Tuesday, December 25, 2007

It's that time of year when lots of people are giving to various charities, and I've been thinking about, if I decided to give, how I would decide. My mental economic model for businesses is that businesses engage in trade. Trade is the exchange of one thing for a more valuable thing. On both sides. From that I've tried to build a model for charities and philanthropy.

Economics of trade

Suppose you (running a business) have a sandwich that you're selling for $2. It's very likely that the sandwich is worth less than $2 to you. If it was worth more than $2, you'd rather keep the sandwich, so you wouldn't be selling it. So let's say it's worth $1 to you. Now suppose I have $2 and decide to buy a sandwich from you. It's very likely that the sandwich is worth more than $2 to me. If it was worth less than $2, I'd rather keep $2 than buy the sandwich. So let's say it's worth $3 to me. Both of us have chosen to trade something less valuable for something more valuable. Before the trade, the sum total of what we had was $1 (your sandwich) + $2 (my money) = $3. After the trade, we have $2 (your money) + $3 (my sandwich) = $5. Trades generate value out of thin air. The world is now $2 richer because of this trade. Free markets make for wealthy countries because people aren't prevented from trading.

Model for Gifts

If you think about trades, they work because both sides have the right incentives. When they both choose to trade, the world is better off. When either chooses not to trade, there's no harm done. Yes, there are some cases where the world is better off if they trade, but they don't both choose to trade, but in those situations a change in price will enable the trade, and both will end up benefiting.

With giving gifts, not only charities but also Christmas gifts, birthday gifts, etc., the person paying and the person receiving the sandwich are not the same. It's much harder to ensure that the trade is actually increasing wealth in the world. Suppose I buy you a sandwich, and I paid $2 for it. I don't know how much it's worth to you. You could say that it's worth $5 to you, but since it's not your money on the line, there's no incentive for you to say the true value. Suppose it's worth nothing to you. So before the trade, the seller has $1 (sandwich), I have $2 (money), and you have $0, a total of $3. After the trade, the seller has $2 (money), I have $0, and you have $0 (sandwich), a total of $2.

Gift giving opens up the possibility that a trade makes the world worse off economically. There are other reasons to give gifts of course, but just keep in mind that economically they're not so good.

Christmas and Birthday gifts are rare; the most common source of economic loss from giving is kids (and to a lesser extent, spouses). Kids might say they really want the $150 pair of shoes, but since they're not the ones working 20 extra hours to buy those shoes, they have no incentive not to make the trade. They make that sad face, or whine, or tell you that all their friends have those shoes, or nag you a great deal. They know how to manipulate parents. Lots and lots of inefficient trades are made by or for kids, causing parents to struggle to make ends meet.

These inefficient trades also can occur with charities. If you're giving to a charity and they “do good” with that money, you don't know whether what they did was worth more than what you gave. You end up judging based on how good it makes you feel, which means charities have an incentive to make you feel good, through special events (seeing lots of people involved in a “special event” makes you much more likely to give money), pictures of needy people (seeing a few people makes that emotional connection that you don't get if you read about helping millions), and other tactics. My view of charities, until recently, has been that there's a high potential for inefficiency and emotional manipulation, just as with kids. I've been trying to form a model that would help me make decisions about whether to give, and to whom, that isn't just to make me feel good, but something that will actually do good.

Externalities

There's something I left out of the model of trades. There's more than just the two parties involved. Maybe the business polluted to make that sandwich. Maybe I littered instead of throwing the sandwich wrapper away. The trade didn't take into account these “externalities” — effects on the rest f the world outside the seller and the buyer.

The standard solution to this is to estimate those effects and charge people for them. Let's take an extreme example, with $5 of pollution for a sandwich. (I think this is extreme for sandwiches but there are probably other industries for which it's reasonable to say the damage to the environment is higher than the value of the product.) Before the trade, there was $1 (your sandwich) + $2 (my money) + $5 (rest of world) = $8, and after the trade there was $2 (your money) + $3 (my sandwich) + $0 (rest of world) = $5. Although the seller and buyer are better off, the trade made the world worse. If we could charge the seller $5 for polluting, the price of the sandwich would be $7 instead of $2. And at $7, I wouldn't buy the sandwich, since it's only worth $3 to me. The trade would be stopped, which is just what we wanted here. To make money, the business needs to figure out how to reduce pollution.

