Lifelong Learning Infrastructure: Who Pays?

In 2024, Singapore enhanced its SkillsFuture Level-Up Programme, adding a new $4,000 top-up credit for mid-career Singaporeans aged 40 and above, on top of the $500 SkillsFuture Credit already available to every adult citizen. The program, funded through the national budget and administered through a digital platform that connects workers to approved training providers, is designed around a simple premise: continuous learning across a working life is not an individual choice that the market will adequately fund. It is a national infrastructure requirement, and it needs to be financed accordingly.

Singapore is a city-state with a population of 5.9 million and a labor market that cannot afford mismatches between workforce capability and economic need. It is not a direct template for the United States. But the policy logic it embodies — that lifelong learning infrastructure is a shared investment with returns that accrue to individuals, employers, and the broader economy — is increasingly difficult to argue against as automation accelerates the pace at which skills become obsolete. The question of who pays for continuous workforce development in the United States has a current answer: mostly employers for those they choose to train, mostly individuals for everyone else. That allocation is producing chronic underinvestment and distributing the cost of that underinvestment to the workers least able to bear it.

How the Financing Burden Is Currently Distributed

The U.S. system for financing adult learning and workforce training is not designed. It is accumulated — a layered set of programs, tax provisions, and employer practices that together produce an investment level well below what labor economists estimate the automation transition requires.

Employers are the largest single funder of workforce training in the United States, accounting for the majority of the $102.8 billion in corporate training expenditures recorded in 2025. But employer training investment is heavily concentrated: large firms with over 10,000 employees spend far more per worker than small and mid-sized employers, who account for the majority of employment but a fraction of training investment. Workers at small firms, which employ the largest share of the workforce, receive less training on average than their counterparts at large corporations, despite facing the same automation-driven skills transition.

The federal government contributes through WIOA-funded training, Pell Grants for eligible certificate programs, the Workforce Opportunity Tax Credit (WOTC), and a range of sector-specific funding streams. The aggregate federal investment in workforce development, approximately $2.8 billion annually through WIOA and related programs, is a fraction of the investment that peer economies make through their public financing systems. Denmark spends roughly 2 percent of GDP on active labor market policy. The United States spends approximately 0.1 percent.

Workers themselves bear a growing share of the training cost through tuition payments, foregone wages during training periods, and individual time investment in credential programs. The World Economic Forum estimates that 57 percent of employees pursue self-directed learning in response to skills shortages, a figure that reflects workers making rational investments in their own capabilities but also a system in which public and employer funding gaps are being absorbed by individuals.

What Other Countries Have Built

The international policy landscape offers several models of lifelong learning infrastructure that distribute the financing burden differently than the U.S. system, with outcomes that are instructive even where direct replication is not feasible.

Singapore's SkillsFuture combines individual learning credits, employer co-investment requirements, and sector-level skills councils that align training supply with labor market demand. The system is co-financed: government, employers, and individuals all contribute, with the government bearing a larger share for mid-career workers facing displacement. The program has achieved participation rates in adult learning significantly above the OECD average, and Singapore consistently ranks among the highest performers on workforce adaptability metrics.

The United Kingdom's Apprenticeship Levy, introduced in 2017, required large employers to contribute 0.5 percent of payroll to a training fund, with the funds accessible for registered apprenticeship programs. The policy logic was sound: making employers pay into a shared training pool reduces the free-rider problem that leads individual firms to underinvest. The implementation produced a more complicated record. Large firms accumulated levy funds faster than they could deploy them into quality apprenticeship programs, leading to a significant underspend of accumulated contributions. Smaller employers, who received a 95 percent government subsidy for apprenticeship costs, saw meaningful gains. The lesson from the UK experience is not that training levies do not work, but that they require complementary investment in program quality and intermediary infrastructure to convert funding into outcomes.

Denmark's active labor market policy provides workers with access to subsidized training during unemployment and facilitates sector-level wage negotiations that include training investment commitments from employers. The system is expensive and dependent on institutional relationships between government, employers, and unions that do not map directly onto U.S. labor market structures. But its core outcome, that Danish workers experience shorter unemployment spells and faster re-employment at wages comparable to their prior earnings following displacement, is directly relevant to the policy challenge the United States faces as automation accelerates.

What the U.S. System Is Missing

The United States has the components of a lifelong learning infrastructure scattered across disconnected programs. What it lacks is the architecture that connects them into a coherent financing system with shared accountability for outcomes.

Individual Learning Accounts (ILAs) are the most widely discussed structural reform proposal. Under an ILA model, workers would accumulate portable training credits over their careers, funded through contributions from employers, government, and individuals, redeemable at approved training providers. The accounts would travel with workers across jobs, eliminating the problem of employer training investment being captured by the employer rather than the worker. Several states, including Washington and Colorado, have piloted ILA-adjacent programs with promising early results. Federal ILA legislation has been introduced in multiple congressional sessions without passage.

Pell Grant expansion for short-term workforce credentials — programs of less than one year that do not currently qualify for Pell funding — has bipartisan support and has advanced further in Congress than most workforce financing reforms. Extending Pell eligibility to workforce certificates that meet quality and outcome standards would make the federal student aid infrastructure available for the credential pathways that the labor market increasingly demands, without requiring the creation of an entirely new financing mechanism.

Sector-based training levies, modeled on the UK Apprenticeship Levy but designed to avoid its implementation failures, would create a shared funding pool for industry-wide training that distributes the cost across competing employers and eliminates the free-rider problem at the sector level. The political prerequisite for this model is employer organization at the sector level, through industry associations or workforce partnerships, that can manage the collective investment and ensure that training programs meet quality standards.

The question of who pays for lifelong learning in the United States is not primarily a technical question. It is a political question about whether workforce development is understood as a shared national investment or as a private good whose costs should fall on the individuals who consume it. The current answer, by default, is the latter. The countries that have answered it differently are producing better workforce outcomes. That fact is increasingly difficult to ignore.

This article is part of the PPV Education and Human Capital series. The previous installment, The ROI of Workforce Certificates vs. Degrees, examined what the earnings data show about which credentials produce returns and under what conditions. The next, Education Systems Built for Jobs That No Longer Exist, examines why the institutional design of education continues to reflect a labor market that has fundamentally changed.

Have research, policy experience, or program models on lifelong learning financing? Reach out.