Welcome to My Weird Prompts. I am Corn, and I am sitting here in our living room in Jerusalem with my brother, Herman Poppleberry.
That is right, Corn. And just for the record, as a donkey, I have a natural affinity for carrying heavy loads, and today we are carrying the heavy load of government fiscal policy.
And I am a sloth, so I will be carrying my usual load of... well, sitting here and trying to make sense of what you are saying. Our housemate Daniel sent us a really interesting audio prompt this morning. He has been working in the world of impact investing for a while, and he wants us to dig into how governments actually pay for things like helping the homeless or reducing crime. Specifically, he is asking if this thing called pay for success is actually a better way to spend our tax dollars.
It is a massive topic, Corn. It touches on everything from how your local city council works to multi-million dollar international development projects. The core question is: are we paying for activities, or are we paying for results?
See, that already sounds like a business school lecture. When I think of the government doing good, I just think of them, you know, building a shelter or hiring social workers. Is that not how it usually works?
Historically, yes. That is the traditional model. For decades, the standard way governments have engaged in contracting for social good is through what we call input-based or activity-based funding. They basically say, here is a bag of money, go provide ten thousand meals to people in need, or provide five hundred beds in a shelter.
And what is wrong with that? If ten thousand people get a meal, that seems like a win.
Well, it is a win in the sense that people are not hungry for that moment. But I would argue it is a very narrow way to look at success. If you are just paying for the number of meals served, you are not necessarily paying to solve the problem of why those people are hungry in the first place. You are funding the process, not the outcome.
I do not know, Herman. That feels a bit harsh. If someone is hungry today, they need a meal today. You can not just wait around for a long-term outcome while people are starving.
I am not saying we should let people starve, Corn. Do not be dramatic. I am saying that the traditional way of contracting often creates a cycle where the organization providing the service is incentivized to just keep providing that service forever. If their funding depends on the number of beds filled every night, they do not actually have a financial incentive to help people move out of those beds and into permanent housing. In fact, if the shelter becomes empty, their funding might get cut.
Wait, so you are saying the current system actually rewards organizations for keeping the problem alive? That sounds like a conspiracy theory.
It is not a conspiracy, it is just bad incentives. It is how the contracts are written. Most government grants are very prescriptive. They tell the non-profit exactly how many staff to hire, what hours to be open, and exactly how to spend every cent. There is very little room for innovation or for shifting resources if they find a better way to do things.
Okay, but at least we know where the money is going. It feels safe. If I give you ten dollars to buy a pizza, and you show me a pizza, I know the transaction was successful. You are suggesting we give the ten dollars and hope that, I do not know, my hunger levels drop by twenty percent over the next month? That is much harder to track.
And that is exactly where pay for success comes in. It flips the script. Instead of the government paying for the pizza, they pay for the fact that you are no longer hungry and have learned how to cook for yourself.
But how does that actually work in the real world? Who pays for the ingredients while I am learning to cook?
That is the clever part. In a pay for success model, also known as a social impact bond, a third party usually steps in. Typically, these are private investors or philanthropic foundations. They provide the upfront capital to the non-profit to run the program. The government stays out of it at the beginning. They just set a target. For example, they might say, if you can reduce the rate of people returning to prison by ten percent over three years, we will pay you back the original investment plus a small return.
Hang on. This is where it starts to feel a bit weird to me. You are talking about private investors making a profit off of people stay out of jail? That feels like we are turning social problems into a stock market.
I knew you were going to say that. But look at it from the taxpayer's perspective. If the program fails, the government pays nothing. Zero. The investors take all the risk. If the program succeeds, the government saves a massive amount of money because they are not paying to house those people in prison, which is incredibly expensive. They take a portion of those savings and pay the investors. It is a win-win.
Is it, though? I mean, what if the investors only pick the easiest people to help? Like, if you are trying to reduce prison rates, they will only take the people who were probably not going to go back anyway. They call that "creaming," right? I read that somewhere.
You have been reading, Corn? I am impressed. Yes, "creaming" or "cherry-picking" is a real concern. But that is why the contracts have to be designed incredibly carefully. You use what is called a randomized controlled trial. You compare the group getting the help to a similar group that is not. You only pay for the improvement above the baseline.
