The psychological lifeline for American labor union survival is the local union meeting (see Ezorsky, 2017; Mellor & Holzer, 2018; Stagner, 1981; Tannenbaum & Kahn, 1958, among many others). Within the context of a union as a voluntary member-driven organization, local meetings are designed to allow and encourage employees to experience safe fulfillment of psychological needs, exemplars of which include opportunity to voice concerns and opinions, to participate in decision-making, to seek and receive help from others, to provide help to others, to be recognized and accepted as an individual who is “one of us,” and importantly, to be valued as someone who shares in and contributes to “our collective effort” (Greenhouse, 2019; McAlevey, 2020; Mellor, 2019; Mellor & Holzer, 2018).Footnote 1 The extent to which local meetings are experienced as safe environments in which employees can strive to fulfill such needs is an issue that we think bears on and provides a solution to a problem that threatens union survival: the problem of chronic low meeting attendance (see Monnot et al., 2011; Rosenfeld, 2014; Tetrick et al., 2007; Wiegand & Bruno, 2018 for like recognition of the issue).

As such, we constructed and tested a prediction model of local union meeting attendance, a model informed by literatures on team effectiveness featuring psychological safety as a psychological construct, meeting design featuring non-psychological safety constructs, and union participation in local activities featuring economic-inspired and attitudinal constructs. As a demonstration of the fit of the model, we collected survey data from employees attending local meetings. Consistent with our aim, we used analytic tests to distinguish the model and to suggest an intervention to address the problem of low meeting attendance. Throughout, we take a decidedly psychological and mediational point of view, in which, consistent with the cited literature, we suggest that the relationship between the experience of psychological safety at local meetings and meeting attendance unfolds through meeting effectiveness as rated by attending employees. Also throughout, we equate local meetings with meetings in nonunion work environments, in which, common to both, meetings are attended by employees who meet on a scheduled basis to coordinate their skills and efforts to effect group outcomes that enhance both self-goals and organizational-goals (see Mathieu et al., 2018; Salas & Fiore, 2012 for parallel definitions).

Before presenting the literature reviewed to construct our model, we should note that hard numeric information about local union meeting attendance is hard to come by. Although local meeting attendance is recorded and archived by unions, the data are considered proprietary (for understandable reasons). As a numeric illustration of local attendance, we turned to the union participation literature. We extracted from the literature American samples of employees eligible to attend meetings (sample N = 19) and recorded the percent of attendance within various reported spans of time (commonly, the last 12 months). The average attendance of employees across samples was 26%; the median was 32%. The range of attendance was between 3 and 43%. To corroborate this information, we contacted various union officials via email through posted websites. Based on a low response rate (15% of 40 sent emails), officials indicated average attendance as low as 10% and as high as 50%.Footnote 2

Literature

Psychological Safety

As a psychological construct introduced by Edmondson (1999; Edmondson & Lei, 2014) to predict “team efficacy and performance,” psychological safety is defined as a shared belief by individual team members that the team is safe for interpersonal risk-taking. The definition is meant to suggest a sense of confidence experienced by team members that the team will not embarrass, reject, or punish any member for speaking up. Not intended as an explicit team goal, this shared belief is thought to emerge from the experience of interactions among team members—interactions in which members encourage each other to engage in conversational turn-taking, to speak roughly in the same proportion, to express and be open to a diversity of ideas and perspectives (even if discussion takes a critical turn or challenges team norms), to exchange personal information intended to allow others to know what feelings are in play and what is being left unsaid, and to exercise a form of social sensitivity, in which attending to and acting upon what others feel—especially in regard to being upset, distracted, or left out—is regarded as a normal part of team life. It is from these kinds of interactions that climate properties are thought to emerge such as interpersonal trust and mutual respect—properties that provide a safe environment for individuals to suggest and explore creative solutions aimed at team goals, to share older and newer knowledge skills (to learn from each other), and all the while to experience, as stated by Edmondson (2018), “a comfortable sense of being themselves.”

Uncommon but not rare, psychological safety as a construct has made the leap from the academic world to the world of application and intervention. As indicated in corporate periodicals and corporate guidebooks (e.g., Understand team effectiveness, 2017; Van Bavel & Packer, 2021), psychological safety is widely regarded as an indispensable tool to diagnose poor team performance and to maximize team effectiveness, outcomes of which are couched in terms of team success.

