NLRB Special Project Posts:
All NLRB postings germane to this EPRN special project will be listed here are they continue to be posted to the EPRN Website. Check back here for updates.
Background Information:
These blogs will serve as helpful background information for drafting your comment.
EPRN Articles Relevant to NLRB Rulemaking:
Deadline to submit comments for this EPRN Special Project: July 22, 2011
Message from Tom Kochan, EPRN Principal Researcher, sent July 3, 2011:
Dear Colleagues,
As you may know the National Labor Relations Board recently issued proposed rule changes for representation election processes and is now in the open comment period. Ellen Dannin, our Employment Policy Research Network Labor and Employment Law Topic Coordinator has written a very good introductory blog on the proposed rules and has started the comment process.
Please read the above postings on NLRB rule making, read the Board's summary of the proposed changes that can be found at http://www.nlrb.gov/node/525 and contribute your analysis either by entering a comment on any of the postings listed above.
Ellen and I will summarize the comments received and forward them to the NLRB during the public commentary period so that our collective expertise is reflected in the process.
It is really important for us to show we can respond to policy/rule proposals in a timely manner (public hearings are scheduled for July 18-19; we need your input well before then) so I hope you will contribute your views and indicate the research base from which your analysis is drawn.
Thanks and have a great 4th of July weekend.
Tom
Thomas A. Kochan
George M. Bunker Professor of Management
Co-Director, Institute of Work and Employment Research
MIT Sloan School of Management
NLRB's Proposed Election Rule Changes
The NLRB Proposed Changes to Union Certification Procedures: Lessons from the Canadian Experience
Chris Riddell
Associate Professor
ILR School, Cornell University
1. Background
Canada has had a variety of experiences with changes in the union certification process, in particular changes in the statutory requirement on what period of time may elapse from the application date to the date when the election takes place – often referred to as “election delay”. The latter will be the focus of my comment. In particular, this comment addresses several components of the NLRB proposal to streamline the certification process in the United States that are specifically geared towards reducing election delay.
It is important to recognize up front that the Canadian union organizing environment may not be an ideal ‘laboratory’ for drawing inferences about how American union organizing would change if the union recognition process were to change. Nevertheless, there are many similarities in labor market institutions and characteristics between the two countries as well as many similarities in the unions (many of which are North American unions) that operate in both countries including the organizing strategies they adopt, and contextual (including employer-related) environment they face (for instance, see Martinello and Yates 2005).
Canada, historically, has used a card-check system where elections are only required if less than a certain fraction – the minimum threshold – of the proposed bargaining unit sign cards. Generally, in Canadian card-check regimes, only around 5-10% of applications require an election. In these card-check regimes, there have, in most cases, been no statutory requirements (or Board policy) for time limits on elections.
The Canadian union certification process has changed dramatically however. Currently, a strong majority of Canadians work in a province where elections are mandatory for determining union certification. In particular, Ontario, British Columbia (henceforth “B.C.”), Alberta, Saskatchewan, Nova Scotia, and Newfoundland all use compulsory elections. All of these provinces except Alberta have a statutory requirement on when the election will take place, which ranges from 5 to 10 business days. Alberta’s law only indicates that the election “will take place as soon as possible”.
Compliance with these statutory requirements is not 100%, however, and has varied across regimes. For instance, in B.C. under the 1987 Industrial Relations Act compliance with the 10 business day election timing statutory requirement was 97%, but in Ontario under the 1995 LR Act compliance was less than 75% (Campolieti, Riddell and Slinn 2007). As well, there have been cases where two provinces have had the same general union recognition – i.e., mandatory votes vs. card-check – but where the election date timing policy differed (see Campolieti, Riddell and Slinn 2007).
In summary, there are two major sources of variation in election timing parameters that researchers have examined: a) comparing the same union recognition regime (cards vs. elections) across provinces where one province had a statutory requirement on when the election would take place and where the other province had no statutory requirement; and b) comparing the same union recognition regime (cards vs. elections) across provinces where both provinces had a statutory requirement on when the election would take place, but where compliance rates differed substantially.