It still feels like a lot of bureaucracy. You need lawyers, evaluators, investors, government officials... it sounds like half the money is going to the people in suits before a single person gets helped.
There are definitely high transaction costs, I will give you that. Especially in the early days. But we are seeing this evolve. In fact, by now in late two thousand twenty-five, the models have become much more streamlined than they were five years ago. We have moved toward things called outcome funds, where the government sets aside a big pool of money for specific results across an entire region, rather than negotiating one-off bonds for every single project.
I still think there is a value in the old way. Some things are just public services. You do not ask if a library is "cost-effective" based on how many people get jobs after reading a book. You just have a library because it is good for the community. Why can we not just have social services because they are good?
Because resources are finite, Corn! We live in a world of budgets. If we spend a billion dollars on a program that does not actually help anyone, that is a billion dollars we can not spend on something that does work. The traditional model often hides failure. Pay for success forces us to confront it.
Let us take a quick break from our sponsors.
Larry: Are you tired of your neighbors looking at your house? I mean, really looking at it? With their eyes? Introducing Eternal Fog! The only home security system that uses a proprietary blend of non-toxic, semi-permanent particulates to create a thick, luscious wall of grey mist around your property twenty-four seven. It is not just privacy, it is a lifestyle. Is it a house? Is it a haunted moor? Your neighbors will never know! Side effects may include mild dampness, loss of depth perception, and an accidental sense of mystery. Eternal Fog. If they cannot see you, they cannot judge you. BUY NOW!
...Alright, thanks Larry. I am not sure if that is even legal under Jerusalem zoning laws, but anyway, back to the topic.
I am still thinking about the mist, Herman. It sounds cozy. But okay, back to the government spending. You were saying that the traditional model hides failure. But does pay for success actually work? Have there been real examples where it was better?
Absolutely. One of the most famous early examples was at Peterborough Prison in the United Kingdom. They used a social impact bond to fund a program to help short-sentence prisoners reintegrate into society. These are people who usually get almost no support and have very high rates of re-offending. The program actually worked. It reduced re-offending by about nine percent compared to a control group. The investors got paid, the prisoners got their lives back, and the government saved money.
Nine percent? That is it? All that work for nine percent?
Nine percent is huge when you are talking about thousands of people and millions of pounds in prison costs, Corn! In the world of social policy, moving the needle by nine percent is a major victory. Most traditional programs struggle to show any statistically significant impact at all because they are never measured this rigorously.
But what happens to the non-profits? If I am a small charity helping kids with after-school programs, I do not have a team of data scientists to track "long-term life outcomes." If the government stops giving me my regular grant and says I only get paid if these kids graduate college in ten years, I am going to have to close my doors tomorrow.
That is a very valid point, and it is one of the biggest criticisms of this model. It can favor large, well-funded organizations that have the infrastructure to handle the reporting. But I would argue it also pushes those smaller organizations to partner up. It encourages them to look at their data and say, "Is what we are doing actually helping?" If it is not, maybe they should be doing something else.
That sounds very cold, Herman. "Maybe they should be doing something else." These are people's lives and passions.
Passion does not pay the bills, Corn, and it certainly does not fix systemic poverty. Effectiveness does. Look at the "Nurse-Family Partnership" in the United States. They have used pay for success models to scale up a program where registered nurses visit low-income, first-time mothers. It has decades of evidence showing it improves child health and maternal economic self-sufficiency. Because they had the data to prove it worked, they could use pay for success to get the upfront money to expand to thousands more families that the government was not ready to fund yet.
Okay, I see the logic there. Using private money to "bridge the gap" until the government is convinced. But I still worry about the "success" definition. Who decides what success is? If it is just "getting a job," what if that job is terrible and the person is miserable? Is that success?
This is where the tension lies. The metrics have to be meaningful. If you define success too narrowly, you get "perverse incentives." Like a job placement program that puts people in temporary roles just to hit the target and get the payout, even if the person is unemployed again two weeks later. To avoid that, you have to build in "durability" metrics. You do not just pay for the job placement; you pay if they are still employed six months later, or a year later.
It just seems like we are trying to turn humans into spreadsheets. I like the idea of the government being a bit more... I do not know, human? Like, "We care about you, so we are going to provide this service," rather than "We are going to provide this service because it has a five percent return on investment for a hedge fund."
