Spotlight: Diego Perez

The Value of Economic Statistics

Assistant Professor Diego Perez

Diego Perez

Every month, the Bureau of Labor Statistics and the Bureau of Economic Analysis publishes statistics on inflation, employment and several other economic variables. Providing macroeconomic statistics entails an economic cost associated with gathering and processing a large amount of data. At the same time, the public provision of economically relevant information is likely to confer certain benefits on society. After all, the availability of timely and accurate economic data is a common feature of any country with solid institutions. In this column I focus on the benefits associated with the provision of macroeconomic statistics. A proper understanding and measurement of these benefits is key to determining adequately how many resources need to be invested in the provision of public information. In a recent paper, Price Setting Under Uncertainty About Inflation, joint with Andres Drenik, we quantify the economic effects of providing precise statistics about the aggregate inflation rate. The provision of accurate inflation statistics enhances the availability of information that firms have when setting prices. Using an economic model we show that more precise information is associated with a more efficient allocation of resources in the economy, which in turn increases social welfare. We also find that the increase in welfare is greater in those economies with more economic volatility and less predictable monetary policies. 

The Economic Effects of the Provision of Statistics

How does the availability of inflation statistics affect the macro economy? The answer revolves around the information firms must have to efficiently determine the prices they should set for goods and services. In making such decisions, firms must use both idiosyncratic information (information specific to their own revenues and costs) as well as aggregate information (information about economy-wide level of prices and employment). Precise statistics about inflation allow firms to disentangle those components of the demand for their products that are idiosyncratic to them and those that are aggregate and common to all firms. For example, in a high inflation environment, a high level of sales is more likely to be attributable to aggregate factors such as an expansionary monetary policy: if there are more dollars circulating, the families will demand more of all goods in the economy. On the other hand, in a low inflation environment, a high level of sales is more likely to be attributable to a higher real demand for the goods that that the firm produces. Therefore, more precise information about inflation helps firms disentangle idiosyncratic from aggregate shocks, as well as monetary from real shocks. With better information regarding the demand for their goods, firms can set prices in a way that better reflects both the production cost and the value assigned by buyers. When prices more accurately reflect the value of demand and the cost of production, inputs of production (labor, physical capital, intermediate inputs) are assigned in a more efficient way. If input factors are more efficiently assigned, the aggregate level of production given the same amount of inputs is higher, which leads to higher social welfare. A symptom of a better access to information about inflation is a lower level of price dispersion for a given type of good. When more precise inflation statistics are available, the price of a cup of coffee should more similar between different stores and brands. The reason is that, with more accurate information about aggregate inflation, when setting prices firms optimally choose to put more weight in inflation statistics, which are common to all firms and hence their prices are more aligned with each other. Once the mechanism through which the availability of public information can affect the macro economy is understood, the next challenge is to quantify it. We can measure this effect if we have: i). an economic model that fits the macroeconomic behavior of an economy and ii). an episode of analysis in which the access to information about the inflation rate changed significantly. The second condition is more difficult to find in the real world: is there an episode with significant changes in the availability of information about the inflation rate? There is, in Argentina.

A Natural Experiment: The Manipulation of Inflation Statistics in Argentina

In 2007 the Argentinean government started manipulating official inflation statistics to prevent figures from reflecting accelerating inflation. The manipulation started in January 2007 with the government’s intervention of the national statistical agency (INDEC) that included the removal of the authorities in charge of computing and publishing the CPI. Since then, local and international media, international institutions and academic circles discredited official inflation statistics.[1] In the paper, we argue that in the past decade Argentina experienced two regimes concerning the access to public information about the level of inflation. Before 2007 the government provided a unique and credible official measure of inflation. From 2007 onwards, official inflation statistics were discredited and there was overall uncertainty about the true level of inflation, in spite of the presence of alternative noisier measures of inflation. Using data on disaggregated prices for Argentina and Uruguay, we found that the manipulation of inflation statistics brought about an increase in price dispersion (measured by the coefficient of variation of prices of similar goods) in Argentina of 13%. With this measured effect in hand, we can estimate an economic model for Argentina tailored to study our episode of analysis and then use it to quantify the value of providing trustworthy inflation statistics. According to our model estimates, the manipulation of inflation statistics lead to a less efficient allocation of inputs of production, that in turn lead to an equivalent permanent drop in the aggregate level of output of 1.3%. With less precise information, the firms that generate the highest value added do not hire more labor or capital. We then explore to what extent are our estimates can be generalized to the US. To do so, we re-estimate our model to match the most salient features of the US economy and then re-do our hypothetical exercise that replicates the manipulation of inflation statistics. We find that the negative welfare effects associated with this policy are more than ten times smaller than those for Argentina. The reason is that the availability of inflation statistics provides less value in the US, since aggregate prices do not fluctuate as much as in Argentina.[2] In other words, the US would not incur such large efficiency losses if firms set their prices by taking into account the long run level of inflation rather than by considering the current level of inflation. In summary, we find that significant welfare gains can be made by providing trustworthy and precise statistical information about the inflation rate, as it helps to allocate resources in a more efficient way. Additionally, these welfare gains are larger in highly volatile economies where prices fluctuate significantly and firms place greater value on information about the aggregate macroeconomic state. The Argentinean episode is ideally suited to measuring these welfare consequences and quantifying their magnitudes.
[1] See, for example, “Acusan al Gobierno de manipular datos del INDEC”, La Nación, February 10, 2007, and “Lies and Argentine Statistics”, The Economist, April 2011. [2] This analysis ignores any potential feedback that the provision of adequate statistics may have on the volatility of monetary policy, which can be a relevant channel in the case of the US.