2. What have we learned about the role of election delay in the union organizing process in Canada?
To begin, what do we know about the effects of election delay on the outcome of the election (i.e., the likelihood of employees successfully certifying)? Table 1 presents a summary of the discussion below. Except where noted, the research discussed here is based on Campolieti, Riddell and Slinn (2007) – henceforth “CRS”.
One piece of evidence comes from comparing the card-check regime in Ontario during the early-mid 1990s with the card-check regime in B.C. over the same period. In the later comparison, we have the identical union recognition process (card-check with election required only if the minimum threshold was not met), and same economic conditions -- except where in Ontario there was no statutory requirement on when the election would take place, but where there was such a statutory requirement in B.C. (at 10 business days).
If we compare these two regimes the actual election delay statistics are striking: the average number of days from application to election was around 10 business days in B.C. but over 50 business days in Ontario. I also note that as these were all elections where the card support was less than 55% so they were more likely to be bids contested by the employer; thus, these are cases that are probably more similar to a U.S. environment than many Canadian union certification contests. The comparative analysis reveals that in B.C. – statutory requirement, minimal election delay – there was no correlation between election delay and the likelihood of employees successfully certifying, but in Ontario – no election timing rule, high election delay – there was a statistically significant and quantitatively large effect of election delay reducing the chance of certifying.
Other evidence points towards employer actions as likely underlying the effects of election delay on the likelihood of employees successfully certifying. Specifically, CSR compared the effects of ULP charges against the employer across these election timing regimes. The results of the comparative analysis reveal that ULPs reduce the likelihood of certifying by a greater magnitude in the high delay/no statutory requirement Ontario regime relative to the low delay/statutory requirement B.C. regime.
The second piece of evidence comes from comparing jurisdictions that have quick-vote procedures as a statutory requirement but with differing compliance rates. In particular, CSR compared elections from the compulsory election regime in Ontario during the late 1990s with B.C.’s mandatory election regime of the mid-late 1980s. The latter B.C. regime, noted above in section 1, was where compliance with the 10 day election rule was virtually 100%. All of the results follow the identical path to those discussed above. In the high compliance/low election delay regime of B.C. there was no statistically significant effect of election delay on the probability of employees certifying, but there was a negative effect in the relatively lower compliance regime of Ontario. Moreover, just as above, ULPs reduced the chances of certifying by a much greater amount in the low compliance Ontario regime.
Thus, to summarize, the evidence from a variety of labor law regime comparisons in Canada demonstrates that when the date of the election is allowed to be delayed beyond two weeks, we start to see election timing reduce the likelihood of employees certifying the union they have signed cards to support. Empirically, this effect can be measured both from election delay itself as well as from ULP charges. Moreover, the recent analysis in Riddell (2010) strongly suggests that the findings of CSR are not spurious correlations; in fact, the causal effect of election delay on certification outcomes appears to be even greater. The question then is why would something that can be best described as a procedural matter influence the actual outcome of the election (especially since cards have already been signed in support)? The result that ULP charges against the employer reduce certification success by substantially more when elections are allowed to be delayed certainly implies that employer actions are at the root of these findings.
Finally, essentially the same result has been documented if we compare card-check regimes to mandatory vote regimes (see Riddell 2004 for this analysis). A mandatory vote system in Canada is virtually the same as the card-check system except that after the card-check has been verified to meet the minimum threshold, an election is held a certain number of days later. In essence, the mandatory vote system adds an extra step to the overall union certification process thereby delaying the stage at which the employee’s (final) decision on unionization is made. It has been empirically established that ULP charges against the employer reduce the probability of employees successfully certifying by twice the magnitude in a mandatory votes system relative to a card-check system.
3. What factors influence election delay?
But why are elections delayed in Canada in the first place and what does this tell us – including what influences non-compliance with statutory requirements on election date timing?
When CSR examined regimes with no statutory requirement and regimes with a statutory requirement but relatively low compliance they found that ULP charges and employer-filed submissions objecting to the application were by far the two key correlates of elections being delayed/non-compliance with a statutory requirement. On the other hand, when CSR examined the highly efficient B.C. regime of the 1980s where compliance was virtually 100% they found that no factor caused elections to be delayed. This of course is hardly surprising since when compliance is virtually 100%, there is no variation in election delay and thus nothing can correlate with it.
