teams a common practice in firms. Despite benefits derived from the members ’ diversity, such teams are also prone to trust deficiency. Advancing prior research, this study focuses on links between multiple referents of trust. We draw on halo and priming effects to suggest that employees ’ trust toward their organization could trickle-down to trust in their co-workers. Moreover, we highlight the moderating role of cultural dissimilarity and relationship length. Analyzing 317 relationships between Turkish employees and their co-workers of Turkish and German cultural background, we present evidence for a trickle-down effect of organizational trust on trust in co-workers. We also find that the trickle-down effect of trust is stronger when cultural dissimilarity is high than when it is low, suggesting that trust in the organizations may outweigh cultural barriers that could hamper trust between co-workers.


INTRODUCTION
Trust between co-workers was found to be key contributor to a wide range of positive organizational outcomes such as work engagement, job satisfaction, and organizational citizenship behavior (e.g., Fulmer & Gelfand, 2012;Lyu & Ferrin, 2018).When employees trust their co-workers and their organization, they are likely to commit to completing their tasks and to contribute directly and indirectly to the success of the company (Cunningham & MacGregor, 2000;Moorman, Darnold, & Priesemuth, 2013).Recent evidence, however, suggests that trust in organizations is at a new record low (Edelmann Trust Barometer, 2021;Fulmer & Ostroff, 2017).Two trends may explain this unfortunate record, with both trends ramping up since the outbreak of COVID-19.First, upward shifts in the volume of remote work arrangements necessitate new approaches to understanding and facilitating trust between co-workers because, compared with traditional work environments, remote environments tend to provide less social and contextual information, which makes trust building and maintaining, challenging (Lu et al., 2017;Romeike et al., 2016).Related, the past decade has witnessed shifting away from traditional, primary life-long career paths, toward more dynamic career schemes, with employees more often moving between jobs and looking beyond local job markets (Papavasileiou, 2017).Indeed, with the accelerated transition to remote work, employers expended their target talent pool to include workers irrespective of spatial (geographical) boundaries (Kramer & Kramer, 2020).
Second, with growing globalization and digitalization and increase in complex and interdependent tasks that require the expertise of and collaboration between professionals, global projects composed of cross-cultural teams have become common employment structures (even more so due to the pandemic), designed to foster flexibility, creativity, and timely response (Henderson, Stackman, & Lindekilde, 2018;Lauring & Selmer, 2011;Neeley, 2015).These structures involve short-and long-term relationships between and within organizations across spatial, temporal, and relational distances with the intention to deliver unique products or services (Bizjak & Faganel, 2020;Fossum et al., 2019).As such, many employees engage in work relationships with co-workers of different ethnicities (cultures) (Lauring & Selmer, 2011;Vuori, Helander, & Okkonen, 2019).Acknowledging the potential advantages of teams whose members are culturally dissimilar, researchers also highlight the need to acknowledge and address challenges that could hamper collegial relations and undermine business success (Breuer, Hüffmeier, & Hertel, 2016;Ferrin & Gillespie, 2010;Jiang et al., 2011).Taken together, these trends suggest that contemporary organizations are struggling with significant trust decline that could undermine their ability to meet business goals.Although trust has been at the core of scholarly investigation on the development of high-quality relationships between co-workers (Cropanzano & Mitchell, 2005), one lacuna that yet limits the extant research is that most studies have taken an isolated view on occurrences within relationship dyads to explain the development of trust and thus neglected the possibility that the organization, as referent for employees' trust, may affect trust between co-workers (Fulmer, 2018;Fulmer & Gelfand, 2012).Especially in uncertain situations, trust in the organization can help to overcome conflicts between co-workers and to build stronger relationships.Comparing employees working in a hightrust company with employees working in a low-trust company, Cohen (2015) found that the majority of employees in the high-trust organization reported feeling closer to and working better with their colleagues.Employees' willingness to put effort in work relationships was also higher when they held high levels of trust toward the organization (Cohen, 2015).More broadly, these findings correspond with research showing how employees' level of trust in their employing organization affects their perception of and behavior in the organization, subsequently contributing to higher engagement and performance (Baer et al., 2021;Klijn et al., 2016;Mangus et al., 2020;Nienaber et al., 2015;Paparoidamis, Katsikeas, & Chumpitaz, 2017;Villena, Choi, & Revilla, 2019;Zak, 2017).
Along these lines, our starting point in this study is that trust in dyadic relationships (i.e., trust between coworkers) is affected by trust in other referents because the relationships are not independent from each other (Gillespie, Fulmer, & Lewicki, 2021;Tan & Lim, 2009).It has been argued that boundaries between different trust referents are somewhat permeable, allowing aspects of one trust referent (i.e., the organization) to spill over to another referent (i.e., co-workers) (Fulmer, 2018;Fulmer & Ostroff, 2017).The idea concerning the possible association between organizational and interpersonal phenomena has been tested across a wide range of topics, including psychological contract breach (Bordia et al., 2010), behavioral integrity (Simons et al., 2007), and ethical leadership (Wo, Ambrose, & Schminke, 2015).These studies are based on the idea that a "[…] flow of perceptions, feelings, attitudes and behaviors [trickles down] the organizational hierarchy."(Wo, Schminke, & Ambrose, 2019, p. 2264).Fulmer & Gelfand (2012) presented a systematic review of trust across multiple levels in organizations and discussed the idea that trust within dyadic relationships is affected by (i.e., trickles up or down to) trust in other referents.In a later work, Fulmer and Ostroff (2017) applied the concept of trickle effects to investigate trust in leaders at different hierarchical levels, empirically demonstrating a trickle-up effect of trust in direct leaders on trust in top leaders.Extending this conversation, Tan and Lim (2009) posited that trust in the organization can be understood as a "global composite body of trust" (p.61), which influences trust in referents within the organization.Based on the concepts of halo and priming effects, it can be assumed that trust in the organization creates a positive environment and fosters a favorable evaluation of co-workers within those organizations (Johnson & Grayson, 2005;Staw, Sutton, & Pelled, 1994).Accordingly, in this study, we propose that trust in the organization can trickle-down to trust in referents within that organization (e.g., co-workers).