I like the charging-for-pollution solution better than laws against pollution, or trade-and-cap systems (I'll have to make another post about that). With a law, the business's incentive is to fight the law, doing as little as possible, because they still make more money if they pollute. Whereas with charging, the businesses makes more money by not pollution. Since profit motivates business, I want the system to give more profit when the business pollutes less. However it's often impractical to measure the impact of pollution, and that's why we have simplistic black-and-white laws in place.

Model for Philanthropy

We have this artificial line between business and charity. We even call the charities “non-profits”, to distinguish them from businesses (even though some businesses don't make a profit). And there's the sense that charities do good for the world and businesses are bad. I think the world is more complicated than that. There's a whole set of potential projects that have both business and philanthropic aspects. For example, microfinance can generate profits and help lots of people at the same time. Tesla Motors might not ever make money, but it could jump-start the market for cleaner cars. Neither counts as a charity; you're not just handing out money. But there's the potential for it to do more good than a charity. The tax laws in the U.S. encourage giving to charity over investing in a “socially reponsible” business that does good things for the world.

If I'm going to do something with money that I don't need, I want to put it into a place that gives the highest return on investment. But in the case of philanthropy, in that return I need to consider not only what I get back, but also what good it does for the world. Externalities are a way of looking at this. Normally an externality is something negative. But there are also positive externalities. If I invested in a business that makes less profit than other businesses, but generates a lot of good in the world, that's a positive externality. The return on investment for me might be low, but the return to the world could be high. Whether it's classified as an investment or a charity isn't relevant (except there are tax implications that affect the return on investment); I want to find things that have a high rate of return. There's also regular shopping. Just as I might avoid “sweatshop“ products, I might favor products that have positive externalities.

Giving

Unfortunately, I haven't found much that can guide me to high-return philanthropy. Businesses report cost and revenue but not externalities. Charities report costs and donations but not the effects they have. I started following GiveWell a while ago because they seemed to be the only ones that came even close to what I was looking for. I was pleased to see they got mainstream press recently. They tend to pick things that are more tangible, and they seem to consider “saving lives” separately from economic benefit, but in general I like what they're saying and doing. I like that they focus on benefits more than cost efficiency, and they're trying to compare different approaches to see what's most effective. I'm also following Acumen Fund, which tries to help people through investments rather than charity.

Businesses looking for customers use emotions, marketing, sales tactics, and they come to you. When they get investors they use numbers, and you tend to go to them. I want to see the numbers side of philanthropic organizations. I don't want them contacting me to tell me how they're “doing good”. I assume almost all of them are doing good. What I'm really trying to decide, by building an economic model, is where to invest. It may be giving to a charity; it may be investing in a business; it may be buying products. It'll probably be all three. The main problem I see is that there isn't enough information, for businesses or charities, and that there's this artificial line drawn between the two. There should be a unified way of thinking about business and charity that finds the best of both, and allows for new types of organizations that don't fit into the current system. Good information about externalities would change the world.

Labels: ,

10 comments:

Josh M wrote at Friday, January 4, 2008 at 10:41:00 AM PST

Great post, Amit! You've hit on one of the most important waves of change for the non-profit world currently. Attempting to quantify said positive externalities is becoming more commonly known as SROI -- social return on investment.

I founded www.MyTwoFrontTeeth.org several years ago to help streamline the process of getting gifts to underprivileged kids during the holidays. In 2003, we competed in two social business plan competitions whose goal was to value new non-profit ideas based on their ability to sustainably generate SROI.

So how do you measure the positive social effects of giving toys to poor kids? Truthfully, we didn't even try. We focused our SROI model instead on how we could more efficiently get the cash & gifts to the kids & family services agencies using an online model. Our social return was largely valued on the thousands of volunteer hours and money we could save the world (which could then be re-appropriated toward other useful SROI generating activities.)

The base assumption here is that the world has plenty of demand (donors) and supply (poor kids) when it comes to charitable gift-giving over the holidays.

So the calculation of (and subsequent ranking/sorting by) SROI is not especially useful if you don't at least generally know toward which causes you want to put your money to work. Once decided however, having all non-profits of like mission agree on a set of SROI measurements would unlock this change -- there's a lot we could learn from GAAP and the SEC here. To boot, it would be awesome if the non-profit world had the market mechanisms to allow for aggressive M&A.