The origin story for the crossover begins with Google’s “quest to build the perfect team” (see Duhigg, 2016; Rozovsky, 2015 for full accounts). In-house researchers at Google set out to determine what factors made for the most effective teams. They collected data from 180 engineering and sales teams to identify skills, personality types, backgrounds, and demographics for team effectiveness. Looking for patterns in the data, they found none—or as stated by one of the lead researchers, “The ‘who’ part of the equation didn’t seem to matter.”Footnote 3 Returning to square-one, these researchers refocused their attention on how employees interact at team meetings. Based on this tack, they finally discovered what did matter: Group dynamics were the key to team effectiveness. Importing new psychological and non-psychological measures to collect data, including Edmondson’s (1999) measure of psychological safety, they discovered that “far and away” the best predictor of team effectiveness in the data was psychological safety (Rozovsky, 2015).

This “discovery” inverted and extended Google’s approach to identifying and maximizing team effectiveness. First, rather than focus on effectiveness as a predictor of team success, focus shifted to psychological safety, in which effectiveness was thought to follow safety (i.e., “safety predicts effectiveness”). Second, with effectiveness as an outcome of safety, team success was thought to follow effectiveness (i.e., “effectiveness predicts success”). This prediction sequence became the center of Google’s intervention strategy to “build more successful teams.” Confirmed by post-intervention data drawn from Google records, ratings by Google executives, and employee data, linked success included higher team productivity (e.g., meeting goals on time; less time on tasks), greater team creativity (e.g., more willingness to consider and incorporate diverse ideas), more stable teams (fewer absences for any reason; lower rates of leaving Google), and greater team satisfaction (a variable that was also shown to “radiate” to other satisfaction targets like customer satisfaction) (see Understand team effectiveness, 2017 for a more complete account). Also, suggested by these stated links—and germane to our study—is implied mediation, in which the effect of safety on success is thought to unfold through effectiveness.

Non-Psychological Safety

Constructs overlapping with psychological safety (e.g., “interpersonal dynamics”) can be found in studies framed in reference to “the science of meetings” (Allen et al., 2015; Rogelberg, 2019), with ensuing diagnostic tools (e.g., Hoffman, 2018; Rogelberg, 2019). Conducted in nonunion environments, this research is comprehensive and includes many constructs that focus on meeting design variables positioned as predictors of meeting success (e.g., Allen et al., 2018; Cohen et al., 2011; Leach et al., 2009; Rogelberg et al., 2014). A sample of such predictors linked to success are meeting composition (only critical personnel should be included at meetings; no one who need not be included should be invited), meeting lateness (chronic lateness to meetings should not be tolerated by attendees; meetings should start and end on times announced in advance of a meeting), meeting agenda (an agenda with stated meeting goals should be distributed to attendees in advance of a meeting along with necessary tools and materials), leader preparedness (leaders should be fully prepared to stick to the meeting agenda, instrumentally guiding attendees to stay on topic), and meeting summary (meetings should end with decision summaries that include when and by whom follow-up work can be expected). To be noted is that these and other design variables show direct links with meeting success, criteria of which include “well-attended meetings.” Our study interest in meeting design predictors of success aligns with our aim to show the effect of psychological safety as an independent and applicable predictor of meeting attendance.

Union Participation

Union participation research has not been silent on introducing variables to predict local activity by employees, albeit most studies that include meeting attendance do not per se indicate predictors of attendance. Rather, in these studies, predictors are reserved for union participation as a global indicator of activity, in which meeting attendance is one of several scored and summed activities (e.g., Hammer & Wazeter, 1993; McShane, 1986; Parks et al., 1995; Wiegand & Bruno, 2018). Also, in these studies, predictors of participation rely on economic-inspired constructs of questionable psychological relevance, constructs such as utility, instrumentality, exchange, and cost–benefit (e.g., Flood, 1993; Klandersmans, 1984; Lund & Taylor, 2010; Tetrick et al., 2007). Although these constructs show links to participation, they present participation as evidence of rational self-interest, with nary a referent to needs or to fulfillment of needs through interactions among local employees (for an exception, see Stagner, 1950). Studies that include attitudinal predictors of participation operate in kind. Predictors such as prounion beliefs and union commitment show positive links to participation, but how could they not—only an irrational employee with strong prounion sentiment or strong commitment would not participate (see Monnot et al., 2011; Tetrick et al., 2007 for relevant studies). Moreover, predictors from this literature have yet to inspire interventions with participation in mind; rather, the interest has been thematically theoretical and explanatory.

