Since the group of super users included leading investment banks and hedge funds, it raised suspicions about the agency sharing market-sensitive information with a closed group. The agency claimed that it did not maintain any list of super users, and that the mail sent by one of its officials was an isolated ‘mistake’.
A subsequent investigation by Ben Casselman and Jeanna Smialek of The New York Times (‘New questions on how a key agency shared inflation data’, 5 April) showed that the official concerned did maintain a list of super users, and had regular email correspondence with them. The official did not share any data in advance.
He simply provided contextual and methodological clarifications on the US retail inflation gauge. “Such details, though highly technical, can be of significant interest to forecasters, who compete to predict inflation figures to hundredths of a percentage point,” wrote Casselman and Smialek. “Those estimates, in turn, are used by investors making bets on the huge batches of securities that are tied to inflation or interest rates.”
Since professional forecasters and even the Federal Reserve have been wrong on the inflation trajectory in the US in recent years, the stakes in getting inflation forecasts right have increased significantly. This seems to have driven up the volume of questions faced by the American bureau of labour statistics.
The unnamed official who created the super user mailing list presumably did so to help such analysts figure out the data better. However, by creating a closed and secret group, he managed to raise suspicions about whether the agency was providing an undue advantage to a select group of investors.
The episode highlights the importance of having standard communication protocols. Statisticians must aim to present as many details as feasible—and in as clear a manner as possible—in their data releases. But often, that won’t be enough for data users. They may have follow-up questions that can only be answered by data producers.
A modern statistical agency needs to institutionalize a mechanism for responding to such questions. One way forward is to follow the example of central banks that publish responses to individual queries on their websites. So, even if a hedge fund analyst raises a query, that analyst gets the information at the same time as the wider public.
The need for such a mechanism is higher in emerging markets such as India where official databases are riddled with gaps and inconsistencies, and where statisticians routinely make use of heroic assumptions to fill in database gaps.
Over the past few years, India’s official statisticians had clammed up in the face of growing criticisms. In a data user conference on national accounts last year, the press was warned not to identify any official in their reports.
No cameras were allowed. As this column had argued earlier (‘Official statisticians must learn to communicate better’, 25 September 2023), the reticence of the ministry of statistics and programme implementation (Mospi) was a defeatist response to the criticisms that were directed at it.
Mospi’s mandarins seem to have finally realized that a culture of secrecy lowers trust in official statistics. Last week, Mospi live-streamed a data users conference on the household consumption survey for 2022-23. This is a welcome departure from the recent past. But it is not enough. Mospi needs to publish a schedule of all its ongoing surveys and forthcoming data releases.
At the moment, such a schedule is maintained only for a few statistical products such as the inflation index and gross domestic product (GDP) figures. After each release, Mospi must open a window for data users to post queries online, and address those questions in an open conference.
If a listed company can be expected to meet basic transparency norms in its dealings with shareholders, there is no reason why citizens in a democracy should not expect the same standards from a statistical agency funded by them. If Mospi becomes more responsive to Indian citizens, it will also be addressing the needs of non-citizens interested in India and the Indian economy.
Some of them may then find it easier to invest in the Indian economy. Statistical reforms can lower the cost of doing business while improving the quality of public discourse. This would also help India differentiate itself from China, where questioning official statistics is actively discouraged.
Some statistical reforms will need legislative changes and wide stakeholder consultation (‘Statistical reforms will not get far without a consensus’, Mint, 11 June 2024). However, there is a lot of scope to improve Mospi’s functioning even while that process unfolds.
It only takes executive action to improve communication protocols, or to establish a code of practice for official statisticians. Such protocols will help Mospi respond effectively to the needs and queries of data users within and outside the government. A responsive statistics ministry will also find it easier to ask for bigger budgets and push for more substantive structural reforms.