When the actual types of objections were examined, the findings were that the strong majority of submissions were challenges to the appropriateness of the bargaining unit or challenges to the voter’s list. This was a surprising finding because most Canadian jurisdictions (including Ontario where most of the election delays were occurring) allow the vote to be held in a timely manner, double-seal the ballots in question and hear evidence at a later date. Clearly, a statutory requirement on election timing greatly reduces such delays, but one needs high compliance with the statutory requirement as well to prevent what overall likely appears to be employer abuse.
4. Key Conclusions
The Canadian experience may be useful in informing the debate on streamlining the NLRB election process. What is the bottom line lesson from the variation in union certification procedures over time and across provinces in Canada?
There does appear to be a direct, causal link between the amount of time that passes from the application date to the election delay and the outcome of the election – and in particular people changing their minds from the card-check stage. Specifically, the longer the delay in when the election is held the lower the probability of employees successfully certifying. Further, employer activity is empirically linked to why elections are delayed (even with a statutory requirement in some cases). However, there is one regime where election delay no longer had a link to the outcome of the election: the case of the 1980s B.C. regime where compliance with the statutory requirement of 10 business days was 97%. Moreover, in this latter regime, there was no evidence that employer-related activities could influence when the election was held; the Board simply did an outstanding job at holding elections at or before the date mandated and addressing objections to the application at a later date. Finally, not only was election delay not linked to the outcome of the election in this regime, but employer ULPs reduced the chances of certification by a much smaller amount. Overall, the implication seems quite clear: when elections take place very quickly after application the ability of employers to launch campaigns against the organizing drive is substantially mitigated.
Indeed, after a careful review of the Canadian literature where a substantial number of different labor law regimes have now been analyzed begs the question: why would a policy-maker not want to adopt a quick-vote system given substantial evidence that election timing is manipulated by the employer in a way that is not in the spirit of Canadian labor law (and the Wagner Act)? The only explanation I can imagine is that employers should have the right to make their case against unionization (if they choose to do so). But, is two business weeks not a sufficient amount of time to present those arguments?
It is important to note that this highly efficient British Columbia regime I am emphasizing here did not have particularly high union success rates from a (Canadian) historical standpoint; in fact, they were much lower than success rates in the British Columbia card-check years – both before and after (see Riddell 2004) . It is therefore incorrect to think that a quick-vote system with virtual 100% compliance is somehow pro-labor or somehow equivalent to card-check, or (from a U.S. comparative standpoint) is basically the Employee Free Choice Act disguised.
References
Bentham, Karen. 2002. “Employer resistance to union certification: A study of eight Canadian jurisdictions.” Relations Industrielles, Volume 57, Number 1, pages 159-185.
Campolieti Michele, Chris Riddell and Sara Slinn. 2007. “Labor Law Reform and the Role of Delay in Union Organizing.” Industrial and Labor Relations Review, Volume 61, Number 1, pages 32-58.
Martinello, Felice and Charlotte Yates. 2005. “Union and employer tactics in Ontario certification applications.” Advances in Industrial and Labor Relations, Volume 13, pages 157-190.
Riddell, Chris. 2001. “Union suppression and certification success.” Canadian Journal of Economics, Volume 34, Number 2, pages 396-410.
Riddell, Chris. 2004. “Union certification success under voting versus card-check procedures: Evidence from British Columbia, 1978-1998.” Industrial and Labor Relations Review, Volume 57, Number 4, pages 493-517.
Riddell, Chris. 2009. “Quick-votes and union certification drives in Canada.” Proceedings of the 61st Annual Meeting of the Labor and Employment Relations Association, University of Illinois, pages 44-47.
Riddell, Chris. 2010. “The Causal Effect of Election Delay on Union Win Rates: Instrumental Variables Estimates from Two Natural Experiments” Industrial Relations, Volume 49, Number 3, pages 371-386.
Thomason, Terry and Silvana Pozzebon. 1998. “Managerial opposition to union certification in Quebec and Ontario.” Relations Industrielles, Volume 53, Number 4, pages 750-769.
Table 1
Summary of Canadian research findings on role of election timing
Elections held in card-check regimes Elections held in compulsory election regimes
B.C.
Statutory quick-vote/high compliance Ontario
No election timing specified B.C.