The joint Berekley Haas/Columbia Global Social Venture Competition site has some great resources and speakers on this topic.

(Post-script -- we took second place in both competitions; full bus. plan here, for those interested: http://tinyurl.com/2kl6oq)

Amit wrote at Sunday, January 6, 2008 at 11:09:00 AM PST

Thanks Josh! I agree with you that calculation of SROI (thank you for introducing me to this term) may not be important in some situations.

On the supply side, (in the mental model in my head) organizations look for a way to increase social well-being. In some situations there are plenty of givers but not enough recipients, and non-profit organizations can help increase the recipients. (For example if there's plenty of investment in business, but there aren't enough businesses, an organization can help people start businesses.) In some situations there aren't enough givers, and the goal would be to increase donations through fundraisers, direct marketing, etc.. In some situations there are plenty of both but the matching is inefficient, and non-profits like MyTwoFrontTeeth and Kiva can solve the matching problem.

On the demand side, (in the mental model in my head), each giver has a different “value function” that directs them to give to things they value; this is the same as in business, where all customers don't buy the same thing because they value different products differently. Since everyone has different values, it may be hard to get organizations to standardize on SROI metrics. Perhaps there are a few axes that many people agree on. Also, what the giver is doing is deciding on organizations. Money going to one organization is not going to another. In contrast, each organization's view of the world is that money is going to them or it's not, so the goal is to maximize what they receive.

I've also been thinking about return vs. risk management. On the return side, I should socially invest in things that produce a high return on investment. But on the risk side, I don't want all my eggs in one basket. Just as I don't buy just one product, and I don't invest my retirement savings in a single stock, I don't think I should give to only one non-profit. I'll continue pondering how to put this into the economic model.

ted b wrote at Monday, January 7, 2008 at 6:35:00 PM PST

Amit,

Interesting thoughts. I have to take issue with some of your suppositions, however.

First, the value of the sandwich is what the market will bear - if it can be sold for $2, the fact that the sandwich maker, with his (perhaps large) stock of sandwiches, would sell a sandwich for only $1 if he had to has no effect on the fact that $2 can be had for it. Therefore, the sandwich's value (to the sandwhich maker and to the market for sandwiches) is the price that can be commanded for it - he can, after all, command and receive $2 for it. The sandwhich maker would not give it to someone else for $1 when he could get $2 for it, therefore its value to him (or her) is $2.

As for the person purchasing the sandwich, they might be willing to pay up to $3 but if sandwiches are going for $2 and could be purchased for that amount, i think the purchaser would say its worth $2 to him or her, not $3. After all, how many times do you hear people say something like

"You want *how much*? ... $3 ... For *that*?... Why, i can get the same thing around the corner for $2."

So, again, i think that if the market price for sandwiches is $2 (and if that's what the market will bear) then, that's the value of the sandwich - to both the buyer and the seller. Nobody values something for less than can be had for it and nobody values something for more than can be had elsewhere.

See for example, http://en.wikipedia.org/wiki/Labor_theory_of_value

Re your comment "When either chooses not to trade, there's no harm done" ... that's contrary to modern trade theory - check out Ricardo

see http://en.wikipedia.org/wiki/Comparative_advantage

I think your thoughts on externalities and that those real costs are not reflected in prices are good ones - i think that's where government needs to play a bigger role.

Thanks for your thoughts.

Ted B.

Amit wrote at Monday, January 7, 2008 at 7:13:00 PM PST

Thanks for your comments, Ted! I may be using the word “value” in a non-standard way; perhaps I should use “utility”. When I say “value” I mean the price at which you are neutral about whether you'd rather have the money or the item. The sandwich maker would prefer to have any amount greater than $1, so the value (utility) is $1. The sandwich eater would rather have a sandwich than any amount less than $3, so the value (utility) is $3.

I still think there is no harm done by not trading. There is a lost opportunity, which means that you're not maximizing your utility, but that's different than performing an action that decreases utility. I do think opportunity cost has a place in the discussion of philanthropic goals; I just haven't put it into my mental model yet.

ted b wrote at Wednesday, January 9, 2008 at 8:13:00 AM PST

Amit,

I still think that the value (or utility) of anything in the context of people buying and selling in a market is the market price - what a good or service can be sold or bought for. Nobody is going to value (or have utility for) something that is less than what they can command in the market place - that would be like saying "I have a $2 bill but I really only value it at $1 and would be indifferent to someone coming along and taking it for a buck" - it doesn't make sense. If a sandwich maker is willing to sell for $1 in a market where sandwiches are being sold for $2, then he will sell for less than $2 and take business away from those who are selling for $2. Eventually, the market price for sandwiches will approach $1 (in a competitive market where there is no product differentiation).