We further contribute to research on trickle effects of trust by situating these effects in the context of international relationships among co-workers.The rise of globalization and the increase in flexible work arrangements described above have changed the way people collaborate at work (e.g., Buengeler & Den Hartog, 2015;Vuori, Helander, & Okkonen, 2019).Employees of different ethnicities who work together often experience challenges having to do, for example, with language barriers, biased assumptions, and differences in interpretation, with the upshot being greater potential for conflicts (Lauring & Selmer, 2011;Obsuwan et al., 2021).Research has accordingly shown that it is particularly difficult to establish and maintain trust in intercultural relationships, in the absence of common ground and where members have varied levels of familiarity (tenure) working with one another (Ferrin & Gillespie, 2010;Priem & Nystrom, 2014).In light of these labor market trends, it becomes critical for managers to understand and facilitate trustful relationships between coworkers.Addressing this need, our study considers two variables-cultural dissimilarity and relationship lengthas two relationship characteristics potentially influencing the strength of trickle-down effects of trust in crosscultural work collaborations.More specifically, we consider employees working together from the same or different culture, as well as relationships of shorter or longer duration, and how these characteristics affect the degree to which trust in organizations drills down to trust in coworkers.Figure 1 graphically illustrates our model.
We test our research questions with a dynamic hierarchical linear modeling, using data collected in two points in time from a sample of 317 business relationships between Turkish employees and their Turkish and German co-workers.Based on Hofstede (1980), this enables us to contrast employees from an individualistic, Western culture (Germany) with employees from a collectivistic, Eastern culture (Turkey), thus investigating cultures that are at opposite ends of the continuum (Hofstede, 1980;Tsui, Nifadkar, & Ou, 2007).This is an interesting and relevant consideration, because research has found differences in the formation of trust between collectivist and individualistic cultures (Wasti, Tan, & Erdil, 2011).
Our study answers the question of whether trust in an organization can trickle-down to trust between coworkers and thus answers the call put forward by several researchers (e.g., Gillespie, Fulmer, & Lewicki, 2021) to consider different trust referents in a single study and adds to the scarce empirical research that takes a multireferent perspective.Furthermore, in bearing with labor market trends and trends in felt trust that appear to be moving in opposite directions, our study factors in the challenges associated with building and maintaining trust between co-workers of different ethnicities and considers dyad-level conditions under which the trickling of trust across referents may be stronger or weaker.We discuss the potential effects of organizational trust for coworkers and how firms can leverage on the diffusion of trust in-between referents.This includes, for example, the adjustment of various human resource practices (recruitment, mentoring, etc.), to establish and continuously reinforce organizational trust, while being mindful to the opportunities and challenges inherent in diversified workforce (balancing between local cultures and global work environment).

HYPOTHESIS DEVELOPMENT Trust and trickle-down effect of trust
The concept of trust has drawn extensive attention from scholars and practitioners, with much research focusing on the consequences of trust for both employees and employers (for overviews, see Kharouf & Lund, 2019;Nienaber, Hofeditz, & Romeike, 2015;Searle, Nienaber, & Sitkin, 2018).A key underlying assumption is that trust is important for fostering cooperative behavior, with the benefits attributed to trust, including better well-being and performance, facilitated by cooperation between various workplace stakeholders (e.g., Bundy, Vogel, & Zachary, 2018;Costa, Fulmer, & Anderson, 2018;Nerstad et al., 2018).
In this study, we consider individual trust in two referents: the interpersonal referent (i.e., a specific coworker) and the organizational referent (i.e., the employing organization).In keeping with Fulmer andGelfand's (2012, p. 1174) distinction between referents of trust, we define employee trust in a co-worker as "a psychological state comprising willingness to accept vulnerability based on positive expectations of a specific other" and employee trust in the organization as "a psychological state comprising willingness to accept vulnerability based on positive expectations of an organization" (see also Rousseau et al., 1998).
When co-worker trust is high, helping behaviors are more common; employees are more likely to exchange resources (e.g., knowledge; Alexopoulos & Buckley, 2013), to view criticism by co-workers as valuable feedback rather than personal attack, and to make effective team decisions (Arnett & Wittmann, 2014;Costa, Fulmer, & Anderson, 2018;Kim, Dirks, & Cooper, 2009;Nienaber, Holtgrave, & Romeike, 2018).When organizational trust is high, employees are more likely to develop a sense of identification with and belonging to the organization, to voice their concerns rather than withdraw from the organization or engage in negative behaviors, and to perceive the organization as authentic (Polat, 2013;Schoorman, Mayer, & Davis, 2007).