However, design variable for “successful meeting attendance” abound in this literature, especially in older studies that suggest “practical recommendations” based on survey and interview data collected from eligible employees (e.g., Dean, 1954; Kahn & Tannenbaum, 1954; Miller & Young, 1955; Purcell, 1954; Rose, 1952; Rosen & Rosen, 1955; Sayles & Strauss, 1953; Stagner, 1956; see also Parker & Gruelle, 1999). Example recommendations we think are yet viable include efforts to extend invitations to employees to attend meetings, especially invitations extended by a local representative (“the local rep”), to perfect meeting flyers (perhaps now also emails and text messages) distributed/sent in advance of a meeting that include the meeting agenda and assurances that workplace issues (e.g., “an unreasonable job demand”) will take priority over national union issues, to communicate to employees how many employees attended the last meeting and are expected to attend the next as a bid to suggest the idea of “missing out,” and, as the sine qua non of recommendations, to plan and conduct “on the clock” short meetings.

Model and Hypothesis

In reference to our psychological view of local union meetings as designed to allow and encourage employees to experience safe fulfillment of psychological needs (exemplars of which have been indicated), we think psychological safety at meetings positioned as a predictor of likely meeting attendance is justified. However, in reference to the cited literature on team effectiveness, equally justified is the prospect that psychological safety indirectly effects meeting attendance through meeting effectiveness as rated by attending employees. This prospect positions meeting effectiveness as a mediator of the effect of psychological safety on meeting attendance (see the hypothesized model depicted in Fig. 1).

Fig. 1
figure 1

Hypothesized model

Full size image

The mediated relationship is hypothesized as follows:

  • Hypothesis: In reference to local union meetings, employees who experience more psychological safety at meetings are more likely to attend meetings in the next 12 months. The path from psychological safety to meeting attendance in the next 12 months unfolds as a sequence with meeting effectiveness as a mediator, such that more psychological safety is associated with higher rated meeting effectiveness, which in turn is associated with more likely to attend.

Method

Procedure

Beginning in June 2021 and ending in December 2021, survey data were collected from American employees. Survey sites included public transportation areas, licensed bingo halls, farmers’ markets, and union- and civic-sponsored community events.Footnote 4

With permission obtained at each site, the researchers circulated flyers with the following information:

Can you volunteer to take this survey? You can if you are employed in the United States and not a full-time student. The survey is anonymous—no names. The survey takes less than 10 minutes to complete. The survey cannot be mailed. $5 is given for taking the survey. Please ask the researcher for a survey.

Employees who responded to the flyer were given an information sheet, a survey, a pencil, and an unmarked envelope. The researchers collected sealed envelopes, paid participants, and conducted onsite debriefing.

Sampling

To ensure that sampling resulted in data appropriate to test the hypothesized mediation, the survey was embedded with eligibility items. We excluded surveys in which responses suggested: (a) noncurrent union membership and (b) nonattendance of at least one regular scheduled local meeting in the last 12 months. An additional check for careless responses resulted in excluded surveys if responses indicated the same scale anchor for long strings of consecutive items.

From a pool of 302 returned surveys with no missing data, 132 surveys were counted as eligible. Eligible surveys included employees from six U.S. States (Connecticut, Massachusetts, New Jersey, New York, Pennsylvania, Rhode Island) and the District of Columbia. Surveys included employees with memberships in 20 unions and 42 locals affiliated with the American Federation of Labor and Congress of Industrial Organizations (AFL-CIO; e.g., American Federation of State, County, and Municipal Employees (AFSCME), American Federation of Teachers (AFT), International Association of Bridge, Structural, Ornamental and Reinforcing Iron Workers (IW), International Brotherhood of Electrical Workers (IBEW), International Brotherhood of Teamsters (IBT), International Longshore and Warehouse Union (ILWU), International Union of Operating Engineers (IUOE), National Education Association (NEA), Service Employees International Union (SEIU), and United Auto Workers (UAW)).

