Diverse group of six people—government officials

When Social Investment Payments Depend on Results: Who Carries the Risk?

Investors usually take the first loss when social investment payments depend on results, but they shouldn’t carry every loss. The fairest deals put each risk with the party best able to control it, and they spell that out before any money moves.

That split rarely happens by accident. Outcomes-based financing risk spreads across governments, investors, providers and evaluators. One vague clause can push a loss onto someone who couldn’t prevent it.

At the Social Investment Taskforce, we share research and guidance for governments, investors and providers building these deals. We put this guide together to help each party see where its share of the risk sits. It covers who sits at the table, how payment models split the downside and how guarantees shift exposure.

Who Sits at the Table in a Social Outcomes Contract?

Government budget meeting in a modern office conference room where

Five parties usually sign or influence a social outcomes contract, and each one holds a different slice of the risk. Their incentives overlap but rarely match, which is where tension in these deals starts.

Below, we’ve explained what each party brings and what it stands to lose.

Outcome Payers: Governments That Only Pay for Verified Results

An outcome payer, usually a government agency, agrees to pay only after an evaluator confirms the program hit its targets. Such a condition keeps public budgets from funding programs that don’t work, which is the core promise of results based social finance.

Still, the payer faces a quieter risk: paying for gains that would have happened anyway. A strong local job market can lift a training program’s numbers on its own (baseline trends often mislead funders).

On the payer side, ask for a comparison group in the evaluation design. With one in place, you pay only for the change the program caused.

Investors Who Front the Working Capital

Someone has to fund the program before any outcome is measured, and that role usually falls to investors. Their working capital keeps services running until results come in, often years later. If the program misses its targets, they can lose part or all of what they put in. For financial institutions used to judging credit risk, this is a different bet. Repayment depends on social results rather than a borrower’s cash flow.

That level of risk explains why investors push hard on contract terms. A multi-year wait for repayment also ties up capital the fund could use elsewhere.

Before committing, check how much capital sits at risk at each payment milestone. Smaller, earlier payments cut the chance of a total loss.

Service Providers Under Pressure to Hit Targets

Service providers deliver the social services themselves, like job training, early childhood education or mental health support. Under a pure social impact bond, investors absorb the financial loss, yet providers still feel the pressure to show results.

Targets can push teams toward easier participants who’d likely succeed anyway (evaluators call this cherry-picking, and it skews the results). Some contracts also tie part of the provider’s fee to outcomes, which moves delivery risk onto smaller organizations.

If you run a provider, the fee schedule shows which share of your income depends on targets you can’t fully control. That gap belongs in your budget from day one.

Independent Evaluators and the Data They Sign Off On

No payment moves until an evaluator signs off. These independent reviewers check whether the program met its targets. They rarely carry financial risk, but their methods for measuring outcomes decide who gets paid.

Anyone who has read a few of these contracts knows the evaluation plan often gets skimmed while the payment schedule gets negotiated line by line. Yet a weak data plan can stall payments for months.

To avoid that delay, agree on data sources, collection dates and access rights at the outset. Clear terms protect every other party at the table.

The Communities Whose Lives the Program Is Meant to Improve

Communities never sign the contract, but they carry the human cost if a program fails or shuts early. A family that loses support halfway through a parenting course receives no benefit from an investor refund.

Even a program that hits its targets can fall short for participants. Targets chosen because they’re easy to measure can miss what people value most. A short-term job placement, for example, says little about whether someone keeps the job.

Bringing community voices into target setting from the start helps close the distance between what a contract measures and what participants need. They know which outcomes will last after the program ends.

Impact Investment Payment Models: How Each One Splits the Downside

With the players mapped out, the payment model decides how losses fall among them. Each model structures social outcomes contract payments in its own way and moves a different share of the downside away from government.

Take a look at how the most common impact investment payment models compare.