A trickle effect of trust refers to the diffusion, spreading, or spilling-over of trust in one referent to another.This approach suggests that the development of trust relationships takes place in a social network of other trust relationships (Ferrin, Dirks, & Shah, 2006) and that employees form trust judgments based on indirect information regarding other referents (Meyerson, Weick, & Kramer, 1996).Prior research discusses various types of trickle effects (e.g., trickle-down, trickle-up or tricklearound; Wo, Schminke, & Ambrose, 2019).Studies taking a bottom-up approach have shown that trust can trickle-up from direct leaders to top leaders (Fulmer & Ostroff, 2017) or from co-workers to organizations (Tan & Lim, 2009).However, to the best to our knowledge, no study has empirically examined a trickle-down effect of organizational trust to trust in co-workers.Tan and Lim (2009) have called for future research to examine this question.We do so and adopt a top-down approach that highlights the idea that the organization, as an entity, shapes trust relationships among its members (Fulmer, 2018;Fulmer & Gelfand, 2012;Gillespie, Fulmer, & Lewicki, 2021).In a trickle-down effect, employees who trust their organization will emulate the perceptions they hold toward their employer by developing trust in other relationships in their work situation, such as their relationships with co-workers (Mawritz et al., 2012).
The notion of the trickling down of trust assumes that when an employee exhibits high levels of trust toward the employing organization, halo and priming effects may occur (Cooper, 1981).The halo effect suggests that because a co-worker is associated with an organization that the employee trusts (the same organization for which the trustor works), a favorable overall evaluation toward the co-worker is created (Crane, 1965;Perrow, 1961).The priming effect suggests that, often subconsciously, individuals with positive feelings are likely to react more favorably to others (Staw, Sutton, & Pelled, 1994)-that is, working for an organization one trusts is likely to create a general positive environment, which is conducive to the development of trust in one's co-workers (Johnson & Grayson, 2005).Taken together, the halo and priming concepts suggest that organizational trust may serve as heuristic, informing employees in judging their coworkers.Accordingly, we propose the following: Hypothesis 1. Trust in the organization is positively related to trust in co-workers.

Cultural dissimilarity and the trickle-down effect of trust
With growing globalization, intercultural teams have become more prevalent, and organizations are introduced with challenges to develop and sustain trust between employees whose cultural background may have little in common (Ferrin & Gillespie, 2010;Priem & Nystrom, 2014).Hofstede's conceptualization suggests that due to distinct socialization processes, members of different cultures share an underlying programming of the mind, which forms beliefs, assumptions, and perceptions (Hofstede, 1980(Hofstede, , 1991;;Peters & Den Dulk, 2003;Straub, Keil, & Brenner, 1997).Cultural differences in values and attitudes influence individual behavior, norms, and needs, as well of interactions between individuals (Erumban & De Jong, 2006).
To reason the impact of cultural dissimilarity on the trickle-down effect of trust, we build on the concept of common ground, which suggests that interpersonal trust is strongly influenced by information held in common between individuals embedded in a relationship (e.g., Priem & Nystrom, 2014).This concept follows the rational of the similarity-attraction theory and more specifically posits that individuals with demographic resemblance (e.g., in ethnicity) are likely to share mutual values and norms, which cause them to positively reinforce one another's beliefs and attitudes and further increase interpersonal attraction (Byrne, 1971;Priem & Nystrom, 2014).Common ground can be defined as the entirety of mutual ideas, values, and information jointly held by two or more individuals (Stalnaker, 2002) and is based on communal and personal sources (Clark, 1996).The most prevalent form of common ground stems from a shared cultural background or cultural similarity (Yuki et al., 2005).It follows that trust relationships may develop faster in relationships between culturally similar parties, compared with culturally dissimilar parties (Dietz, Gillespie, & Chao, 2010).Zaheer and colleagues (Zaheer & Kamal, 2011;Zaheer & Zaheer, 2006) provide evidence for cultural differences in trust foci and general trust levels, which can result in misunderstandings and trust asymmetries.Such asymmetries in turn can hinder fruitful exchange relationships.Paradoxically, however, trust is markedly important in intercultural relationships due to high uncertainty and complexity having to do, for example, with communication (semantic barriers and word connotations) and work ethics/norms (Dietz, Gillespie, & Chao, 2010).
Prior research, investigating intercultural peer relationships, found evidence that the organizational setting (e.g., organizational culture) is key for addressing challenges that could arise from cultural dissimilarity between employees (Gerhart, 2009).Along these lines, we suggest that the trickle-down effect of organizational trust may be particularly salient for co-workers, who are culturally dissimilar.Our explanation here is twofold.
First, organizational trust corresponds with willingness to rely on organizational rules and procedures-even if they differ from cultural norms (Fulmer & Gelfand, 2012;Six & Sorge, 2008).Referring to this idea, Salazar and Salas (2013) stated that a shared identity formed by the perception that everyone is working on the same global work team shapes a common understanding of tasks and goals.This suggests that potential barriers due to cultural dissimilarities may be reduced by-or have less profound effect-in the presence of organizational features that are salient to the individual employee, such as perceived organizational trust (Liden, Anand, & Vidyarthi, 2016).Second, trust in organizations was found to improve quality of personal relationships among employees and to tone down differences between parties (Van der Berg & Martins, 2013).These effects may be more pronounced in intercultural relationships because the feeling of sharing organizational membership reduces "outsider" perceptions of a culturally dissimilar co-worker.Indeed, from an in-group and out-group perspective, in a shared context, an out-group member shares at least the same (organizational) values and rules as in-group members (Spector & Jones, 2004).