Measures

Demographics

Assessed demographics included age (indicated in years), gender (coded as either man (0) or woman (1)); ethnic group (coded as either non-ethnic (0, White, European American) or ethnic (1, African American, Asian, Pacific Islander American, Latinx American, Middle Eastern, Arabian American)); English as a second language (coded as either English as a first language (0) or English as a second language (1)); socioeconomic status, in reference to “education level (highest degree), contribution to family income, and occupational job status” (response options: lower class (1), lower middle class (2), middle class (3), upper middle class (4), upper class (5)); and employment status (coded as either part-time (0, less than 35 h a week) or full-time (1, 35 h or more a week)).

Ninety-five percent of employees were age 25 years or older (the median age was 47; the range in years was 21 to 75). Fifty-one percent were men employees. Seventeen percent identified themselves as ethnic. Five percent identified English as a second language. Sixty percent identified themselves as middle class or lower (the median class was middle class; no one identified themselves as upper class). Eighty-one percent were full-time employees.

Assessed demographics specific to local unions included length of local membership (indicated in years); local officer status (response options: member only (0) or officer (1)); local meeting size (response options: less than 25 members (1), somewhere between 25 and 50 members (2), somewhere between 50 and 100 members (3), over 100 members (4)); and local meeting attendance in the last 12 months (calculated as the percent of regular scheduled meetings attended either in-person or online).Footnote 5

The average length of membership was 11.60 years (the median length was 7 years; the range in years was 1 to 46). Twenty percent identified themselves as officers. The median meeting size was somewhere between 25 and 50 members. The average meeting attendance in the last 12 months was 42% (the median attendance was 47%; the range was 13% to 100%).

To estimate the representativeness of the sample with respect to the 2021 population of American union employees, the 2022 January issue of the U.S. Bureau of Labor Statistics (BLS) was consulted (Union Affiliation, 2022). In doing so, we compared the percentages in the sample for age group, gender, ethnic group, and employment status with reported national percentages. The results indicated that employees 25 years or older and women employees were oversampled by 3% or less, + 0.0015, + 0.0288, respectively. The results also indicated that ethnic employees and full-time employees were undersampled by 9% or less, -0.0641, -0.0921, respectively.

Psychological Safety at Meetings

To assess psychological safety at local meetings, we asked employees to respond to 7 items adapted from versions of the Psychological Safety Scale developed by Edmondson and her colleagues (Edmondson, 1999; Gavin et al., 2008; Nembhard & Edmondson, 2006; Tucker et al., 2007; see also Parker & Gruelle, 1999 for related items). The items focus on the experience of employees at meetings (in-person or online) in the last 12 months exempting non-psychological safety (see the Appendix for a list of items).

The items were prefaced with the statement:

“We are interested in a frank and accurate description of how your local meetings were run based on your experience—not hearsay from others—strictly and exclusively based on your experience.”

The statement was followed with a response instruction (“Check (√) one blank”) and an item stem:

In reference to local meetings I have attended in the last 12 months, the following describe what I experienced at these meetings . . .”

An example item is:

“. . . meetings wherein members were at no risk of embarrassing themselves even when they couldn't always express themselves clearly.”

Response options were “yes” or “no”.

A principal components analysis was performed on the psychological safety items. The analysis produced one eigenvalue greater than 1.00, eigenvalue = 4.748, percent of variance explained = 67.831, item loadings ≥ 0.559. The Cronbach’s α for the items was 0.92.

As a result of these analyses, responses were averaged, yielding continuous psychological safety scale scores from 0 (experienced less) to 1.00 (experienced more).

Non-Psychological Safety at Meetings

To assess non-psychological safety at local meetings, we asked employees to respond to 5 items adapted from union and nonunion design meeting scales, taxonomies, and commentaries (Hoffman, 2018; Lund & Taylor, 2010; Miller & Young, 1955; Parker & Gruelle, 1999; Rogelberg, 2019; Rose, 1952; Twarog, 2007). The items focus on the experience of employees at meetings (in-person or online) in the last 12 months exempting psychological safety (see the Appendix for a list of items).

The items were interspersed randomly with the psychological safety items, following the same preface statement, response instruction, and item stem.

An example item is:

... meetings wherein members did their best to begin and end meetings on time (as scheduled).

Response options were “yes” or “no”.

A principal components analysis was performed on the non-psychological safety items. The analysis produced one eigenvalue greater than 1.00, eigenvalue = 2.573, percent of variance explained = 52.454, item loadings ≥ 0.482. The Cronbach’s α for the items was 0.72.