Statutory quick-vote/high compliance Ontario
Statutory quick-vote/low compliance
Election delay not correlated with election outcome
Election delay found to reduce the likelihood of employees successfully certifying
Election delay not correlated with election outcome
Election delay found to reduce the likelihood of employees successfully certifying
ULPs associated with about a 26% decline in odds of employees voting in favor of certification
ULPs associated with about a 46% decline in odds of employees voting in favor of certification
ULPs associated with about a 33% decline in odds of employees voting in favor of certification
ULPs associated with about a 47% decline in odds of employees voting in favor of certification
ULPs not correlated with election delay
ULPs associated with longer election delays
ULPs not correlated with election delay
ULPs associated with longer election delays
NOTES: The specific years and legislation in effect referred to above is as follows: column 1 – B.C. 1993 to 1998, Labour Relations Code (1993); column 2 – Ontario 1993 to 1995, Collective Bargaining and Employment Act (1992); column 3 – B.C. 1987 to 1992, Industrial Relations Act (1987); column 4 – Ontario 1995 to 1998, Labour Relations Act (1995).
NLRB's Proposed Election Rule Changes
In Support of the NLRB’s Proposed Election Rule Changes
Statement before the NLRB’s Open Meeting on Proposed Election Process Rules
Paul F. Clark, Professor and Head Department of Labor Studies and Employment Relations
Penn State University
July 19, 2011
As a university faculty member I have observed, studied, and taught about the American system of employment relations for many years. For the majority of American employees, the legal framework for that system is spelled out in the National Labor Relations Act.
Each time I introduce a new set of students to the Act I begin by having them read Section 1. This section provides the rationale for the Act’s passage. Central to that rationale is the concern that “The inequality of bargaining power between employees who do not possess full freedom of association or actual liberty of contract and employers who are organized in the corporate or other forms of ownership association substantially burdens and affects the flow of commerce, and tends to aggravate recurrent business depressions, by depressing wage rates and the purchasing power of wage earners .”
It seems clear that in writing this legislation Congress recognized that when employers held all of the power in the employer-employee relationship (as they had for the previous 100 years of U.S. history), when they made all decisions unilaterally, not only did individual employees suffer, but so did society at large. The danger of concentrating power in any one institution is something that the architects of our political system clearly recognized, and it is the basis of the system of checks and balances that have been one of the foundations of American democracy.
The architects of our system of employment relations recognized this danger as well. The opportunity to organize a union and bargain collectively that the National Labor Relations Act extended to American workers represented a check on the absolute power of employers in the workplace, and it served as a mechanism for balancing the interests of employers and employees.
Senate Majority Leader Harry Reid recently referred to the principle of checks and balances in a statement of support for the changes the Board majority has proposed. Let me just state here that I believe I am making a slightly different point than the Speaker made. My point is that the right to organize and bargaining collectively IS itself a check on unilateral employer power in the workplace. If employees believe that an employer is exercising the power they have responsibly by employing good human resources practices, providing reasonable pay and benefits, and using their right to employ at will judiciously, they will likely forgo their right to organize a union. However, if an employer does not exercise its power in a responsible way by not employing good HR practices, not paying reasonable pay and benefits, or abusing its right to employ at will, its workers have a legally protected way to do something about it — they can organize a union and try to impact the employer’s practices for the better.
One of the aphorisms about employment relations that I first heard in graduate school, and have heard many times since, is that “if an employer gets a union, it deserves one.” The idea being that employees, in almost every case, organize a union because, in their view, the employer has not lived up to its responsibility. I think this was actually a pretty insightful aphorism that applied for the first 40 years or so of the Act’s existence. For much of that period unionism grew to the point that up to one-third of eligible workers exercised the right to organize and bargain extended by the Act. And for the two-thirds of employers without a union, the threat that their workers might follow suit provided a great incentive to provide good pay and benefits and otherwise engage in good HR practices.
Regrettably, the thoughtful system of employment relations that the Act created, and that served this nation well for several decades, no longer functions as intended to the detriment of our employment relations system.