Taken together, the aforementioned arguments suggest that the trickle-down effect of organizational trust may become more evident when there are cultural differences, because culturally dissimilar co-workers have a competing focus of reference and identification (the organization whom they trust), such that they are more likely than culturally similar co-workers to bestow felt organizational trust on their co-workers.Accordingly, we propose a stronger trickle-down effect among employees who are culturally dissimilar to their co-workers.
Hypothesis 2. The trickle-down relationship between organizational and co-worker trust is strengthened among employees who are culturally dissimilar to their co-workers and attenuated among employees who are culturally similar to their co-workers.

Relationship length and the trickle-down effect of trust
Relationships at work may vary in length of time during which co-workers work with one another, ranging from very short relationships, where co-workers perceive each other as merely acquaintances or complete strangers, to very long relationships, where co-workers perceive each other as close friends (Dibble, Levine, & Park, 2012).Relationship length incorporates aspects having to do with the development and maintenance of interpersonal relationships and was found to be a key indicator affecting the nature of relationships between parties in various contexts (e.g., mentor-mentee and buyer-seller) as well as the outcomes associated with such relationships (Claycomb & Frankwick, 2004;Holtgrave, Nienaber, & Ferreira, 2017).For example, co-workers engaged in relationships of longer duration were more likely to share knowledge (Alexopoulos & Buckley, 2013), and mentees showed better academic performance later (than earlier) in the mentorship process (Grossman et al., 2012).Similarly, in marketing research, Dagger, Danaher, and Gibbs (2009) observed a relationship-maturity effect, in which customer-reported relationship strength was higher for longer duration relationships.
Intuitively, one may expect that trust would be positively influenced by relationship length (e.g., Lewicki & Bunker, 1996).When two people are engaged in longer interaction, they learn more about each other and their relationship matures, such that trust may grow.Then again, there could be reasons why trust may not increase over time.For example, early in the relationship, people often use cognitive cues derived from group memberships, stereotypes, and so forth, which may lead to high or low trust.Similarly, as a relationship progresses in time, people observe more of others' behavior and may conclude that these others are more or less trustworthy (e.g., McKnight, Cummings, & Chervany, 1998).In light of these conflicting arguments, it is not surprising that research has found no direct association between relationship length and trust (see meta-analysis by Dirks & Ferrin, 2002).Thus, instead of focusing on main effects, authors have sought to integrate relationship length as a moderator.For example, Levin, Whitener, and Cross (2006) found that relationship length changed the relative importance of different trust bases (e.g., reciprocity and shared goals), such that the influence of these trust bases on trust varied depending on how long individuals have known each other.
Building on these findings, we propose that relationship length plays a moderating role in the diffusion of trust from the organizational to the interpersonal level.We draw from the filter theory of relationship development (Duck et al., 1991) to suggest that the relative importance of trust in the organization to trust in coworkers (i.e., the trickle-down effect of trust) may shift as co-worker relationship develops.Filter theory suggests that different information about one's partner is important and available at different phases of a relationship.The theory explains how deeper level personal information about the parties becomes increasingly available, and is increasingly sought, as relationships progress overtime.Yet at the early stages of relationships, parties have relatively little personal information with which they can form meaningful ties with one another.As such, newly formed relationships are often characterized with uncertainty (partners are not yet certain of the nature and outcomes of their relationship), and the organization may be a more salient source for identification.In such circumstances, individuals rely more on the organization to provide guidance in interpreting the work situation, with perceptions of organizational trust more easily trickling down to trust in co-workers.In their conceptual model of trust formation in organizations, McKnight, Cummings, and Chervany (1998) referred to the role of organizational trust in laying out the foundations for trust between members, also considering the time dimension.They suggested that the institution in which trust takes place is particularly important for establishing high levels of interpersonal trust in initial phases of relationships.
However, as relationships continue, individuals observe more of others' behavior, processing this deeper level information to either support or disconfirm earlier inferences (and potentially, end poor relationships) (Anderson & Weitz, 1989;Duck et al., 1991).Thus, as individuals (re)interpret their observations, they increasingly base their trust less on broader, general, organizationally derived information and more on the information they are processing directly from their observations of one another's trustworthy behavior.More nuanced perceptions about others, which are based on behavioral exemplars, are then invoked (Levin, Whitener, & Cross, 2006).Co-workers thus have a more ground upon which to evaluate each other (as trustworthy or not)-irrespective of their view about the organization, and so organizational trust may be less important in prescribing co-worker trust.And we propose the following: Hypothesis 3. The trickle-down relationship between organizational and co-worker trust is attenuated as a function of relationship length.

METHOD Sample and data collection
Data were collected from Turkish employees working in the Istanbul offices of multinational firms with coworkers from Turkey and Germany.We sought to contrast two cultures that are rather distant from one another.Turkey is a traditional and patriarchal country in the process of modernization, a predominantly Islamic country that has re-oriented itself to the West in the past decades (Ertürk, 2008;Sozen & Shaw, 2003).Germany is a developed, largely egalitarian, and secular country, although Christianity helped form its culture (Diehl, Koenig, & Ruckdeschel, 2009).The countries rely on different types of economic systems, a government-guided market economy (Turkey), and a social market economy (Germany) (Pudelko, 2006).Notably, Germany and Turkey differ along the individualism/collectivism dimension of Hofstede's cultural model (1980).Employees of individualistic cultures (Germans) often view themselves as independent from others, such that they consider their own thoughts, needs, and feelings as unique and possibly different from those of others.They appreciate personal fulfillment and embrace change and deviation from the status quo (Hofstede, 1980).Employees of collectivistic cultures (Turkish), however, tend to define their identity by membership of social groups such as family or social class (Hofstede, 1980;Top, Akdere, & Tarcan, 2015).