As a result of these analyses, responses were averaged, yielding continuous non-psychological safety scale scores from 0 (experienced less) to 1.00 (experienced more).

Meeting Effectiveness

To assess local meeting effectiveness, we asked employees to respond to 2 items adapted from Hammer and Wazeter’s (1993) Global Scale of Local Union Effectiveness. The items focus on the experience of employees at meetings (in-person or online) in the last 12 months (see the Appendix for a list of items).

Prefaced with a response instruction (“Circle one number”), the items followed the psychological safety and non-psychological safety items.

An example item is:

“Based on how your local meetings were run, how effective do you think they were in doing the business your local needed to do?”

Responses were based on Likert scaling with 1 to 7 anchors (Not very effective to Very effective).

A principal components analysis was performed on the meeting effectiveness items. The analysis produced one eigenvalue greater than 1.00, eigenvalue = 1.918, percent of variance explained = 95.914, item loading = 0.979. The Cronbach’s α for the items was 0.96.

As a result of these analyses, responses were averaged, yielding continuous meeting effectiveness scale scores from 1 (lower effectiveness) to 7 (higher effectiveness).

Meeting Attendance in the Next 12 Months

To assess local meeting attendance in the next 12 months, we asked employees to respond to an item to indicate how likely they were to attend regular scheduled meetings (see Flood, 1993; Kahn & Tannenbaum, 1954; Lund & Taylor, 2010; McShane, 1986; for similar one-item measures in reference to 12 months).

The item followed the meeting effectiveness items:

“Based on how your local meetings were run, how likely are you to attend meetings in the next 12 months?”

Responses were based on Likert scaling with 1 to 7 anchors (Not very likely to Very likely), yielding continuous meeting attendance in the next 12 months item scores from 1 (less likely) to 7 (more likely).

Controls

To control for sample-specific statistical associations between the demographics and psychological safety at meetings, psychological safety scale scores were regressed onto the demographics, and in the same analysis, onto non-psychological safety scale scores. From this analysis, the unstandardized residual psychological safety scores were used to represent psychological safety (the predictor variable, x) in model tests.

In the same vein, significant zero-order correlations between the demographics and meeting effectiveness, and between meeting attendance in the next 12 months, were used to select covariates (see Sauer et al., 2013 for an overview of covariate selection). As such, meeting effectiveness scores were regressed onto gender, officer status, and meeting attendance in the last 12 months. In a separate analysis, meeting attendance in the next 12 months scores were regressed onto gender, membership years, officer status, and meeting attendance in the last 12 months. From these analyses, the unstandardized residual scores for meeting effectiveness (the mediator variable, m) and for meeting attendance in the next 12 months (the outcome variable, y) were used in model tests (see the tested model depicted in Fig. 2, with covariates listed in grayscale).

Fig. 2
figure 2

Tested model

Full size image

Results

Raw score zero-order correlations, means (Ms), and standard deviations (SDs) for all study variables are presented in Table 1.

Table 1 Zero-order correlations (rs), means, and standard deviations
Full size table

Descriptive Tests

To discern significant mean differences in demographics in relation to model variables, we performed t-tests, using median splits for non-dichotomous variables to construct demographic subgroups. On average, women employees (vs. men employees) were less likely to experience psychological safety at meetings, a difference also seen for employees in which English is a second language (vs. employees in which English is a first language), ts(130) ≤ -2.733, ps < 0.01. In contrast, on average, local officers (vs. member only) were more likely to experience psychological safety at meetings, a difference also seen for employees who attended a higher percent of meetings in the last 12 months (vs. employees who attended a lower percent), ts(130) ≥ 2.303, ps < 0.01.

Also, on average, women employees were less likely to indicate higher meeting effectiveness, t(130) = -2.473, p < 0.01. In contrast, on average, local officers and employees who attended a higher percent of meetings in the last 12 months were more likely to indicate higher meeting effectiveness, ts(130) ≥ 4.059, ps < 0.01.

And, on average, women employees were less likely to attend meetings in the next 12 months, t(130) ≥ -3.627, p < 0.01. In contrast, on average, local officers, employees with more membership years, and employees who attended a higher percent of meetings in the last 12 months were more likely to attend meetings in the next 12 months, ts(130) ≥ 2.319, ps < 0.01.Footnote 6

Preliminary Tests

The zero-order correlation between psychological safety at meetings and meeting attendance in the next 12 months was positive and significant (r = 0.49, p < 0.01), as were correlations between psychological safety and meeting effectiveness (r = 0.49, p < 0.01) and between meeting effectiveness and meeting attendance in the next 12 months (r = 0.72, p < 0.01), results that are consistent with the Hypothesis.