The evidence suggests that this is because the check and balance the Act offered to employees — the opportunity to form a union and engage in collective bargaining — is now unattainable for many American workers. It is unattainable because the process for exercising that right has become a minefield and a marathon and many employees who might want to explore the option of organizing a union simply choose not to because the price is too high. Various studies and polls over the last decade have suggested that up to 60 million non-union workers would like to organize, but most will not try. The minefield they face consists of the many sophisticated elements of the modern anti-union campaign skillfully designed by attorneys, psychologists, and communications specialists. All too often unfair labor practices, ranging from threats and intimidation to the firing of union activities, are a part of these campaigns. And the marathon aspect of the process, of course, is caused by the endless delays that have become part and parcel of the process, a phenomenon identified in a number of studies, including a recent one at the University of California.
The fact that the employment relations system created by the Act does not function as intended serves the interests of employers; but it does not serve the interests of employees or of our larger society. I believe the changes proposed by the Board majority are a small, but important, first step to restoring the opportunity for employees to choose union representation and collective bargaining. I believe the proposals designed to reduce unnecessary delays in the election process are particularly important. They would help restore a workers right to choose union representation, without interfering with the employer’s right to communicate with their employees. I urge the Board to adopt the proposed changes.
The Not So Partisan Board
Professor Block:
I agree that the Board's seeming flip flopping reduces its legitimacy with the two communities and the public. But that partisan pendulum is a bit of an illusion.
In reality, in most cases, the Board's casehandling isn't politicized, ideological, or partisan. In the vast majority of cases - lower salience run of the mill cases - the handling of cases is objective and fair. As a Liebman attorney, we could depend on fair, objective, largely apolitical handling of cases, particularly by Meisberg, Acosta, and Kirsanow. Schaumber and Battista were more ideological, but in most cases, that wasn't an issue.
But the public doesn't hear about this. The Board makes news when it hands down high-salience controversial decisions - Brown University, Oakwood, etc. When we had these kind of cases at the Board, we knew that the two sides would follow their marching orders. Liebman and Walsh would do what Kennedy, AFL-CIO and Change-To-Win wanted. The Republicans would do what the Bush Administration, NWTWC, Congressional Republicans want. These are the cases, that the public and the communities hear about.
Why is the Board fair and objective and relatively apolitical in low salience cases?
First, Board Members build trust and reciprocity in these cases. The quintessence of this are the superpanels that the Board uses in late summer and early fall as the fiscal year end approaches. With Congressional oversight looming the Board tries to move cases fast. Rather than the standard process - review by staff counsels, bench memo, presubs by the staff counsel and the front offices, subpanel with votes and rationales submitted by the staff attorneys, decisions and dissents - the Board moves cases FAST.
At superpanel, they get in the board room, the staff counsel presents the question, the Board Members get on or off the case allowing a majority (like Schaumber might opt out and allow Kirsanow to replace him, to create a 3-0), and the Board votes to adopt short form (short boilerplate decision, no dissent, can get out in about a week). The normal case process can take 6 months to a year or years. These superpanels move cases in days! But no one in the public and many practitioners don't know about this cooperative process.
Importantly, the Board Member's staff attorneys are a stabilizing, non-partisan force. Whether Republican or Democratic, they are specialists, they work together on cases for years, they tend to become less ideological as the spend more years on the Board, and they view their task to be professional handling of cases in an objective way. In my time at the Board, I knew and worked with about 100 Member staff attorneys and only a handful were politicized or ideological, whether Democratic or Republican.
One of the sources of convergence of staff attorneys who come from a union-side background and those who come from management is the educative effect of the union participation of the Republican, management types. Board staff attorneys are members of the NLRBPA, and involvement in the NLRBPA changes many of the GOP counsels over time. Attorneys who opposed unions learn the value and efficacy of unions in the advancement of their self interest. For example, a very ideological attorney on the Schaumber staff did great service to the NLRBPA as a grievance chair and an attorney who started out very political on the Battista staff became an invaluable asset to the PA over time. Such cases are myriad.
Just as union matters bring Board staff attorneys closer, personally and ideologically, the internal labor relations of the Board bring Members closer. Sometimes in odd says. Her public pro-union image notwithstanding, Member Liebman was more vociferously anti-union in her role as Board manager than anti-union Schaumber. When Schaumber heaps high praise on Liebman (in places like his recent House statement on the Board's new rules), some of the mutual respect is a product of their shared project of busting and harassing the Board's unions in their management role. They were joined in that project by Solicitor William Cowen (a former GOP Board Member). The public just doesn't see the mutuality of interest that Wilma and Peter and Bill had as anti-union managers in the Board's internal labor relations.
