In 1995, the New Voice administration quickly set a new pace for organizing. “The AFL-CIO has proposed a fundamental, even radical, urgent institutional shift,” wrote Richard Bensinger, the new administration’s first organizing director. “Our goal is to organize millions of workers in the next decade.”

Sweeney started by laying out a “changing to organize” program at fourteen regional conferences that called for affiliate unions to work toward spending 30 percent of resources on organizing. Previously, only a handful spent more than even 10 percent. By 1997, the federation was spending up to $20 million a year on organizing, an eightfold increase compared to the Lane Kirkland years, and funded in part by the proceeds from a new credit card marketed to union members. It launched Union Summer, which put young people directly on campaigns and poured resources into the Organizing Institute (OI), which trained thousands of recent college graduates and rank-and-file activists to serve as an army of new organizers. The AFL-CIO Executive Council called for the organization of a million new members a year.

There were a number of high-profile campaigns and even some big wins. First was a nationwide campaign to organize 20,000 strawberry workers, forging new partnerships with progressive community groups. The organizing department played a lead role in the successful effort to unionize and win a contract for Avondale shipyard workers in New Orleans and at the Los Angeles airport. It supported organizing among Las Vegas construction workers, South Florida health care workers, hotel workers, and a successful effort among 10,000 US Air workers. When 74,000 home health care workers unionized with the Service Employees International Union (SEIU) in 1999, it was the largest union victory in decades, and it seemed to herald a new day.

Yet, the Sweeney efforts were never able to turn the ship around. The absolute number of union members dropped by a million from 1995 to 2009, and union density dropped 2.5 points. Most telling is that during Sweeney’s term, the labor movement brought fewer workers to National Labor Review Board (NLRB) elections than before he took office.

Unions brought an average of 252,000 workers to election each year in the Kirkland/Thomas Donahue years, but only an average of 168,000 workers a year in the Sweeney years. The percentage of the nonagricultural workforce voting in union elections dropped during Sweeney’s tenure to a mere 0.07 percent, compared to 0.2 percent in when he took office. The NLRB figures do not include all organizing efforts, such as public sector campaigns or victories under the Railway Labor Act. In these years, some unions increasingly began to turn to employer recognition campaigns that bypassed the broken NLRB system.

One study finds that as many as 40 percent of union campaigns employed this tactic at the turn of the twenty-first century, which could mean that as many as 280,000 workers were involved each year in organizing campaigns in the Sweeney years. Nevertheless, unions in the Sweeney years never matched the organizing activity of previous decades.

From the 1950s through the 1970s, an average of a half a million workers a year lined up each year to vote in labor board elections, and unions routinely brought over a full 1 percent of all the nation’s workers to election. Today organizing activity is lower than in the Sweeney years. Even with the largest upsurge in organizing in decades, such as at Starbucks and Volkswagen, only 107,000 workers voted in union elections in 2024.

The Sweeney administration quickly realized how difficult it was for a voluntary confederation of unions to move affiliate unions to organize. The AFL-CIO can serve as a bully pulpit, can use inspiration and shame to try to force change, but at the end of the day, organizing workers into unions with collective bargaining agreements is a process that rests with its member organizations.

Very few unions ever embraced the call to dedicate 30 percent of resources to organizing or even took seriously the executive council’s call to organize a million workers a year. Frustration with slow progress on organizing spurred the 2005 Change to Win (CTW) split in which five unions left the federation, calling for the AFL-CIO to force mergers and triple the amount it spent on organizing.

Within a decade, however, the CTW coalition had fallen apart, and also never achieved the turnaround in union organizing and membership growth that it sought to achieve. In fact, one analysis found zero statistical difference in the CTW unions’ organizing success after the split. SEIU recently announced its reaffiliation with the AFL-CIO, finally closing this chapter of fruitless disunity in labor’s ranks.

Despite their best efforts, Sweeney’s critics were no more successful than he had been. In retrospect, it is clear that both the New Voice coalition and CTW were trying to stand on shifting sand. The raft of neoliberal free trade agreements in the 1990s and early 2000s decimated manufacturing, the traditional stronghold for labor. During Sweeney’s tenure, the United States lost a third of its manufacturing base, or about six million jobs.