Also, as a baseline model check for mediation, residual scores for meeting attendance in the next 12 months were regressed onto residual scores for psychological safety at meetings.Footnote 7 The unstandardized coefficient was positive and significant, B = 1.330, standardized b = 0.315, standard error (SE) = 0.352, t = 3.778, p < 0.01, R2 = 0.099; F(1, 130) = 14.276, p < 0.01.

Model Tests

To test the hypothesized mediation, we ran a series of customized regression-based analyses derived from PROCESS 4.0 written for SPSS by Hayes (2022). In each regression model, we used 10,000 bootstrap (Boot) samples to generate Boot 95% confidence intervals (CIs) for direct and indirect effects.Footnote 8

Sequential Analysis

We first tested a model that estimated the direct paths to and from model variables in each instance, the two hypothesized direct effects (psychological safety → meeting effectiveness, meeting effectiveness → meeting attendance in the next 12 months), and the non-hypothesized direct effect (psychological safety → meeting attendance in the next 12 months). This all-inclusive analysis included the hypothesized indirect effect of psychological safety on meeting attendance in the next 12 months through meeting effectiveness (psychological safety → meeting effectiveness → meeting attendance in the next 12 months).

The results of this analysis are presented in Table 2. As shown, the direct effects were significant (i.e., the CIs did not include zero), with the exception of the path from psychological safety to meeting attendance in the next 12 months, direct effect = 0.4170, Boot SE = 0.2898, Boot CI [-0.1563, 0.9903]. This nonsignificant path indicates that when the path from psychological safety to meeting effectiveness and the path from meeting effectiveness to meeting attendance in the next 12 months are included in the model, the effect of psychological safety on meeting attendance in the next 12 months is nonsignificant. Also, as shown, the indirect effect of psychological safety on meeting attendance in the next 12 months through meeting effectiveness was significant, indirect effect = 0.9126, Boot SE = 0.2305, Boot CI [0.4866, 1.3906].

Table 2 Regression results: sequential analyses
Full size table

Based on the results of this analysis, we tested a model that estimated the hypothesized direct effects and the hypothesized indirect effect only (i.e., the psychological safety → meeting attendance in the next 12 months path was fixed to zero). The results of this fully-focused analysis (see Table 2) indicated that the direct effects were significant, as was the indirect effect of psychological safety on meeting attendance in the next 12 months through meeting effectiveness, indirect effect = 0.9600, Boot SE = 0.2392, Boot CI [0.5244, 1.4461].

Summary of Model Tests

The results of the preliminary analysis and the sequential analyses provide support for the hypothesized mediation. Consistent with the Hypothesis, employees who experienced more psychological safety at meetings are more likely to attend meetings in the next 12 months, but the path from psychological safety to meeting attendance unfolds through meeting effectiveness, such that more psychological safety is associated with higher indicated meeting effectiveness, which in turn is associated with more likely to attend.

Supplemental Analyses

To illustrate the distinctiveness of the hypothesized sequence, we ran a residual regression analysis with non-psychological safety scale scores regressed onto demographics plus psychological safety scale scores. Using these residual scores, we reran the all-inclusive analysis and the fully-focused analysis with non-psychological safety as the predictor variable (x). The results of these analyses indicated no significant direct or indirect effects involving non-psychological safety (i.e., the CIs included zero).Footnote 9

Discussion

Overall Summary

Our view that local union meetings provide employees an opportunity to experience safe environments in which they are encouraged to fulfill psychological needs through interactions with other employees can be suggested as linked to the problem of low meeting attendance. Our model results are unequivocal. As hypothesized, employees who experienced more psychological safety at meetings are more likely to attend meetings in the next 12 months, a relationship that unfolds through meeting effectiveness as rated by employees attending meetings. That the relationship unfolds through effectiveness should surprise no one, in that effectiveness as a mediator is implied in the literature on team effectiveness in nonunion environments. Also, in reference to the confirmed effect, it should be noted that the effect is independent of the experience of non-psychological safety at meetings, and that the link between non-psychological safety and likely meeting attendance as mediated by effectiveness is not seen in our data.

