IN THIS week’s print edition, we explain why we have decided to drop Argentina’s official inflation statistics and publish a private-sector estimate, State Streets’ PriceStats Index, instead. The PriceStats method involves an automated daily trawl of huge numbers of internet prices oakley sunglasses, instead of the traditional government approach of identifying a representative basket of goods and then sending dozens of mystery shoppers out to buy those things monthly. It was dreamed up by an Argentine, Alberto Cavallo, who set up a website, Truth in Argentine Statistics , and did the research needed to validate the method during his studies at Harvard University. You can find out more about the theory at the Billion Prices Project, a collaboration between Mr Cavallo and Roberto Rigobon, another Harvard economist an economist at MIT Sloan School of Management. The two set up PriceStats to commercialise the idea, and now the company produces daily inflation information for 19 countries, which are available from State Street, an investment bank a financial services firm.

To sum up, supermarket prices turn out to provide an excellent match for government inflation figures, even though they can only substitute directly for around two-thirds of what appears in the official basket of goods—everywhere except Argentina. The following chart shows the difference between two inflation rates—PriceStats’ and the official one—for a range of countries. In Argentina, the difference is more than 13 percentage points. The second-highest difference found among the 19 countries PriceStats covers is also shown: it is in Venezuela, and comes in at just over one percentage point. Generally, online prices turn out to move a bit earlier than official data (this is in fact PriceStats’ selling point). But those differences do not propagate over time; offline prices catch up with online prices, rather than diverging.

The close fit between the two series everywhere else suggests that none of the innocent explanations for the big difference in Argentina holds up. You might wonder whether PriceStats is only measuring inflation as experienced by rich people, since poor ones do not shop online, or whether the wide gap is somehow a phenomenon of high-inequality countries, or ones with a history of high inflation. But Brazil and Venezuela tick all those boxes too, and PriceStats’ figures match their governments’ figures extremely well.

Argentina is not the only country where people complain that official inflation figures do not match their own experiences. A single basket of goods simply cannot fit all pockets and lifestyles: poorer people tend to experience higher inflation when food prices spike, because they spend a bigger share of their income on food; pensioners suffer disproportionately when fuel prices soar foakleys, but are usually protected from big hikes in travel costs. But again, the differences tend to fade over time. Sometimes, though, they lead to a more general suspicion that official statisticians are succumbing to political pressure to report a rate lower than reality. But what PriceStats have managed to show is that official statisticians generally do a stellar job—except in Argentina.

From the end of this month, PriceStats will start publishing its data for both Argentina and the United States on their website, with only a short lag. We’ll certainly be keeping a weather eye out for any sign that Argentina’s official statisticians are raising their game.