Development activist Deborah Doane said yesterday on comment is free that Goldman Sachs should admit it drives food prices up through speculation. She excoriates the finance giant for what she sees as its role in 44m across the world being in food poverty, suggesting its charitable efforts amount to little given its speculative activities. She assembles an array of sources purportedly supporting her case.

But the way Doane does this is perplexing. She implies in her first paragraph that the World Bank agrees with her conclusion, but the link she includes mentions only a UN official's opinion, making no reference to the WB. And while, apparently, more than 100 studies support her argument, her article provides no evidence this is the case. Moreover she makes no attempt to deal with the basic economic argument against her thesis.

On the one hand the most limited version of Doane's thesis—that speculation increases prices—is undeniable. When speculators buy into the market, they raise the price then. But the overall case makes little economic sense. If speculators' influence is big enough to boost prices when they buy in, it is big enough to cut prices when they sell out. That is, speculators both add to, and take away from, prices.

A speculator makes money by buying in times of relative plenty, when prices are low, and selling in time of relative scarcity. For helping society ration effectively—making sure the differing scarceness of a good is reflected in its price, thereby improving individual decision-making—the firms earn a return. If a speculator, by contrast, buys in at the top of the market, reducing supply when it is most needed, and sells at the bottom, when it is least needed (relatively) they lose money. This is how the profit and loss system, in a good institutional structure, encourages and rewards socially-minded behaviour. And speculation should smooth volatility in markets. A jump in price will encourage sales from speculators, bringing the price back down. A dip in price will encourage speculators to buy, bring the price back up. This result dates back to a 1953 paper from Milton Friedman, which is hard to find online, despite being cited 2411 times according to google scholar.

Having said all this, it's possible there are some circumstances where this simple common sense argument fails to obtain. A widely-cited 1990 paper by Andrei Shleifer, Larry Summers, Brad DeLong and Robert Waldmann finds that the way other traders buy and sell can change the way speculation works. If so-called noise traders' buy, rather than sell, when prices rises hit, i.e. their strategies include elements of positive feedback, early buying from speculation can trigger a spike. Anticipating this, speculators will have an incentive to buy extra, generating a bigger price increase and a bigger return, along with more volatility in the face of new information.

The authors look at some hugely interesting survey, experimental and real world evidence supporting the assumption that some traders might use positive-feedback strategies, and even argue that the strategy could persist in the long-run, despite its irrationality. Though on average those pursuing the strategy will be driven out of the market as they lose money, investors see different episodes differently, and some will enjoy huge returns through holding extra risk. So it is actually rational for speculators to target price movements driven by others' irrationality, rather than market fundamentals.

All this said, it's unclear this model delivers the results Doane relies on. Certainly markets working in this way would produce bigger spikes in both directions, but there's no reason to expect prices would be higher overall. If anything, the suggestion seems to be that prices would be below where fundamentals suggest in times of scarcity and above in times of relative plenty—alleviating shortages more than under the more straightforward model. The extra profits speculators make would come out of the pocket not of the poor commodity consumers, but from the noise traders, following their irrational strategies.