Tag Archives: incentives

Thinking Velocity Vs. Efficiency

Image representing GlobalGiving as depicted in...
Image via CrunchBase

Global Giving is a deservedly widely-hailed institution, highlighted by such luminaries as the Clinton Foundation and Bill Easterly. Its model is relatively simple. Philanthropists-to-be browse the website, and look for a (prescreened) cause they want to donate to; the causes are posted with amounts needed, where and what the money will be spent on, and a reassurance that you’ll get regular updates. And a couple guarantees – notably, that your money will be “on-the-ground” in 60 days. Now, we might expect that getting spent quickly is a good idea; after all there is a fear that the money will be retained by those getting it for personal gain, or that it’ll be whittled away on bureaucratic red tape. And some pressure must surely be put on organizations to avoid waste, undue delay or excessive red tape.

However the constant pressure to get money out the door, let alone  an absolute requirement, can lead to dangerous incentives as Good Intentions are Not Enough illustrates:

I worked for a large international charity after a major disaster. As part of the senior management team I attended monthly meetings to discuss our “burn rate”. The burn rate referred to how quickly we spent… our money. There was no similar meeting on the quality or impact of our aid….[We wanted to] ensure that all money was spent by the five year anniversary of the disaster. My organization knew that if there were any unspent donations at that time they would get a lot of negative publicity. If the public heard that they hadn’t spent all of the money donated for that disaster they would be less willing to donate to the next disaster. Spending money quickly was paramount. Unfortunately this caused a lot of problems in the field for our partner organization…Our staff was caught in the middle between the very real needs of the partner organization and the demands of HQ. A large part of the funding decisions were based on the ability of the project to burn through money rather than on the quality of the program or what type of assistance was needed the most.

I suggest you read the whole thing: the real problem of the constant push to get money out the door irrespective of quality, need, efficacy etcetera is demonstrated in the details. The incentive to burn money is also amply demonstrated by a conversation one of Good Intention’s are Not Enough’s commentators had with their organization: “[I said] ‘But if that’s the best way of doing it, why don’t we do that everywhere?’ ‘Because it takes too long and it doesn’t cost enough.’” We usually think of corporations as doing things quick, dirty and cheap. Well it seems that charities often are incentivized to do things quick, dirty and expensive.

So while I think Easterly is right to hail GlobalGiving for its concentration on solving small problems in doses, rather than on the next panacea, I think we should recognize that velocity should not take precedence over efficacy.

Again, I think that GlobalGiving does do a wonderful job and that its model is something to consider very strongly over the usual scaled international aid. We should just think very seriously about what we push aid to do.

Leave a comment

Filed under Organizational Models

A Hayekian Challenge For Charities

Debris in the streets of the Port-au-Prince ne...
Image via Wikipedia

One of the most famous works by the economist Friedrich Hayek is his short essay, The Use of Knowledge in Society. In it he explains that an economy exists not merely to solve a known problem of scarcity, slotting x number of pegs into y number of holes, but rather as an enormously complex machine for transmitting knowledge from distant individuals. The problem that any economy exists to solve, then, is created because

the knowledge of the circumstances of which we must make use never exists in concentrated or integrated form but solely as the dispersed bits of incomplete and frequently contradictory knowledge which all the separate individuals possess…it is a problem of the utilization of knowledge which is not given to anyone in its totality.

This problem, of course, exists in all economies. But, as this recent disaster in Haiti has shown, the dangers of complexity are hardly restricted to the private sector:

The earthquake in Haiti sparked one of the biggest international aid efforts the world has seen, but the sheer number of charities involved caused problems with communication and coordination…With dozens of charities operating all over Haiti, there was a risk of their work overlapping. The UN operated a cluster system based around areas of need – for example, water or shelter – so that charities could share information, collaborate and set common standards on, say, the cost of materials. But the clusters have brought their own problems.

Hayek’s solution to this problem for economies is the price system, a set of numbers continuously in flux and decided by millions if not billions of bidders, changes in demand, the ease of manufacture and many other factors. But while this may work for the incentivized capitalists of the private sector, the more convoluted incentives and operations of the public sector make this a difficult trick to pull off. So, what are some practical ways to make charities do a better job coordinating?