Internal labor relations makes strange bedfellows in the other directions. Kirsanow, despite the public notion that he was an ideologue, helped the PA delay the implementation of an unfair evaluation system by telling the other Board Members that he believed the system would fare badly at arbitration. Walsh, a former President of the NLRBPA, was the only Member other Kirsanow, that was reputed to make the NLRBPA's case now and then. The public doesn't know that now and then Board Members like Kirsanow and Walsh tried to put the brakes on the hard nosed anti-union internal labor relations stance of Liebman and Schaumber.
Lafe Solomon, then head of the R-unit, and Les Heltzer, the Executive Secretary were also the voices of reason in internal labor relations. Lafe played an important role in creating a best practices memo that helped fix the unfairness and discrimination in the unfair evaluation system. Imagine, Kirsanow and Lafe aiding the cause of the NLRBPA union by being the voices of fairness against the hard-core stance of Liebman and Schaumber!
What does the public see? Recently Kirsanow is running around crying chicken little, repeating the party line, like Schaumber and Hayes, that the election rules changes will "eviscerate" employer Section 8(c) rights. On the other side? Lafe Solomon of Boeing infamy and Wilma Liebman. Looks very partisan. But looks are deceiving.
So while I agree with the basic premise that politicized high salience Board cases create an image of a flipflopping Board, there is another reality to the Board. The public and many practitioners see the tip of the iceberg, but the low-salience cases that no one sees, or the cases pushed through rapidly, cooperative, and in a bipartisan manner at superpanels, are a different Board. Board attorneys add predictability and objectivity, and the Board's internal labor relations makes for "cross cutting cleavages" that bridge ideology and party.
Tripartite committees? Great idea in theory. They are used in many other countries and were used in this country during WW1 and 2 in this country. But the fact that they are now used in social democracies would likely kill them here. The right will red bait tripartite committees the way they do universal health care.
But it is great to hear someone suggesting something creative and new. The unidimensionality of our labor-law debates is itself an oppressor.
NLRB Representation Proceedings Rules Changes - General Observat
These changes in representation case processing procedures seem reasonable. Most importantly, they are within the Board’s discretion to administer the representation process under the Act.
The intensity of the debate over these rules supports my view that the major issue with the NLRA and NLRB at this time is not the election rules, a General Counsel decision to file a charge (such as Boeing), or a decision on any particular doctrine (voluntary recognition such as Dana). The real issue facing the NLRB is the perceptions of partisanship and distrust that are visited on the Board. The employer community does not have confidence in the Board when it is dominated by Democratic appointees and the union community does not have confidence in the Board when it is dominated by Republican appointees. Thus, these proposed rule changes, which, in my view, should be viewed as a routine administrative matter, resulted in an editorial in the Los Angeles Times (http://www.latimes.com/news/opinion/opinionla/la-ed-nlrb-20110711,0,3040...). The Boeing charge resulted in hearings before a committee of Congress. Reading the dissenting opinion on the rules proposal, it is not unreasonable to believe that these rules will be rescinded at such time as as Board is dominated by a Republican majority.
This lack of trust, perhaps exacerbated, if not caused by, the hyperpartisanship in Washington, is destroying the credibility of the Board as an institution. This is no way to administer an industrial relations system in a modern, globalized economy where it is necessary that firms be competitive and employees be provided with the right to bargain collectively. I do not see how we can continue with a situation in which, at any point in time, one of the two major actors to the collective bargaining system believes that the Board is hostile to its interests. There is no reason to believe these problems will not continue as long as a majority of Board is composed of members appointed by a President associated with one party.
A reasonable solution is a legislative restructuring to create a tripartite Board consisting of labor representatives, management representatives, and established neutrals with expertise in collective bargaining and labor law. Employers and unions will know that their views and position will always receive a fair hearing. Over the long run, a tripartite Board will likely result in stability in Board doctrines and increased acceptability of Board decisions.