But Corn, is it more "human" to keep funding a program that does not work just because it makes us feel good to provide it? I think the most human thing you can do is ensure that the help we provide actually changes someone's life for the better. If that requires a spreadsheet to prove it, then I am all for the spreadsheet.
I guess I can see that. But what about the cost-effectiveness part? Daniel asked if it is a "brighter and more cost-effective alternative." Is it actually cheaper for the government?
In the long run, yes, because you are only paying for what works. You stop flushing money down the drain on ineffective programs. But in the short term, it can actually be more expensive because of the interest you have to pay the investors and the costs of the evaluation. It is an investment in the future. You are paying a premium now to avoid massive costs later.
I am still not entirely convinced. It feels like we are privatizing the social safety net. If the government starts outsourcing all its "good deeds" to private investors, what is the government even for?
The government is still the one setting the priorities, Corn! They are the ones deciding which problems need solving. They are just changing how they buy the solution. Think of it like this: if the government needs a new bridge, they do not usually hire a bunch of people and buy some cement and try to build it themselves. They contract it out to a company that knows how to build bridges, and they only pay when the bridge is finished and safe. Why should social services be any different?
Because a bridge is a physical thing! You can see if it stands up. A human life is complicated. You can not just "build" a successful person. There are so many factors outside of the program's control. What if the economy crashes? The program could be doing everything right, but if there are no jobs, they fail the metric. The investors lose their money, the non-profit goes bust, and it is not even their fault.
That is exactly the risk the investors are being paid to take! That is why they get a return. They are taking on the risk of economic shifts or unforeseen circumstances so the taxpayer does not have to. If the economy crashes and the program fails, the government has not lost its budget for that program. It can take that money and pivot to something else.
I suppose. It just feels like a very stressful way to run a charity. Always looking over your shoulder at the data, wondering if you are going to hit your "target" so you can keep the lights on.
It is stressful, but it is also empowering. Imagine being a social worker and knowing for a fact that your work is having a measurable impact. Instead of just "doing your best" and hoping for the change, you have the data to show exactly how many families you have helped stay together or how many people you have helped find stable housing. That is a powerful motivator.
I think we might have to agree to disagree on how "empowering" a spreadsheet is, Herman. But I do see your point about the traditional model being a bit of a "black box" where money goes in and we are not always sure what comes out.
Exactly. And that is the "brighter future" Daniel was asking about. It is a future where we are more honest about what works and what does not. It is a future where we are willing to try new things because the financial risk is managed by people who can afford to lose it, rather than the public purse.
So, what are the practical takeaways here? If someone is listening to this and they are, I do not know, a voter or someone who works for a non-profit. What should they do?
First, I think we should all be asking for more transparency in how our tax dollars are spent on social programs. We should be asking, "What is the intended outcome, and how are we measuring it?" Even if it is not a full "pay for success" model, that mindset is crucial.
And for the people in non-profits?
Start getting your data in order. Even if you are not doing a social impact bond, the world is moving toward "outcomes-based" everything. If you can prove your impact with numbers, you are going to be much more successful in getting funding from any source, whether it is the government, a foundation, or a private donor.
I would add a takeaway for the government folks, too. Do not forget the "human" element. Data is great, but do not let the metrics become so rigid that you lose the ability to help the people who do not fit into a neat little box. Sometimes the "success" is just being there for someone, even if it does not show up as a percentage increase in the state's gross domestic product.
I can get behind that, Corn. The metric should serve the mission, not the other way around.
Well, I think we have covered a lot of ground. From Peterborough prisons to Eternal Fog.
Do not buy the fog, Corn. I am pretty sure it will just make our laundry damp.
Too late, I already have the brochure. Anyway, thank you to our housemate Daniel for sending in that prompt. It was a deep dive, but I think I understand the world of government spending just a little bit better now. Even if it makes me want to take a very long nap.
You always want to take a nap, Corn.
True. You can find My Weird Prompts on Spotify, or visit our website at myweirdprompts.com. We have an RSS feed there for subscribers and a contact form if you want to send us your own weird prompts. We are also available on all major podcast platforms.
And if you are an investor looking to fund my "More Carrots for Donkeys" social impact bond, you know where to find me.
No one is funding that, Herman. Goodbye, everyone!
See you next time!
Larry: BUY NOW!
