CPI measurement problems dwarfed by other concerns - Capital Economics
Global Economics

CPI measurement problems dwarfed by other concerns

Global Economics Update
Written by Gabriella Dickens

The impact of the lockdowns on the ability of statisticians to collect price data means that measures of inflation will be less accurate than normal. However, this will probably have little impact on economic policy because these problems are dwarfed by concerns about the scale of the economic slump.

  • The impact of the lockdowns on the ability of statisticians to collect price data means that measures of inflation will be less accurate than normal. However, this will probably have little impact on economic policy because these problems are dwarfed by concerns about the scale of the economic slump.
  • Not only has the coronavirus plunged the economy into turmoil, but it has thrown up a number of challenges for statistics offices when it comes to measuring the damage. For consumer price indices, which are already notoriously difficult to produce, the shutdowns mean much of the data are simply unavailable. And even where businesses are still trading, surveyors are often unable to collect the data in person, leaving potentially up to half of the CPI data missing in most advanced economies. So, in the short term at least, it is perhaps fair to say that we are in a period of “no-flation”.
  • What’s more, even if the data were available and could be collected effectively, the lockdowns have drastically altered households’ spending patterns. So, the CPI basket may no longer be representative of average household expenditure.
  • Admittedly, price data for most advanced economies appeared relatively unaffected in March. In the US, for example, three-quarters of respondents still replied to the survey, with around 40% of that number collected in person. But while the number of respondents held up fairly well at the aggregate level, certain categories saw the number plunge, including new vehicles and certain hospitality sectors. What’s more, with the lockdowns taking full force towards the end of March, the CPI indices will probably become less reliable in the months ahead.
  • Of course, statistics offices will find some of these challenges easier to tackle than others. For instance, surveyors will be able to fill in some gaps using online sources. But it becomes more of a challenge where data are not available at all due to the shutdowns.
  • Methods to “fill in the blanks” do already exist and, in normal times, are fairly effective. The most common is imputation where similar products are used to estimate price changes for the actual item included in the CPI basket. For example, with international flights suspended between China and New Zealand in February, New Zealand’s statistics office estimated the price change using all other flights over the same period.
  • But these are not normal times. And with almost 50% of most CPI baskets severely affected by the shutdowns in DMs, a sizeable portion of the data will have to be estimated. For instance, French inflation data for April showed that roughly 45% of the price data were imputed.
  • What’s more, with whole sectors ceasing trading, substitute products will be harder to find so some prices will have to be “carried forward” meaning that the data from the previous month are used. While this will probably be a somewhat realistic scenario in certain areas, including bars where the prices will be unchanged when they reopen, for others, changes in the true price just won’t be reflected. It stands to reason then that price indices are unlikely to accurately reflect inflation during the lockdowns.
  • Inaccurate price data could have two important implications. First, if statistics offices face similar problems when constructing their deflators, inaccuracies could be introduced in a whole host of national accounts and household finances data. Second and most obviously, it will make it difficult for policymakers to judge where inflation sits relative to central bank targets.
  • However, given the exceptional circumstances, even if the CPI data are way off the mark, any implications for policy will most likely be minimal. After all, monetary policymakers target inflation over the medium term. So, as long as the strict lockdowns end soon, inaccuracies in the CPI data won’t seriously inhibit policymakers’ ability to gauge inflation further ahead. And of course, for now at least, the immediate concern for policymakers is stabilizing output given the scale of the economic disruption.
  • When lockdown measures start to ease, though, we expect weaker demand will more than offset any shortages in supply causing inflation to edge lower. (See our latest Inflation Watch.)

Gabriella Dickens, Assistant Economist, gabriella.dickens@capitaleconomics.com