We identified participants based on a list provided by the chamber of commerce and industry for Turkish-German business relations.Based on this list, we contacted the Istanbul offices of 600 randomly selected firms, sending them an introductory participation-request letter and summary of the proposed research and asking them to provide us with the names of a Turkish employee who would be willing to participate in our study.Initially, 441 firms agreed to participate (response rate 71.89%), and we contacted the corresponding employees via personalized mailings.
Data were self-reported.To minimize concerns related to common method variance, we gave specific instructions on how to fill out the survey, assured respondent anonymity, applied well-established items with alternating scales, included reverse-coded items, and separated exogenous from endogenous variables.Finally, we applied temporal separation by collecting data in two points in time, 6 weeks apart (Podsakoff et al., 2003).Unique codes were assigned to the sample participants to match their Time 1 (T1) and Time 2 (T2) data.
At T1, participants were first asked to rate their level of trust in their employing organization and answer several questions concerning control variables.In addition, they were asked to name four co-workers with whom they work regularly and that were similar to them in as many attributes as possible, including job roles, location, and hierarchical level.Of these, two had to be of Turkish ethnicity and two expatriates of German ethnicity.For T2 survey, we randomly choose two of the employees' selected co-workers-one of Turkish and one of German ethnicity.The survey participants were then asked to rate their levels of trust in these two co-workers.From the initial 441 firms, we obtained usable responses to the questionnaires from 317 Turkish employees, providing us with data on a total of 634 employee-co-worker relationships of either an entirely collectivistic nature (Turkish-Turkish) or mixed collectivistic-individualistic nature (Turkish-German; Hofstede, 1980).
The main industries included in our sample were manufacturing (24.9%), consumer products (22.7%), information technology (7.3%), medicine/pharmaceutical (5.0%), and finance (4.7%).Of the 317 firms, 238 (75.1%) were publicly listed and 79 (24.9%) were privately owned.Regarding the selected Turkish employees who responded to our survey, 159 (50.2%) were female.The majority of respondents held nonmanagerial position (66.6%), while the rest of the respondents belonged to the middle management.We did not include top management or owners of a company as those are often seen as substitutes for organizational trust.The time the respondents had been working with their Turkish coworkers ranged from 2 months to 19 years.The time the respondents had been working with their German coworkers ranged from 1 month to 19 years.
The low number of German colleagues from which we could choose did not allow for an exact like-for-like comparison in the selection of the Turkish and German co-workers.By instructing our respondents to select two co-workers similar in as many attributes as possible, we did our best to ensure a high comparability of the chosen co-workers.However, a paired-samples t-test revealed significant differences between the groups of Turkish and German co-workers' hierarchical levels and the time the respondents had been working with the selected coworkers.Hence, we controlled for these and other coworker-related aspects in our analyses.

Measures
We followed Brislin's (1970) recommendations for the translation of the English-language original items into Turkish.The questionnaire was pretested through indepth interviews.Based on the feedback received in these interviews, few items were revised to enhance clarity.All constructs, unless otherwise stated, were answered with a 5-point Likert response scale.
We measured organizational trust by adopting the original 10-item scale of Searle et al. (2011).We adapted the scale to our context and had to drop three items as they were dropped due to unsatisfactory loadings.Example items include "My firm is capable of meeting its responsibilities" and "Employees" needs and desires are important to my firm'.Data analysis confirmed that onefactor solution fitted our data best (Table 1).To measure co-worker trust, we adopted the 10-item scale by McAllister's (1995).A pretest of the original scale revealed that Turkish employees felt uncomfortable answering items like "Me and my colleague are emotionally involved in our working relationship."These items were thus dropped, and we proceeded with a refined sevenitem version (e.g., "My colleague responds constructively and caringly to my problems.").For co-worker trust too, we found that one-factor solution fitted our data best (see Table 1).This could be caused by the adaptation of the scale due to our intercultural setting, as culture is strongly linked to the understanding of trust (Dietz, Gillespie, & Chao, 2010).
Cultural dissimilarity.In line with Jiang et al. (2011), cultural dissimilarity was captured via a dummy variable, coded 1 for an individual's rating of a co-worker with a different cultural background (Turkish-German relationship) and 0 denoting otherwise (Turkish-Turkish relationship).To capture their nationality, we did not ask participants to report their country of origin.Rather, we requested that they self-categorize themselves, that is, each participant affiliated himself/herself as Turkish, German, or other nationality ("Please indicate the nationality with which you are affiliated: 'German', 'Turkish' or 'Other'").This approach is consistent with prior research on teams and dyadic relationships: It is common for culture to be ascribed, not measured, using nationality as a proxy for culture (e.g., Connaughton & Shuffler, 2007;Stahl et al., 2010;Zellmer-Bruhn & Maloney, 2020).For example, Hardin, Fuller, and Davison (2007) applied Hofstede's values theory in their study and used nationality as a proxy for culture, upon which they attributed cultural values of individualism and collectivism to compare differences between team members from Hong Kong and the United States.
