Do Incentives Actually Create More Apprenticeships?

The money is moving 

The US Department of Labor (DOL) is backing a 2025 executive order to reach 1 million active apprentices. On November 2, it opened applications for six pay-for-performance incentive funds, worth $162 million in total, that offer sponsors up to $6,000 for each eligible new apprentice. 

The money goes to apprenticeship sponsors: either employers or intermediaries that run programs on employers’ behalf. Many states already use a mix of subsidies and tax credits to target employers more directly. Apprenticeship.gov lists 25 states with tax credits for hiring apprentices. South Carolina offers a tax credit of up to $4,000 per apprentice per year. Illinois offers up to $3,500 per apprentice, with $1,500 more in underserved areas. New Jersey reimburses up to half of a new apprentice’s wages, to a maximum of $12,000 over 52 work weeks. The DOL expects its new incentive funds to stack on top of these state initiatives, for potential totals of roughly $10,000 to $18,000 per apprentice depending on the state. 

These incentives aim to lower the cost of starting or expanding a program. That cost is sometimes apprentice wages, but more often related training, administration, and on-the-job learning, work that intermediaries often handle for multiple employers.  

The question is – do these incentives create new apprenticeships?  

What the evidence says 

US evidence is thin 

No rigorous US study has yet compared employers and intermediaries that received incentives with similar ones that did not, which is what we would need to know how many apprentices were truly additional. More evidence may be coming: the first priority question in the DOL’s FY 2027 Evidence Plan is how pay-for-performance and incentive models can get employers to adopt Registered Apprenticeship, including the right retention period, incentive size, and payment recipient. 

Abroad: small effects, or high costs 

In England, 76% of employers who applied for or received a hiring incentive said they would have recruited the same number of apprentices anyway. The report estimated the incentive raised recruitment by about 11%, falling to about 6% once hiring that was merely brought forward is set aside. That is a small boost for the money. 

Australia saw a much larger effect. A recent evaluation (preprint) of its COVID-era apprentice wage subsidy found about 70% more apprenticeship starts than expected without the program. But it cost about $40,000 AUD per additional start (roughly $28,000 USD), and non-trade apprentices were about 7 percentage points more likely to cancel within 18 months, against a typical rate of 30% to 40%. In other words, the subsidy brought in many more people, but a larger share of the added non-trade apprentices did not stay. Taken together, the international evidence suggests incentives can raise apprenticeship starts, but either by small amounts or at a high cost per additional start.  

Even without incentives, apprenticeships often pay off for employers. A study of 68 employers found that about two thirds recouped their investment. The typical employer got back $1.44 for every $1 it spent. But results varied a lot. In the study’s medium estimate, one in four employers lost about 7% or more, while one in four gained about 121% or more. Most of the cost comes early, when apprentices are still learning and contribute the least. 

So the case for public money is not that apprenticeships are a bad bet on average. It is that employers cannot be sure they will come out ahead, and that uncertainty may deter exactly the firms we most want to encourage to start an apprenticeship program: small employers, sectors new to apprenticeship, and occupations with long ramp-up periods.

Completion is the bigger lever 

Getting apprentices in the door is only half the job. Social Finance’s analysis of federal apprenticeship data shows that of the roughly 167,000 apprentices who started a registered program in 2017, only 46.8% finished within six years, and cancellations spike in the first few months. 

In conversations with unions, workforce boards, and funders, we heard the same reasons apprentices leave: lack of childcare or transportation, financial hardship, sudden life events, and poor job fit. 

Supports aimed at those problems show promise. These include pre-apprenticeship programs, mentoring, help with childcare and transportation, and living stipends or emergency funds. In a 2024 evaluation of Oregon’s highway construction workforce program, receiving supportive services was associated with being about 10% more likely to complete. For those who got both financial and non-financial support, the figure was 13%. Not every service helped equally: childcare was associated with a 10% gain, while gas and travel assistance showed no gain. To target supports well, we need to know which ones work for whom. That means tracking the services each apprentice receives and linking them to whether they finish. 