Also, that a mediated relationship is shown in relation to psychological safety at meetings but not in relation to non-psychological safety at meetings should surprise no one, least of all us. First, consider in reference to the link between effectiveness and attendance how work meetings in nonunion environments differ from local meetings. To wit, at work meetings, attendance is likely mandatory whether or not meetings are rated as effective by attendees. Not so with respect to local meetings; they are without exception nonmandatory. If local meetings are rated as ineffective by attending employees, no one would expect anything but low attendance. This mandatory versus nonmandatory basis of meeting attendance, spliced with the link between effectiveness and attendance, puts the role of meeting effectiveness in sharp relief. Second, consider the nature of meeting outcomes stemming from local meetings. A distinctive truism of local meetings is that meeting outcomes are group outcomes (e.g., approval of a wage adjustment); they apply equally to and benefit all eligible employees whether or not they attend meetings. This truism begs the question: Then, why attend? Our answer to this question is rooted in the experience of employees at local meetings in reference to how they interact. Above and beyond group benefits to be had, we think individual benefits are to be had—benefits that are associated with fulfillment of psychological needs. Our data are clear on this. To the extent that employees experience psychological safety at meetings conducive to such fulfillment—independent of non-psychological safety at meetings—more psychological safety is linked to higher rated effectiveness, and in turn, such rated effectiveness is linked to more likely to attend.

Literature Contributions

Centered on our interest in solving the problem of low local meeting attendance, our study contributes to the cited literature in several ways. Foremost, our importation and adaptation of constructs from the cited literature on team and meeting success featuring group dynamics represents a first attempt to bridge literatures in nonunion and union environments. As seen in our model, we adapted constructs from these literatures to predict likely meeting attendance. In doing so, we showed the expansiveness of the constructs for prediction of meeting success in union environments. In particular, psychological safety as a construct is given a boost by our work in regard to external validity. Without hesitation, we can confidently state that applications of adapted constructs featuring group dynamics are now open for prediction of meeting success in union environments.

The union participation literature also is a direct beneficiary of our crossover work. Having taken a decidedly psychological view, we think the participation literature has been mired in economic-inspired and attitudinal constructs with explanatory merit but bereft of insight drawn from attention to how employees interact at local meetings—insight that we think once shown in relation to prediction can be used as an intervention resource to address and solve the problem of low attendance. Moreover, the inclusion of psychological safety as a predictor of likely meeting attendance represents an important correction to the oft cited view of why employees attend (or do not attend). Put in simple terms in reference to our model results, we doubt that employees attend meetings solely for economic and attitudinal reasons grounded in rational self-interest. To be included are psychological reasons grounded in fulfillment of psychological needs. In our boldest (and we hope clearest) statement, we think that psychological safety entered into the prediction equation in regard to meeting attendance not only addresses an untapped need-based psychological connection between employees and unions but also provides an additive answer to the question of “Then, why attend.”

Study Limitations

Our study is not without caveats and limitations, all of which we think can be addressed in future studies. As a caveat, our study is intended as a demonstration of how covariation unfolds in regard to the sequence of variables indicated in our model. It is not intended as a demonstration of causality. Such a demonstration would, at minimum, require measurement of local meeting attendance at two points in time (e.g., the last 12 months and the next 12 months; see Mathieu et al., 2008 for example data). Also, as a first attempt to introduce psychological safety at local meetings as a predictor of likely meeting attendance, we made no attempt to rule out other predictors. Doing so would be wide of our study aim. But we do encourage researchers who have interest in making causal claims to collect data using time-lapsed measurement designs, with only one caveat. Because of the proprietary nature of local attendance data, we anticipate that these data are hard won (difficult to obtain), a reality that we think provides perspective on the value of our data.

Mediation can also be misinterpreted as part of a causal chain (see Hayes, 2022 for a thorough discussion). No such chain is implied by our model tests. We openly invite researchers to expand our model in future studies in reference to both predictors and mediators. Based on the idea that there is no “true predictor” or “true mediator” in regard to variation in local meeting attendance, we view psychological safety at meetings as joining a set of known predictors linked to participation such as instrumentality and union commitment. As for mediators, easily envisioned are multiple intervening variables such as emotional investment in the union movement and sense of civic duty applied to unions (see Rose, 1952; Tetrick et al., 2007 for discussions). Moderators of the shown mediation are especially welcomed in future studies. As seen in our data, women employees are less likely to experience psychological safety at meetings and are more likely to rate meeting effectiveness lower. Also evident in our data, employees in which English is a second language are less likely to experience psychological safety at meetings. We think both of these demographics represent important markers of conditional differences to be explored in future studies.