As for "no harm in not trading" thats not true - if two entities are able to produce goods at different relative production levels, then they benefit from trading. Total output is increased as are relative consumption levels.

Example: Country A and Country B both produce 2 goods, clothes and food.

Say Country A can produce 100 tons of clothes and no food or 100 tons of food and no clothes or any combination in between (say 50 tons of each, for example).

Say the same is true for Country B but say that they are twice as efficient at producing food than clothes, i.e. they could produce 200 tons of food if they focused on doing just that and made no clothes. And say that they are just as efficient at producing clothes as Country A - they could make 100 tons of clothes if they focued soley on that, just like Country A.

If there was no trade, and if they divided their efforts between making food and clothes, then total output would be 150 tons of food and 100 tons of clothes (since Country A could make 50 tons of clothes and 50 tons of food and country B could make 100 tons of food and 50 tons of clothes).

If they specialized in what they were good at then total output would be increased (200 tons of food and 100 tons of clothes). Then if they traded (at, say, an exchange rate of 1 ton of food for 2/3 of a ton of clothes) then both of their consumption levels would be greater than if they didn't trade. They both could consume 50 tons of clothes each and country A could consume 75 tons of food and Country B could consume 125 tons of clothes.

Trade with specialization increases total output and allows for greater consuption. Distribution of the extra consumption capacities inside country A and B is a different matter.

See http://en.wikipedia.org/wiki/Comparative_advantage

Best regards,
Ted B

Daniel wrote at Monday, February 18, 2008 at 4:31:00 PM PST

I meant to comment on this a month ago, but I haven't been sick until now. :-)

I think you're playing fast and loose with your definition of utility. In your market model, it's definitional that the voluntary trade only occurs because there's an increase in net utility. The dollar values you give are just stand-ins for utility. For example, if I have a contagious disease and you know it, you'll require more than $2 from me, which has nothing to do with the sandwich and everything to do with making the utility equation work.

But your model for gifts suddenly loses all this sophistication. After the trade, the seller has $2, you have $0, and I have $0. But that's not possible. Since these dollar values are just stand-ins for utility, you have more than $0. In fact, the utility for you of giving me the sandwich must have been at least $2 -- it cannot have been less than $2 or you would have given me $2 instead of buying a sandwich for me and giving me the sandwich.

I think the missing thing that unites both markets and gifts is that we have expectations about what a thing is worth to other people. The sandwich seller will not long believe that his sandwiches are worth $1 if you and everyone else buy them for $2: if I come along and offer him $1.50, he'll be unlikely to trade.

And your assignment of a utility >= 2 to giving me the sandwich is probably dependent on your expectations about what that sandwich is worth to me. Or, if your utility is independent of your expectations about me, then utility is maximized when you give me the sandwich even if it has no value to me.

As an aside, expectations are key in a capital model. When the sandwich maker asks you to invest in expanding his sandwich production, you will destroy value because of your false expectations. But I'm guessing that you're still okay with the capital model even though it can destroy value when investors have false expectations. :-)

But here's the kicker: false expectations also cause value to be destroyed in the market model. If I have a contagious disease and you _don't_ know it, your trade with me may destroy a whole lot of value. You can call that an externality if you want... but if you do, I'll respond that your lack of knowledge about the value of the sandwich to me is also an externality in the gift/charity context.

Utility is a great theoretical construct in the absence of error: the reason he acted was because the action had net value. But utility arguments can never suggest that a voluntary action (including a gift) destroyed utility.

So my claim is that effective giving, as well as effective trading and investment, requires a ballpark understanding of the (subjective) utility that others will attach to an action.

To further that understanding, I'd like to point out some of the costs involved in accepting a gift. Accepting a gift usually involves time, sometimes a lot of time. It often involves some kind of emotional attachment or obligation. In real life, accepting a gift always requires knowledge, ranging from "knowing the gift is there" to "knowing how to unwrap it" to "knowing how to make it fry your bacon" and "knowing how to put out the fire when the bacon fat overflows the skillet."