1. Have the UN devote more resources to making communication efficient, and inclusive. The common supposition that low overhead is a good sign when choosing charities is misplaced; often the reverse it true. So invest in efficient communication and incentivize participation – say, give $5000 to charities over a certain size which join.

2. Learn from some similar examples. For instance, the US military currently has a problem coordinating drone strikes: there’s just too much data. But they’re coping with this using some relatively simple social networking tools and wikis. Now, I don’t think anyone is of the opinion that the UN is as efficient an organization as the US military – but they could certainly pivot off these ideas by trying to improve coordination with these technologies.

3. Look to technologies down the line. This is definitely a little utopian, but with enough information flowing in, organizations may in the future be able to take advantage of IBM’s question-answering computer. This machine, known as Watson, is good enough to go toe-to-toe with Jeopardy! greats, and is being considered for use in extreme scenarios like answering split-second medical questions in an emergency room. Why not have it collate information during disaster relief? For instance, if one group is currently working on freeing earthquake victims in one sector of Port-au-Prince, they could let Watson know with a GPS and time stamp, along with a status line. Then if another group is looking for the best place to find people, Watson could at least narrow the field. Again, this solution is fraught with problems. But it might be the best of a bad set of options.

Now, I realize the irony of these answers – they’re totally anti-Hayek. They’re centralized, bureaucratic, and try to deal with the problem of knowledge by consolidating it in one place. I just think streamlining a centralized database may be our only choice. But, just in case, here are two answers Hayek may have liked better:

4. Create cash prizes to distribute for verifiable goals. As with many free market ideas, many people will strongly dislike this one. It seems to make mercenaries out of altruists, and it may indeed open the door for fraud. Still, creating incentives for verifiable goals like people with access to potable water, vaccines given, people in acceptable housing, etc. may be extremely useful to push aid organizations on. This has numerous dangers like those mentioned above as well as possibly leaving non-incentivized goals under-pursued. Nevertheless, on a limited scale this may drive organizations to better work.

5. Create prediction markets to establish answers to pertinent aid questions. When will those five wells be dug? When can permanent housing be expected to be finished in this quadrant? Should we fund nutrition packs or local farmers in the coming month? As Robin Hanson argues, prediction markets – forums where parties can place bets on certain outcomes – can be very effective in situations like these, because these are exactly the sort of questions that make interset groups fight for their organization over the greater good. Making the choosers have clear, definable upside if they are correct and downside if they’re wrong makes the choice more trustworthy.

I’m sure there are many more ways to help coordinate in these difficult situations. But by applying some of these lessons we may hope to avoid tragedies like this and move more in this direction, of incremental improvements in very tragic circumstances.

Hat tip: Good Intentions are Not Enough, Matt Yglesias, Marginal Revolution

Leave a comment

Filed under Uncategorized

Bribery and the Tragedy of the Commons… or Rent Seeking as a Problem of the Anti-Commons

Red Square, Moscow, Russia
Image via Wikipedia

We’ve talked before about the problems of corruption, and of nonprofit innovative solutions to the problem – specifically, an Indian group’s spreading of a Zero denomination Rupee to be handed to bribe-asking officials. Recently, corporations are trying to do their part, pledging to boycott bribery:

Dozens of international firms doing business in Russia have pledged not to offer bribes, in a move aimed at fighting corruption collectively. The accord, signed at an official ceremony in Moscow, was initiated by the companies, not the Kremlin, said the Russian-German Chamber of Commerce.

According to the article, about $300 billion is extracted every year in bribes. Of course, this doesn’t count all the inefficiencies introduced into the system by firms being chosen over competitors for reason of bribe rather than superiority.

Now, this does seem to be a nice little pact, and we can all snap some pictures and shake hands and go home. But as far as I can see through a number of news stories there’s nothing enforceable or transparent about this pact; it’s just some words. In fact, the head of Transparency International in Russia admitted:

any company working in Russia could find itself in a situation when it did not want to give bribes but was forced to do so if it wanted to carry on doing business. The agreement would not solve situations such as these, she said.

I believe the main problem of bribery and incentives is one of the commons. The tragedy of the commons is a well known economic problem. Essentially, when a rent producing commodity, such as a lake or the air, is unowned there is little incentive to preserve it because all actors know that their effect is small and their incentives to gain the short term big. This leads to overfishing, air pollution etc. Typically this can be solved by adding property rights; if someone owns the lake they’ll fish at a constant and conservative rate in order to maintain the long term population of fish. Giving property rights works: observe the massive success of the cap and trade system in stopping acid rain.