Relationship length.Following Jiang et al. (2011), the duration of the relationship between an employee and his/her respective co-worker was operationalized as the total number of months an employee has been working with the Turkish and German co-workers.
Control variables.We controlled for several variables that may influence co-worker relationships, including hierarchical level and communication frequency (the frequency of face-to-face and virtual communication).We also controlled for respondents' tenure and gender, as well as organizational ownership, size, and age.

Measure validation and preliminary analyses
Before analyzing our hypotheses, we assessed the validity and reliability of our measures following the guideline of Anderson and Gerbing (1988).First, we computed the reliability coefficient α (Cronbach's alpha) for each multi-item scale.All coefficients exceeded the commonly accepted benchmark of .70 (Ambrose, Hess, & Ganesan, 2007).Second, we conducted confirmatory factor analyses to evaluate the convergent validity and reliability of our measures.As noted above, three items with poor loadings were dropped from the organizational trust scale.All other factor loadings were above .40(p < .01),indicating the unidimensionality of our measures (Bagozzi & Baumgartner, 1994).The composite reliabilities (ranging from .82 to .95) were all well above the standard of .60 (Bagozzi & Yi, 1988), and the average variance-extracted (AVE) indices were all greater than the .50benchmark (Fornell & Larcker, 1981) except for individual trust in the organization (.415).However, due to a good Cronbach's alpha, we kept this factor, as other authors suggested (e.g., Ping, 2009).Additionally, we assessed discriminant validity by confirming that each construct's variance shared with other constructs was lower than its AVE (Fornell & Larcker, 1981).Multicollinearity also did not seem to pose a problem, as the highest VIF value was 2.65.Lastly, we assessed the overall fit of our model.As shown in Table 1, our hypothesized four-factor model provided a good fit to the data (χ 2 (df) = 46.98 (28); χ 2 / df = 1.68; comparative fit index [CFI] = .99;Tucker-Lewis index [TLI] = .99;standardized root mean square residual [SRMR] = .027;root-mean-square error of approximation [RMSEA] = .033).We then used the nested model testing procedure to examine the adequacy of our measures by comparing the four-factor solution to several alternatives.These included a model combining co-worker and organizational trust into one-factor (three-factor model) and a single-factor model.All these models resulted in a significant decrease in fit as indicated by the chi-square difference test.Hence, we proceeded with the four-factor solution.
Because our data are nonindependent (each Turkish employee rated his or her trust in two co-workers, one of the same and one of different [German] cultural ethnicity), analyses that do not consider the nesting of data may potentially misrepresent the effects tested in our model (Klein, Dansereau, & Hall, 1994).Addressing this issue, we used random-effects regressions to analyze our trickle-down effect.This is important, because all our key variables were measured at the individual level.We note that contrary to fixed-effects models, random-effects models assume that any individual-specific unobserved factors are uncorrelated, an assumption that we found to be valid for our data as indicated by the Hausman test (T H = 6.16, p = .29;Hausman, 1978).

RESULTS
Mean, standard deviations, and correlations of the study variables are shown in Table 2.
As indicated by the change in chi-square (χ 2 ), the full model including interaction effects (Model 3) fits our data best.The results of the random-effects regression are summarized in Table 3.
Consistent with Hypothesis 1, Model 2 shows that the trickle-down effect is positive and significant (β = .28,p < .01).Hypothesis 2 proposed a moderating role for cultural dissimilarity in the trickle-down effect.Our results offer support for the proposed moderation effect.The results (see Model 3) suggest that the more culturally dissimilar the co-workers (the less common cultural background they share), the stronger the effect of organizational trust on co-worker trust (β = .07,p < .05). Figure 2 illustrates this interaction effect.
Finally, Hypothesis 3 was not supported.We did not find evidence for a moderating effect of relationship length on the association between organizational trust and co-worker trust.

DISCUSSION
Globalization has had a significant impact on how, when, and where work is done, bringing many workers into frequent contact with colleagues from different countries, in what had been termed "the age of global team-based work" (Wildman & Griffith, 2015, p. 1).And while cultural heterogeneity can result in positive outcomes like greater responsiveness to customers and more creative solutions (Lauring & Selmer, 2011), managerial attention and skills are required to leverage on the complexity inherent in cross-cultural teams to elevate their performance (Breuer, Hüffmeier, & Hertel, 2016;Ferrin & Gillespie, 2010;Henderson, Stackman, & Lindekilde, 2018;Neeley, 2015;Wang et al., 2019).One of the key determinants of the success of these teams is trust, which was found to contribute to the development of high-quality relationships (Cropanzano & Mitchell, 2005).In this study we, first, sought to respond to the call set by a number of scholars for the integration of multiple trust referents within organizations (Fulmer, 2018;Gillespie, Fulmer, & Lewicki, 2021).We proposed and tested the possibility that trust in the organization trickles down to trust in coworkers, thus helping to overcome potential difficulties in multicultural work environments.Our results provide the first empirical evidence of a trickle-down effect, indicating that trust judgments about co-workers are shaped in part by the trustworthiness of an organization.More broadly, this finding could nourish future studies interested, for example, in how organizations' trust in a network (e.g., strategic alliances) might trickle-down to trust in any member organization (by its employees) or between partner organizations (e.g., Balboni, Marchi, & Vignola, 2018;Kharouf & Lund, 2019).Second, to provide a nuanced understanding of when or for whom trickle-down effects are more likely to occur (Wo, Schminke, & Ambrose, 2019), we considered two relationship characteristics that are particularly relevant in the context of current labor market trendscultural dissimilarity and relationships length-as moderators influencing the strength of the trickle-down effect of trust.Drawing on the concept of common ground, we argued that cultural dissimilarity moderates the trickledown effect.Our results show that the more culturally dissimilar co-workers, the stronger the trickle-down effect of organizational trust on co-worker trust.Thus, organizational trust becomes even more important in companies employing ethnically diversified workforce as the level of organizational trust is more likely to be instilled in trustworthy co-worker Our multireferent analysis underlines the significance of halo and priming effects (Crane, 1965;Perrow, 1961) for individual perceptions, not pertaining on common ground.The halo effect is evident in our results showing how a favorable overall evaluation of the organization positively influences trust between co-workers.And the priming effect occurs in our study because, potentially subconsciously, individuals with positive feelings are likely to react more favorably to others.Consistent with the findings of other scholars (Crane, 1965;Johnson & Grayson, 2005;Perrow, 1961;Staw, Sutton, & Pelled, 1994), we observed how organizational trust functions as heuristic, informing employees in judging their co-workers.