The North American Free Trade Agreement (NAFTA), China’s entry into the World Trade Organization in 2001, the Free Trade Agreement of the Americas (FTAA), and more took an enormous toll as union members lost their jobs, and workers who did not yet have unions became more scared to organize in this new climate. Meanwhile, the gig economy meant that many workers held precarious jobs, outside the reach of collective bargaining, or labored in “fissured workplaces” where they did not have clearly defined employers to negotiate with. Fewer and fewer workers were thus even eligible for union membership.

Meanwhile, the doorway through which workers could enter unions remained incredibly narrow. Starting in the 1970s, America’s employers ramped up resistance to workers’ union organizing efforts and began to bend and break labor law at new levels, and the law was too weak to stop them. By 2009, when faced with union organizing efforts, nearly 90 percent of employers forced workers to attend one-sided meetings against the union, more than half threatened to shut down, and a third illegally fired activists. Even when workers won a union election, only a little more than half ever got a first contract.

Employers so effectively circumvented the nation’s toothless labor law that it had become a nearly Sisyphean task for workers to organize using the National Labor Relations Act (NLRA). Even if unions had poured in greater resources and brought more workers to election, there is little evidence that this would have turned the tide on union membership, given the structural obstacles to winning union elections and getting contracts within the NLRA system.

The last big surges in private sector union membership were in the 1930s, when the founding of the CIO unions and a then functional NLRA meant that millions of industrial workers had new access to membership, and during World War II, a brief historical period when the federal government encouraged union membership and actively discouraged employer resistance. Absent such new organizational forms, a functioning labor law, and strong government support, would more resources have ever done the job? There is no real historical or even global precedent for such a turnaround.

The Sweeney administration tried to change the landscape. It waged a full-fledged battle to win labor law reform in the Employee Free Choice Act (EFCA), legislation that would have increased employer penalties for labor law violations and allowed workers to choose a union by signing cards rather than through the fraught NLRB election process. Yet, labor never could build a filibuster-proof majority in the Senate, even with President Barack Obama in the White House. The broken NLRB status quo remained in place.

Some argue that the biggest failure of the Sweeney administration around organizing was less structural than cultural. Sweeney pledged to revive a “culture of organizing,” urging unions to look outward, not inward. Yet, much of the program’s focus was on reallocating resources within a staff-heavy organizing model that never emphasized rank-and-file leadership and agency. Even “whole-worker” organizing, however, faced the often-insurmountable obstacles of organizing within the United States’s broken labor law system and had to operate within a neoliberal economy stacked against workers.

Perhaps, then, the Sweeney administration’s greatest miss was that it never bypassed the broken system and opened up new paths to build worker power within capitalism’s latest transmutations. It is imaginable, after all, to have a union movement with low collective bargaining density, but higher worker engagement. A movement can also raise wages and win expansive benefits, outside the traditional collective bargaining paradigm.

Bargaining for the Common Good, for example, reimagines bargaining to include both workplace and community issues. At a time when many workers may be reluctant to organize because they remain unsure whether a union can really do anything for them, this community-based model has helped teachers, janitors, retail workers, and more gain more leverage over the issues most relevant to them and their communities, like housing and education.

Another model that bypasses the broken system is the Fight for $15, SEIU’s successful campaign that helped millions of workers win higher wages, inspired dozens of states to raise their minimum wage, and helped put as much as $150 billion in workers’ pockets since 2012. Sweeney campaigned on the slogan that America needs a raise, but the person who delivered that raise was Mary Kay Henry, former SEIU president and Sweeney mentee. Most importantly, workers won this victory outside the traditional paths to growth that were the locus of most of the AFL-CIO’s organizing  efforts. This suggests that labor could not actually get a raise for America using the current structures but instead needed to build new ones.

Sweeney called for such new models in the beginning of his tenure. “As unions expand into new sectors of the workforce — from low-wage, largely immigrant workers to high-tech professionals — they’ll find new ways to build organizations that meet their needs,” he predicted in his book, America Needs a Raise. “For those of us already in the labor movement, our obligation is to help them organize themselves — and also to offer some models of organization which they can choose among and, no doubt, improve upon.”

