Do more-durable public goods select for or against cheating?
"I don't think taxpayers should have to pay for the bailout; let the government pay for it!" -- An irate taxpayer, interviewed after an earlier corporate bailoutThis week's paper is "Molecular and regulatory properties of a public good shape the evolution of cooperation", recently published by Rolf Kummerli and Sam Brown in the Proceedings of the National Academy of Sciences. The full paper is freely available on-line, making it an example of a public good.
Public goods are benefits that are widely available, rather than available only to those that produce them. A school, a bridge, or a medical treatment developed in a government lab may be used by someone who didn't contribute to its cost, such as a child, a tourist, or a wealthy tax-cheat. All three of these examples are durable, continuing to providing benefits for many years, although the school won't be much use unless we continue paying teachers. How does the durability of public goods affect the prevalence of "free-riders", who take the benefit without contributing to the cost?
This week's paper discusses a public good used by bacteria, namely, pyoverdin, which picks up scarce iron from the environment and releases it to iron-hungry bacteria. Iron is often scarce inside the hosts infected by these bacteria, because hosts actively limit iron levels to prevent infection. Starving humans may have low enough iron levels that we shut down the iron-reducing system, but then relief supplies arrive, including iron pills....
Pyoverdin is a public good because it can be used by bacterial "cheaters" that don't release any of it into the environment. Making pyoverdin has a cost, so you might expect "cheaters" to proliferate. That does tend to happen, but it depends on how durable pyoverdin is (that is, how many times each molecule can transfer iron before it breaks down) and on the production options available to the bacteria.
Bacterial neighborhoods where everyone makes pyoverdin tend to get invaded by cheats that never make pyoverdin. If pyoverdin breaks down quickly, though, those neighborhoods fall apart, killing off the cheats, while more-cooperative neighborhoods survive longer. So durability undermines cooperation. ("We already have a bridge -- no more taxes!")
But things are different if the "cooperators" make pyoverdin only when it's in short supply. Then the cost of cooperation is lower, so cheaters don't invade as easily. ("Let's inspect the bridge yearly, but make costly repairs only if needed.")
The "results" just described are actually predictions from their computer model. But Kummerli and Brown also looked at real bacteria that differ in pyoverdin production. Although different pyoverdins differ in durability, they simulated increased durability mainly by adding pyoverdin. Consistent with the computer model, cheaters that never make pyoverdin out-competed cooperators, but only when pyoverdin levels were low. With more "durable" bioverdin, cooperators stopped making bioverdin, eliminating their competitive disadvantage. So increasing the durability of public goods can actually favor cooperation, so long as cooperators don't over-invest in public goods.
A large fraction of fundamental scientific research is currently supported as a public good. When people started dying in large numbers from AIDS, we didn't know what caused it. Some people thought it was caused by "being gay." In theory, individual drug companies could all have tried to find the cause and kept their results secret, each hoping to gain a monopoly on effective AIDS treatments. But probably no company would have invested enough money to make much progress, knowing that another company might be doing similar work and further along. So the task of finding the cause fell mainly to government researchers and government-funded university researchers, who then published their results. Drug companies then stepped in and used that information to develop a variety of effective treatments.
So scientific knowledge is mainly a public good and a relatively durable one. Today's technology is based on scientific knowledge generated over many decades. How much do we want to invest today in generating new knowledge, only some of which will be used immediately?