Payment modelWho absorbs a shortfallWhat to check before signing
Social impact bondInvestors lose capital if outcomes fall shortThe minimum result needed for any repayment
Outcomes-based contract with provider feesProviders lose the outcome-linked share of feesHow much of the fee rides on results
Outcomes fundPooled funders spread payments across many projectsWhether one project’s failure drains money from others
Development impact bondInvestors, with donors paying instead of governmentsWho pays if a donor changes priorities
Rate cardProviders or investors, at a fixed price per outcomeWhether prices cover the cost of harder cases

In a first-time deal, the social impact bond row usually deserves the closest look, because investors set the terms everyone else works within. Providers face a different picture, since the fee-based contract hits their budget hardest. Price your services with room for missed targets.

Lawyers and finance professionals gathered around a glass table

Shifting Outcomes Based Financing Risk With Guarantees and Concessional Capital

Once you know which party absorbs a shortfall, you can start moving part of that loss elsewhere. In many larger deals, concessional finance does that job. It mixes cheaper public or philanthropic money with private capital.

Development agencies have already changed how they use that cheaper money. According to Convergence data on guarantees, grants made up 41% of development agencies’ concessional capital in 2022 but just 10% in 2024. Over the same period, guarantees rose from 11% to 27%, and senior debt grew from 19% to 24%.

Because a guarantee caps an investor’s loss without spending money up front, a funder can back more deals with the same budget. Raise private capital for an unproven program and you’ll often find a first-loss layer decides whether investors stay in.

A first-loss layer is only one option, though, and most deals stack several protections to cover different gaps:

  • First-Loss Capital: A foundation or donor absorbs the earliest shortfalls before private investors lose anything. That cushion often persuades cautious investors to back an unproven program. Ask what share of losses it covers, and confirm it sits under the investors who need it most.
  • Partial Guarantees: A partial guarantee covers a set share of investor principal if outcomes fall short. Most stop at principal, so investors can still lose the returns they expected. Check whether returns are covered too, and price the deal on what the guarantee actually protects.
  • Recoverable Grants: These grants become repayable only if outcomes land, so the funder gets money back from programs that succeed. That money can then back the next new program. Funders testing a new model can use one to share early risk without demanding a fixed return.
  • Tranches: Senior and junior tranches give investors different risk levels within the same deal. Junior holders take losses first, while senior holders accept lower returns for stronger protection. Match each layer to its holder’s appetite for loss before you set the rates.
  • Payment Caps: This limits what the government owes if results beat targets. While the limit protects public budgets, it also trims the upside investors can earn. Agree on the cap before launch, because renegotiating after strong results usually strains the partnership.

Match each tool to the risk you want to transfer. A guarantee on the wrong layer protects the wrong party. Placing a first-loss layer under senior investors, for example, does little for a small provider whose fees ride on outcomes.

So list every party’s exposure before you choose a tool, then check that each protection lands where a shortfall would hurt most.

Measurement Disputes Decide Who Loses When Results Fall Short

Every guarantee and payment condition depends on a number someone has to measure. Under the US Treasury SIPPRA program, the federal government pays only when an independent evaluator validates that a project met its predetermined outcomes. Those policy terms make the evaluator’s data the hinge of every payment.

In practice, disputes start long before the final report. Data arrives late, or a partner agency changes how it records outcomes. Participants also drop out and leave the comparison group too small to show a clear effect.

In practise, whoever holds the raw data controls how the dispute ends. That gives providers and government agencies an edge over investors, who usually see results only through summary reports.

So write a dispute process into the contract before launch. Name a tiebreak reviewer, set deadlines for data corrections, and agree in advance how missing records count. Those steps stop a measurement argument from becoming a loss for the party with the least control.

Government policy analyst seated alone at a cluttered desk

Where the Risk Should Sit Before You Sign

Outcomes deals work when each party takes responsibility for the risks it can manage. Investors take on capital risk, providers own delivery, and governments own the targets and the data rules.

Before you sign, map every risk to a named party and check that the contract says the same thing. Then test the payment and measurement clauses against a bad year.

The Social Investment Taskforce shares research and guidance for governments, investors and providers building social investment deals. Contact the team to talk through your structure and spot gaps before they cost you.