Interestingly, and contrary to notions conveyed in the filter theory of relationship development (Duck et al., 1991), relationship length had no role in the trickledown effect of trust.In this respect, our findings somewhat divert from findings reported in earlier studies (e.g., Alexopoulos & Buckley, 2013;Dagger, Danaher, & Gibbs, 2009;Levin, Whitener, & Cross, 2006), which attributed varying impact of trust depending on whether co-workers work with each other for short or long period of time.The finding that the effect of organizational trust on trust between co-workers remains at the same level, not becoming less important as collegial relationships mature, may be explained by theories of swift or presumptive trust (Meyerson, Weick, & Kramer, 1996;McKnight, Cummings, & Chervany, 1998).These theories are based on the assumption that trust develops prior to interaction, not through interaction (i.e., prior to any knowledge of team members' behavior).It is argued that initial trust can occur among co-workers who have very recently met, for various reasons, such as their shared membership in the same organization.This does not mean that their level of trust in the organization would more easily drill down to their trust in co-workers at an early stage of relationship, but possibly, that belonging in the same social category (employment in the same firm) provides a good starting point on which they can build initial trustirrespective of their trust toward the organization at this point.This explanation underlines the idea of "initial trust," which focuses on the relevance of trust in the early stages of relationships (McKnight, Cummings, & Chervany, 1998).
Finally, we like to raise a note about the role of communication in trust between co-workers.As described above, frequency of communication was controlled for in our analyses, consistent with prior research.Not surprisingly, we observed a positive effect of face-to-face communication on co-worker trust (e.g., Postmes, Spears, & Lea, 1998).However, our results also reveal no negative effect on trust when co-workers engage in frequent virtual communication.These results add to emerging evidence that challenges prevailing assumptions about how trust develops in geographically distributed and remote teams (Wilson, Straus, & McEvily, 2006).Specifically, developments in information and communication technology (ICT) and technology-enabled meetings (e.g., video conferences) have made these media rich, accessible, and easy to use.Our findings suggest that ICT may have become less challenging for capturing and facilitating relationships.
F I G U R E 2 Interaction effect of cultural dissimilarity in the trickle-down effect

THEORETICAL AND PRACTICAL IMPLICATIONS
The results of our study provide two important theoretical implications.First, our results contribute to the discussion on the complex ways in which trust relationships can develop, and in particular, on how trust in one referent can trickle-down (i.e., have a spillover effect) into trust in another referent (Hardin, 2013).Whereas prior research has shown that trust can trickle-up (e.g., from co-workers to the organization; Tan & Lim, 2009) or trickle-around (from one organization to another organization; Høyer & Mønness, 2016), no study to date has prospectively investigated to potential benefits of topdown spirals of trust.As such, our study adds another piece to the puzzle related to trust dynamism, further establishing trust as a phenomenon that processes in multiple directions across different referents (co-workers, mangers, and organization).
Second, we found that cultural dissimilarity is a boundary condition that influences the magnitude of the trickle-down effect of trust.With this, we are able to extend the conversation on the impact of cultural differences, illustrating how employees who view their organization as trustworthy may replicate such trust perceptions in their relations with co-workers-despite cultural dissimilarities.In other words, this finding emphasizes the importance of organizational trust as a means to compensate for lack of common ground among culturally different co-workers, trickling down to trust felt between them (Gerhart, 2009).
In terms of practice, the evidence provided in this study for trickling down of trust between organizational and interpersonal referents highlights the need for organizations to invest resources to develop their trustworthiness.First, the absence of trust can strain interpersonal relationships, giving rise to conflicts and fallouts.Managers could demonstrate organizational trustworthiness via observable benevolent and integrity manifestations such as promise keeping, fair and nondiscriminating procedures, and willingness to admit mistakes and to take responsibility for failures (Biron, 2010;Searle et al., 2011).Organizations can also position themselves as trustworthy by internally and externally communicating about and rewarding activities that are judged as socially legitimate, ethical, and consistent.
Second, it is important to include trust-building mechanisms early in the recruitment and socialization processes to help prospective employees and newcomers to form trust perceptions (Van der Werff & Buckley, 2017).For example, newcomers should be introduced to the organization's vision and values as principles that are expected to guide interactions between members (Six & Sorge, 2008).Newcomers may also be mentored by other employees who may highlight topics related to integrity and ethics.Notably, organizations should continually identify and remedy trust-breaching behaviors-by employees and supervisors-to signal expectations from and commitment to members.