Also, in reference to our sample size, and based on our broad but limited sampling of any one union, the generality of our model results requires replication with larger samples and, under ideal conditions, data collection that represents an entire union. In this vein, a suggested side benefit of our results is their use as a means to justify a request for collaborative research directed at a union, in which researchers and local reps work together to collect anonymous in-house local attendance data in pursuit of a common goal: to forge a solution to the problem of low meeting attendance.

Suggested Intervention

As indicated, interventions featuring psychological safety or attention to interpersonal dynamics abound in nonunion environments, in which targeted are work teams with enhanced meeting success in mind. These interventions include standard assessment tools showing how assessment is delivered to attendees, how feedback is gathered from attendees, and how implementation is driven through involvement by attendees. Moreover, the nuts and bolts of these interventions are available in the public domain, features of which can be extracted from online documents or popular press books (see Hoffman, 2018; Rogelberg, 2019; Rozovsky, 2015; Understand team effectiveness, 2017 for prepared material).

To target local unions with enhanced meeting attendance in mind, herewith is a suggested intervention outlined for a union environment.

First, we suggest that the local reps meet to work out an announcement to be distributed before a meeting. The announcement should indicate an agenda item stated in bold print that “the local will conduct an assessment to discern how employees would like to see their meetings run, with attention to enhancing the experience of employees at meetings.” At this meeting, we suggest that the local reps preselect a local employee who is known and respected to introduce an invited guest (an outside researcher) who will administer an anonymous survey to be filled out voluntarily, a survey that is intended for “you to indicate how you would like to see your meetings run.” Using the foregoing statement beginning with “indicate” as the stem for the survey items, we suggest that all 12 items be used from our study (the 7 psychological safety items interspersed with the 5 non-psychological safety items), with yes–no response options (Likert scaling could also be used). Before the survey is distributed, it should be indicated and underlined that survey results will be reported at the next meeting using average item responses rather than individual responses, and that the interpretation of the results will be opened to the floor for discussion by those attending.

Next, we suggest an announcement be distributed by the local reps before the next meeting reminding employees that “a substantial part of the meeting agenda will be devoted to the results of the survey taken at the prior meeting about how employees would like to see their meetings run.” Importantly, the announcement should stress that all eligible employees are encouraged to attend the meeting whether or not they attended the last meeting and whether or not they volunteered to take the survey. At this meeting, “the researcher” should be absent, having reported the average responses to items (without descriptive tests) in a written document submitted to the local reps and distributed by the reps to employees before the meeting. This meeting should include only eligible employees—a closed-door meeting that also excludes non-local officials (perhaps to be briefed later). The key to this meeting is the creation of an open and informal atmosphere of discussion without concern that the conversation should be limited to one meeting. If a second meeting on survey results seems to be in order, it should be called. It is entirely possible that some locals during this phrase of intervention may take several meetings to “air out” and “settle issues” as related to how “we would like to see our meetings run.”

We envision variants of the intervention as outlined and we invite suggestions and refinements from all involved parties, including feedback on how meeting announcements are best constructed to generate interest and how surveys are best distributed to include the faintest voices. Also, as an addendum to the intervention, it should be noted that formal recommendations are not recommended. As indicated, psychological safety is not intended as an explicit team goal. Rather, by opening discussion about how all employees can expect to experience a safe environment at meetings, we expect that emergent are enriched climate properties of mutual respect for psychological needs and a renewed sense of trust that “ours is a collective effort to serve and fulfill the needs of all members.”

Final Note

On the relationships indicated in our model, we do not take the point of view that we have isolated the predictor or the sequence by which the experience of employees at local union meetings is linked to meeting attendance. We consider our model as a work-in-progress, in which a suggested path to attendance may yet be altered or reconfigured in reference to future modeling. In relation to our interest in strengthening the psychological lifeline between employees and unions, and in relation to our interest in developing interventions, we encourage researchers and local reps to view our model as a starting point to take in the psychological experience of employees at meetings and to consider associations between psychological safety and local attendance.