The AFL-CIO’s most ambitious effort to explore new organizational structures is Working America, an associate membership organization founded by the Sweeney administration in 2003 that anyone can join for free. Its largest impact has been political. Working America relies heavily on thousands of canvassers who have doorstep conversations with voters in key battleground states, with a particular focus on the white working class. By 2008, Working America could claim one in ten Minnesota voters as members, for example, and also worked on state- and city-level minimum-wage campaigns.

On balance, however, Working America has developed as mostly a political mobilization group. And it is just one lonely piece in what ought to be a complex jigsaw puzzle of new labor structures within the federation’s orbit. The union movement knew as early as the 1980s that it needed to develop associational membership models, which would allow people not covered by a collective bargaining agreement to join unions.

Most unions balked at the idea of pouring resources into working people who would not end up paying full union dues, though a few unions like Communication Workers of America, American Federation of Teachers, and Actor’s Equity have experimented with associate or open membership models. Sweeney, as well as his successors at the AFL-CIO, have shied away from challenging unions to develop such alternative models, preferring instead to put the focus and resources on Working America.

Perhaps, the AFL-CIO’s reversal on immigration policy in 2000 was one of the most important contributions of the Sweeney administration to a new kind of organizing, even though it was outside the purview of the organizing department. For decades, the official policy of the labor movement on immigration was some version of what one Houston-area building and construction trades leader asserted in the early 1980s: “If they’re illegal, then they shouldn’t be in our union, and we shouldn’t be bothering with them.” Under Sweeney, the AFL-CIO Executive Council made an abrupt turn on immigration, calling instead for a legal path to citizenship for millions of undocumented workers and repeal of the law that criminalized hiring them. The impact was enormous.

Even though this immigration reversal did not usher in millions of workers into collective bargaining agreements, it set a new tone and expectation around how the US labor movement was oriented toward millions of immigrants in America’s workplaces and put the weight of the federation behind a national policy shift. This new understanding helped give much-needed space and momentum for the burgeoning worker center movement that developed in the early 2000s and is now strung together through national networks like the National Day Laborer Organizing Network (NDLON), National Domestic Workers Alliance (NDWA), Restaurant Opportunities Center (ROC), and more.

Immigrant workers make up a significant proportion of these groups’ memberships. Though some unions initially regarded these groups as competitors, partnerships have grown as unions have banded together with worker centers to organize workers who were previously outside unions’ radars. The AFL-CIO has brought the National Taxi Workers Alliance (NTWA) onto its executive council and supported a special fund that experiments with new forms of membership representation and organizing.

Any measure of union membership, and NLRB statistics, show that the Sweeney administration never came close to meeting its ambitious union organizing goals. Union membership, union density, and the numbers of workers voting in NLRB elections all fell on his watch. But maybe we should measure the New Voice administration’s impact beyond statistics that chart growth within a narrow and broken labor law system. After all, the percentage of workers covered by a collective bargaining agreement is a poor gauge of success when so few workers ever have a real chance to form a union.

Ultimately, an assessment of organizing leads to the larger, overarching goal of the union movement. Is the purpose of organizing to increase union membership, or is it to rebuild a working-class movement that can effectively make life better for the vast majority of working people? “My goal is to leave a legacy similar to the advantage my own generation enjoyed,” wrote Sweeney. “A social movement and a social contract that will once again lift Americans and bring us together.”

Though the Sweeney administration legacy is mixed, its humane immigration policy now seems more monumental than ever and helped give breathing room for new kinds of worker organizations at a crucial moment. Under Sweeney, the federation also helped to set the stage for the Affordable Care Act by punching above its weight in the election of President Obama.

But the AFL-CIO could not shed the shackles of the broken labor law system or forge a new social compact that centers on working people’s prosperity. America’s workers and their unions still have this task ahead of them, and their best path to success, especially in the next four years, is likely outside the Sisyphean task of organizing within the broken US labor law regime.