What looks promising in the new funds, and what we still need to learn 

The new federal incentive funds already have several interesting design features, which Apprenticeships for America helpfully summarized in this quick-response guide. Taken together, they offer a set of natural comparisons, each with different payment sizes, timing, and targets, though they are not randomized, so differences in sector and sponsor mix will complicate comparisons. Four things stand out as promising, each with further areas for testing and learning:

1. Paying for growth versus relabeling

The funds are designed to pay only for new apprenticeships. Under the D’s recently updated FAQ for the program, a sponsor is paid only for apprentices enrolled above a “substantial growth threshold,” defined as 125% of the median active apprentice count over the 12 quarters before July 1, 2026. Programs registered after that date are paid for every new apprentice. The aim is to fund expansion instead of subsidizing programs with stable counts. That principle is right, though the formula is imperfect and may not work as intended.  

There is always a difficult tradeoff between simplicity and precision when designing a pay-for-outcomes program like this one. The substantial growth threshold should be evaluated on whether it strikes that balance well: how burdensome is it to calculate, and does it separate real growth from relabeling in practice? 

2. Targeting the money (e.g., smaller employers, newer sectors, and harder-to-enroll populations)

Several of the funds build in some targeting. Florida’s shipbuilding fund pays a $2,000 baseline and raises it to $6,000 for programs that meet at least one priority condition, such as serving a rural community or a high-poverty area. The same idea appears in DOL’s manufacturing fund, which launched earlier and served as a model for the newer funds. It caps support at 200 apprentices per sponsor, which spreads the money across more sponsors. 

What we still need to learn is whether the money reaches sponsors facing the greatest risk. Applications are rolling until funds are obligated, so we should watch whether small employers and sectors new to apprenticeship actually apply, or whether established sponsors with more capacity capture most of the funds. 

3. Paying for outcomes that matter

Incentive payments are spread across milestones, not paid on enrollment alone. Jobs for the Future’s fund pays $3,750 at 90 days, the early window when cancellations spike, and $2,250 at completion or 12 months, whichever comes first. The telecommunications fund goes furthest, with separate payments for retention, for completing related technical instruction, and for finishing the program. Comparing designs can help show whether payments reduce cancellations and improve completion rates, and which retention period and incentive sizes are right. 

But a milestone payment alone may not be enough. Keeping apprentices enrolled means tackling the reasons they leave, with supports such as mentoring, childcare assistance, transportation, and living stipends or emergency funds. A payment of a few thousand dollars is unlikely to cover those costs, and it arrives only after the apprentice has already stayed. Funding for retention supports may be needed up front, alongside the milestone payments.

4. Using data to learn what works

DOL’s incentive funds push sponsors toward better use of data. Because payments depend on verified enrollment, retention, and progression milestones, sponsors now have a financial reason to track whether apprentices stay and finish, not just whether they enroll. This is a real step forward, but more is needed.  

The required data does not capture the supports apprentices receive, such as childcare, transportation, or mentoring, or longer-term outcomes like post-completion wages. Adding both would let sponsors see which supports help which apprentices finish and let evaluators compare designs across the funds. That evidence would help prove ROI, strengthen the case for more investment, and open the door to even more creative and ambitious outcomes-based funding models, such as downside insurance that covers more of an apprenticeship’s early costs, employer-repayment models, or additional public payments tied to apprentices’ earnings gains. 

The bottom line 

The evidence so far is mixed: incentives have raised apprenticeship starts abroad, but either modestly or at high cost per additional apprentice, and US evidence is not yet available. Incentives deserve a role in growing apprenticeships, but we are still learning how to design them well. The design of the new federal funds shows promise: the funds pay for growth, reward retention and completion, and require better data. Whether they create apprenticeships that would not otherwise happen and keep those apprentices through to completion depends on questions the funds themselves can help answer. For some employers and emerging sectors, a few thousand dollars in incentive payments may not be enough. Still, the data these funds are building lays the groundwork for bolder models that share more of that risk. 

Designing an incentive fund or evaluating one? Social Finance helps agencies and funders structure pay-for-performance models, track the supports learners receive, and link them to completion. Get in touch to learn more >