I relocated to Harlem in the summer of 2011, and upon moving in promptly called the cable company. It turned out the cable installer had grown up in a public-housing complex just a few blocks north of my apartment. He’d completed an apprenticeship program several years earlier, and as we chatted it was clear he was acutely aware of how lucky he was. He had a steady income and decent pay—a rarity among his friends—and he owed it to a cousin who had stumbled upon the training program. A happy coincidence, but one that highlights a major weakness in today’s middle-skill training efforts.

Broadly defined, “middle-skill” jobs are those that require more than a high-school diploma but less than a bachelor’s degree—including occupations as diverse as electricians, police officers, radiologic technicians, paralegals, and cable installers.

Although middle-skill jobs as a share of overall employment have shrunk, they still make up 48 percent of all employment in the country today—nearly 70 million jobs. Further, wage gains in these occupations often outpace national averages—with annual salaries in some categories topping $70,000. Finally, many of the middle-skill industries—welding, speech therapy, plumbing, and a whole host of others—are not easily outsourced, offering job security in addition to income gains. Simply put, these are good jobs.

So while we have to generate an estimated 21 million jobs by 2020 to return to full employment, the United States also needs to address the looming shortfall in skilled workers—an estimated three million by 2018. If we take advantage of the moment by building a robust middle-skill education and training architecture, we may be able to speed the recovery and rebuild the core of a sustainable middle class.

The decades-long decline of unionization weakened the training architecture for middle-skill workers; the funding and number of U.S. apprenticeship programs (which often relied on union infrastructure) decreased by 36 percent over the past 15 years. Further, private-sector investment in training is constrained by a range of market failures, in addition to legitimate concerns that training investments cannot be recouped if workers subsequently move to competitors.

The good news is that we have a blueprint for a comprehensive U.S. middle-skill training architecture. Examples abound: In the utility sector (in which up to 60 percent of the workforce will need to be replaced in the next ten years), it is the Center for Energy Workforce Development, a collaboration launched in 2006 between multiple utilities and unions to create a pipeline of technicians. In Kentucky, it is Metropolitan College, launched in 1998 by UPS, the state, the city of Louisville, and several community colleges to train around 2,000 students each year for UPS air-hub operations. In health care, it is the initiative sponsored by Kaiser Permanente and its unions, which offers more than 50 training programs across the health-care field with a 95 percent retention rate of the registered nurses it trains. These programs, whether centered around a region, an industry cluster, or a community college, share some common features: linking training to specific hiring needs; providing financial support for students (especially adult learners); and focusing on post-training placement.

The problem isn’t the lack of ideas. It’s scale. While every state has some kind of skills initiative, none of these efforts has the capacity to address the anticipated 2018 shortfall. The Obama Administration has taken several steps, including the launch of Skills for America’s Future, a public-private initiative to facilitate partnerships among employers and community colleges nationwide. And the Department of Labor is granting $2 billion over four years to finance expansion of community-college training programs.

However, it is the Administration’s more recent proposal that could be transformative. The $8 billion “Community College to Career Fund” would be co-administered by the Departments of Education and Labor to establish partnerships between community colleges and businesses to train two million workers.

Like many ideas in Washington these days, the fund proposal has fallen by the wayside, yet another victim of the near-absolute preoccupation with fiscal austerity. In theory, this idea should have strong bipartisan support—last time I checked, it’s pretty good politics to be in favor of facilitating work. However, recent disagreements over reform of the Workforce Reinvestment Act, currently the primary workforce-training program at the federal level, suggest philosophy is trumping action at least in the short term.

For now, we must rely on initiatives that do not require significant federal funding—with the support of state and local government, the private sector, unions, and philanthropy to pick up the slack. Here are three ideas that don’t need federal action and can help make progress while we agitate for action in Washington:

Accelerate common standards. Workforce development experts have long identified the need for a national transferable and stackable certification system. A stackable certification—the kind that has become standard for companies like Microsoft, Oracle, and Cisco—provides a clearly defined system of industry-recognized competencies tied to job opportunities and career advancement. More recently, a range of industry associations, educational institutions, and states have developed such skills-certification systems. But we now may be suffering from too much of a good thing. Workers and employers struggle to make sense of the 700,000 different certificates available nationally. A logical fix is to simplify and streamline training by industry sector. The National Governors Association could take the lead on this, borrowing the idea of the common core standards initiative in the K-12 arena. By making the credentialing challenge a major priority, it could corral the multitude of stakeholders across industries and states—and do so quickly.

Apprenticeship investment fund. Union membership is at a 97-year low. Yet over the last few years the annual Manpower Inc. talent shortage survey shows that the skilled trades are the hardest employment sector to fill. Although the AFL-CIO and Change to Win-affiliated unions sponsor apprenticeship programs across the country, such efforts should be scaled up further. To take advantage of the current union talent on the cusp of retirement, a union-financed National Apprentice Investment Fund could focus on expanding such programs in every sector facing shortages. This could be modeled roughly after the Employer Investment Fund structure used in the United Kingdom’s film and other industries to facilitate employer involvement in skills training. A percentage of union dues (possibly allowing partial repayment from employers and dues from new union members) could be used as a financing source.

An X Prize for data integration. We need to use twenty-first-century IT tools to meet the challenge of matching workers with training opportunities and potential employers. The lack of transparency surrounding educational and employment data is puzzling. I can find a dim sum restaurant in Boise, Idaho in two seconds, but somehow a national jobs database is elusive. Philanthropists should challenge the tech industry to develop a simple, national, searchable database of available jobs, requirements, and credentialing opportunities (the latter with user rankings a la Yelp). In the ideal world, users would complete a simple checklist of current skills that would yield a list of available jobs and certifications that match. Once that is built, we then need to challenge our best minds and leading tech companies—Google, Facebook, Amazon—to deploy it creatively on a national basis.

There is no real substitute for the federal government in bringing public policy ideas to scale. But we cannot wait for Washington to get its act together. The United States has a window of opportunity right now to build a stronger national architecture for middle-skill training. If we move with national purpose, we could have a “three-fer”: addressing looming shortages in critical skills; moving low-skill workers up the economic ladder; and helping to reduce poverty and spur growth. If we act now, we can ensure the future cable guy doesn’t have to rely on a savvy cousin to find a path to the middle class.