So before you give me that leftover fruitcake that's taking up space on your shelf, spend some time getting to know me, discovering what I like, what's important to me, and whether the lack of fruitcake in my house is because it's an unattainable luxury or because I gag on it.

That's a nontrivial task. It doesn't happen automatically. But neither does the research that enables the sandwich seller to set his price, or you to set yours. The deeper question is, which research do people _want_ to do?

kerspoon wrote at Monday, January 26, 2009 at 5:06:00 PM PST

I like the idea behind what you are saying. It seems to make logical sence to me. It's not quite what you are looking for but the following may be of interest to you:

http://www.kiva.org/app.php?page=businesses

Ryan Sitch wrote at Wednesday, June 3, 2009 at 3:50:00 PM PDT

I agree very much with your view of charities, and that they really need to be doing something, not just making a show. I try to give to local charities, that are on the spot, like Christmas Cheer. Where you can actually buy food, take it down with your reciept and get a tax donation slip, and watch your food go directly into a hamper which goes to a family the following week.

Anonymous wrote at Wednesday, October 28, 2009 at 3:14:00 AM PDT

check out the "parable of the broken arrow" http://en.wikipedia.org/wiki/Parable_of_the_broken_window

Daniel wrote at Wednesday, November 3, 2010 at 10:49:00 PM PDT

I meant to comment on this a month ago, but I haven't been sick until now. :-)

I think you're playing fast and loose with your definition of utility. In your market model, it's definitional that the voluntary trade only occurs because there's an increase in net utility. The dollar values you give are just stand-ins for utility. For example, if I have a contagious disease and you know it, you'll require more than $2 from me, which has nothing to do with the sandwich and everything to do with making the utility equation work.

But your model for gifts suddenly loses all this sophistication. After the trade, the seller has $2, you have $0, and I have $0. But that's not possible. Since these dollar values are just stand-ins for utility, you have more than $0. In fact, the utility for you of giving me the sandwich must have been at least $2 -- it cannot have been less than $2 or you would have given me $2 instead of buying a sandwich for me and giving me the sandwich.

I think the missing thing that unites both markets and gifts is that we have expectations about what a thing is worth to other people. The sandwich seller will not long believe that his sandwiches are worth $1 if you and everyone else buy them for $2: if I come along and offer him $1.50, he'll be unlikely to trade.

And your assignment of a utility >= 2 to giving me the sandwich is probably dependent on your expectations about what that sandwich is worth to me. Or, if your utility is independent of your expectations about me, then utility is maximized when you give me the sandwich even if it has no value to me.

As an aside, expectations are key in a capital model. When the sandwich maker asks you to invest in expanding his sandwich production, you will destroy value because of your false expectations. But I'm guessing that you're still okay with the capital model even though it can destroy value when investors have false expectations. :-)

But here's the kicker: false expectations also cause value to be destroyed in the market model. If I have a contagious disease and you _don't_ know it, your trade with me may destroy a whole lot of value. You can call that an externality if you want... but if you do, I'll respond that your lack of knowledge about the value of the sandwich to me is also an externality in the gift/charity context.

Utility is a great theoretical construct in the absence of error: the reason he acted was because the action had net value. But utility arguments can never suggest that a voluntary action (including a gift) destroyed utility.

So my claim is that effective giving, as well as effective trading and investment, requires a ballpark understanding of the (subjective) utility that others will attach to an action.

To further that understanding, I'd like to point out some of the costs involved in accepting a gift. Accepting a gift usually involves time, sometimes a lot of time. It often involves some kind of emotional attachment or obligation. In real life, accepting a gift always requires knowledge, ranging from "knowing the gift is there" to "knowing how to unwrap it" to "knowing how to make it fry your bacon" and "knowing how to put out the fire when the bacon fat overflows the skillet."

So before you give me that leftover fruitcake that's taking up space on your shelf, spend some time getting to know me, discovering what I like, what's important to me, and whether the lack of fruitcake in my house is because it's an unattainable luxury or because I gag on it.

That's a nontrivial task. It doesn't happen automatically. But neither does the research that enables the sandwich seller to set his price, or you to set yours. The deeper question is, which research do people _want_ to do?