How does this relate to bribery? Consider corporations as partakers of the government; they are fisherman in the lake of government. As each of them interacts with government, many of them lose money as a result of bribery. But because of the small effect it has on each of them individually – though its effect may be large in the aggregate – it isn’t worth it for them to create enforceable, binding pacts against it. And in fact, some of them gain from the system, as HP seems to have when it essentially bribed itself into a contract. Because there isn’t any ownership of the long run management of the government, rent-seeking firms can manipulate it to their advantage and others have to simply pay to play. Their stake in the game isn’t valuable enough to risk bucking the system.

Alternatively, one could see this as a problem of the anti-commons. These are instances of the opposite problem: over-ownership. As Michael Heller discusses in his new book, when a property is too subdivided it can cause massive gridlock in a system. For example, if a new type of rice is created mixing a couple genes, it may be impossible to bring to market because each of those genes may be separately patented. Apparently, in the creation of online banking over 10,000 patents were involved. Ownership can spur innovation, but over-ownership cuts it down.

This would relate to bribery so: if there were a very small number of firms who were tied tightly into the bribery scheme, it might make sense for all them to collude and boycott bribery. However, as the number of firms explodes it makes sense for those firms with current contracts gained through bribes to avoid binding pacts against bribery because on an individual level it helps those firms using it to maintain their positions. However, in the aggregate this rent-seeking simply harms the economy.

Of course, there is a pact. But due to its lack of transparency or enforceability, it’s hard to see how its rules won’t be bent.

Still: both these explanations, I think, cannot be right. There can’t be too much and too little ownership. So which is it?

Reblog this post [with Zemanta]

1 Comment

Filed under Uncategorized

Charities and Companies: A Tale of Two Incentive Structures

Wall Street taken above steam stack road works.
Image via Wikipedia

The comparison between for profits and not for profits is complicated, and the way we understand their differing incentives is of vital importance.

Companies have a simple goal, simple metrics for understanding success, and clear incentives to produce results. The actual process of making money is very difficult but to understand when a company is doing well all one really needs to do (at least in a mature company) is to look at its books. If it’s making a steady and growing profit, good. If it’s continually losing money it’s clearly failing. Lose enough money and the shareholders will try to do something about it or the business will simply go bankrupt.

Non profits, meanwhile, are messy entities. They have complicated goals – as differing as giving food to the hungry and education to the least educated – and measuring how they’re doing is difficult. Moreover, without shareholders closely monitoring their cashflow their incentives are also strange. The service they provide to donors is ultimately to make them feel like they’re doing good – and that may not be identical with fulfilling their charitable goal to the fullest.

We deal with this mismatch of incentives and outcomes – of getting charities to serve the needs of the needy – by trying to get an informed donor community. As we’ve mentioned numerous times on this blog Give Well, Charity Navigator and Philanthropedia all try to highlight the very best of the charities – and also the very worst – so that donors will start shifting resources to where they’re most helpful. What this should do is bend the incentives towards a high bang for your buck on meeting charity goals.

Now with this in mind, there’s the additional question as how a donor should think when compared with an investor. Because of the weakness of feedback the nonprofit sector does not operate like an efficient market – that is, unlike Wall Street there should be a huge amount of “arbitrage opportunities” in charities. The “price” of a nonprofit (which in this case would be its donations) should not be as informed as a publicly traded stock which allows for bargains. While there are still strong arguments for diversification in donations (to reflect some doubt on our ability to evaluate charities) there is still compelling logic to placing most of our donations in a small number of excellent charities.

Added 4/21/10: My cousin reminds me that nonprofits are not all charities. There are many excellent organizations from religious, educational and entertainment oriented (which may, as this podcast suggets, simply be subsidizing rich people through museums, operas and the like) which qualify as 501(c)(3)s. In these cases too the incentives problem persists. As Tyler Cowen has pointed out, museums are often very visitor unfriendly because being arcane and subtle fits donor preferences for high art. Places that rely more on visitor receipts will respond by being easier to enjoy oneself in.


Reblog this post [with Zemanta]

2 Comments

Filed under Organizational Models