An ad hoc panel of the National Academies of Sciences, Engineering, and Medicine will review measurement issues in the Bureau of Labor Statistics (BLS) Consumer Price Index program, and provide guidance for its continued modernization. The study will examine the potential to improve CPI methodology by incrementally transitioning from traditional survey-based data collection to an approach that blends multiple (survey and non-survey, government and commercial) data sources. The panel will consider opportunities to apply new data sources to improve the construction of specific elementary item-area indexes as well as to improve index aggregation along several dimensions.
Many data sources have emerged during the past 20 years (since the last National Academies' review of the CPI), that could be used in the construction of the 7,000+ elementary item-area indexes in a way that improves the accuracy, timeliness, and detail of resulting price statistics, or reduces costs in the CPI program. The panel will identify specific areas where new kinds of data may be harnessed in a relatively straightforward way to improve price measurement of some items such as food and electronics. The panel will also propose solutions for some historically difficult-to-measure expenditure categories, particularly for which the availability of alternative data create opportunities for improved price measurement.
The panel will consider opportunities to use new data sources to improve aggregation of the elementary item-area indexes and also to mitigate upper-level substitution bias in the CPI-U and the CPI-W—for example, by taking advantage of the simultaneous availability of quantity and price information to update baskets and weights with shorter lags. As part of this task, the panel may revisit concerns about data sources used to estimate population item expenditure weights. The panel will also assess the prospects for creating new index aggregates that would present information about prices paid and expenditure weights for goods and services by households across the income distribution (by decile, or perhaps by quintile).
Finally, the panel will offer forward-looking thoughts about what price measurement may look like in 20 years and what BLS can do to anticipate future research and policy needs. As part of its information-gathering activities, the panel will gather input from data users, stakeholders, and survey experts. The panel will produce a consensus report with conclusions and recommendations.