Third, our findings stress the need for organizations to be minded of the cultural composition of their workforce and to be aware of the critical role that organizational trust can prevail to foster interpersonal trust in intercultural settings.As more and more organizations go global and with intensified workforce mobility worldwide, cultural diversity in organizations has become paramount and will most likely become even more critical in the future.Given the effect observed in our research for cultural dissimilarity, the establishment of organizational trust is particularly valuable against the backdrop of an increasingly intercultural workforce (e.g., Lauring & Selmer, 2011).Challenges to establish trust may differ depending on the size and geographical outreach of the organization.Small (er) size firms that are not geographically dispersed may try to bring nonlocal employees closer with the local culture, for example, by exposing employees to local traditions, food, and art.By exchange, this may help members feel more welcomed in and connected to the organization.At the same time, multinational organizations should opt for the formation of global, transnational work environment, alongside a clear image of the organization-as a vivid entity-in the eyes of its members from around the world and supported by public relations and communication (logo, in person annual conferences, virtual events, trainings etc.) might translate into strong organizational identity and affinity (Lauring & Selmer, 2011;Wilkins, Butt, & Annabi, 2018).
In sum, recognizing the substituting effect that organizational trust may have in enhancing trust among culturally dissimilar co-workers, our recommendations here illustrate the idea that a positive shared work environment helps firms to build trust that can minimize the negative association often observed in global teams between cultural distance and interpersonal trust.

LIMITATIONS AND OPPORTUNITIES FOR FUTURE RESEARCH
Our study has limitations, which offer avenues for future research.First, we relied on self-reported data by Turkish employees, which may lead to common method bias or cause selection effects.However, as we are primarily interested in investigating psychological processes (Conway & Lance, 2010), we believe that our methodological approach is adequate to understand trickle-down effects of trust in organizations.
Next, although we collected data at two points in time, our design does not capture the full dynamics within trust relationships (or the evolution of trust over time; Searle, Nienaber, & Sitkin, 2018).This means that the inferences within this study still largely rely on a theoretical rather than empirical foundation (Podsakoff et al., 2003).This may also suggest that instead of trust trickling down between the organizational and the interpersonal referent, a trickle-around effect could be considered in the future.In this respect and while we feel confident that theoretical evidence (e.g., Fulmer & Gelfand, 2012) and prior research in related fields (e.g., Wo, Ambrose, & Schminke, 2015) provide sound basis for our trickle-down model, we encourage researchers to apply a longitudinal design to better capture trust's dynamic.This would also allow for accounting for internal events (e.g., team conflicts) or external events (e.g., political developments, such as changes in the relationship between Turkey and Germany over the past decades) that may have influenced the trust level between co-workers.
Further, our investigation of cultural dissimilarity is limited to Turkish and German cultures, and there is a need for future research to explore whether the results are applicable to employees from other countries, especially countries that are of greater cultural dissimilarity to Germany and Turkey.Related, our results may offer a one-sided view by focusing on Turkish employees only.Contrasting our findings with the individualistic (e.g., German) perspective may provide another promising avenue for future research.However, according to Hofstede's conceptualization, the results are generalizable to a wider range of relationships between employees characterized as more individualistic, who are working with rather collectivistic employees, like German and Turkish employees in our sample.Indeed, based on prior research on cultural distance and its impact on differences in attitudes and behavior, it is likely that our findings can be generalized to employees of nationalities that rank close to Germany and Turkey in terms of cultural values (e.g., Cheng & Leung, 2012).Thus, our results are likely to be of interest to many organizations worldwide.Finally, it is important to note that the trickle-down effect was observed in our data irrespective of the cultural aspect.As such, this finding can generally prompt organizations to acknowledge and make use of different trust referents, as discussed below.
Another limitation is that our participants were embedded in project teams and thus may have been influenced by other team-related factors such as support from team leader or team cohesiveness.The increase of the use of such teams in organizations makes our study particularly timely, and future research may further investigate how variables at the team level could impact trickle-down (and trickle-up) effects of trust.
Finally, researcher may broaden our model, which has focused on cultural similarity-a rather fixed feature-by considering other personal characteristics that could establish a sense of common ground.

CONCLUSIONS
Our study provides empirical evidence for a trickling down of employees' trust between organizational and interpersonal referents.We thus capture interrelations between employees' trust in the organization as an entity and trust in their co-workers by taking a multireferent perspective.These findings highlight the importance for future research to continue to consider the interrelationship of different trust referents within one study.Further, we highlight cultural dissimilarity as a relationship characteristic that shapes the magnitude of trickle-down effects of trust.Overall, this study emphasizes how important it is for managers to build high-trust organizations particularly to establish trust among members of culturally diverse teams.

ORCID
Viktoria Maria Baumeister https://orcid.org/0000-0003-2119-6762 Random-effects regression analysis on trust Above reported coefficients are standardized.Robust standard errors are reported in parentheses.Continuous variables used in the interaction terms have been mean-centered.Cultural dissimilarity is coded 1 for German co-workers (i.e., different cultural ethnicity) and 0 for Turkish co-workers (i.e., same cultural ethnicity).Organizational ownership is coded 1 for publicly listed firms and 0 otherwise.Gender is coded 1 for female and 0 for male.