Introducing Morning Consult’s Unemployment Index: The U.S. Edition

Since August 2020, Morning Consult has collected daily data on U.S. consumers’ employment status. This survey uses the same definition–though different methodology–as the Bureau of Labor Statistics (BLS) in determining the national unemployment rate. As a result, Morning Consult’s high-frequency data is useful for measuring real-time fluctuations in joblessness. Having evaluated the robustness of this relationship over time, we are pleased to introduce Morning Consult’s newly established U.S. Unemployment Index.

About the survey

Morning Consult’s means of collecting information on employment status differs from government statistical agencies. Among the most notable differences, the BLS conducts live interviews with a lengthy set of questions from which unemployment classifications are derived, whereas Morning Consult’s employment data is collected through online surveys with a simplified question framework (shown below). However, both methods are designed around identifying job status using the exact same definition of unemployment: per the BLS, “people are classified as unemployed if they do not have a job, have actively looked for work in the prior 4 weeks, and are currently available for work.” On balance, the survey results are closely comparable, but not the same. 

Why an index?

Morning Consult’s calculated share of unemployed U.S. adults overall has been highly correlated with the BLS’s national unemployment rate, registering a .8 historical correlation during BLS reference weeks (monthly data for the Current Population Survey (CPS)  is typically collected during the week encompassing the 12th day of each month). This strong relationship, observed over almost four years, has fortified our confidence that our simplified question framework is in fact picking up a useful signal in terms of employment flows. At the same time, methodological differences contribute to persistent differences leading to a higher unemployment share in Morning Consult’s data compared with the BLS. 

As we publish this analysis, we aim to prevent assumptions of false equivalency between Morning Consult’s unemployment data and the BLS’s unemployment rate. We therefore converted our data to an index to reduce potential confusion that might arise from displaying this data as a percentage share.

The Unemployment Index

In addition to clearly identifying Morning Consult’s data as distinct from the BLS’s unemployment rate, the Unemployment Index is designed with ease of interpretation in mind. The index’s base value of 100 is intended to be consistent with a “neutral” (or, in economic terms, natural)  unemployment rate. This value would equate to the lowest level of joblessness the economy could theoretically sustain without triggering elevated inflation. Estimates in recent years have placed this rate somewhere between 4% and 5% in the U.S.; for our purposes we chose to target the 10-year average BLS unemployment rate over the past 10 years (4.8%).

Since Morning Consult’s data series began in August 2020, the BLS’s unemployment rate came closest to this threshold in September 2021, when the monthly rate was 4.7%. We therefore selected this month as the base period for our index. To calculate the index scores, we re-base our unemployment data so that the average of the September 2021 values are equal to 100; prior and subsequent index values are calculated relative to this underlying value:

The resulting index values are generally consistent with economic views on the trajectory of unemployment in recent years. The Unemployment index history began in late 2020, with above-neutral unemployment levels as the country recovered from the Covid-19 pandemic disruptions. The index then declined to below neutral (100) and predominantly stayed below that threshold through early 2024. In recent months, however, the index has been trending higher as various factors including persistently high interest rates contribute to cooler demand for labor. The index has predominantly remained above 100 since February of this year, and averaged 102.7 over the four weeks ending June 2024. 

Methodological considerations for demographic-level indexes

One important feature of Morning Consult’s data, including the unemployment index, is the rich demographic detail available for analysis. To allow for cross-demographic comparisons in the U.S., all demographics’ unemployment indexes are calculated relative to a common base value–namely, the unemployment share for all U.S. adults identified as the “neutral” level in September 2021. Consequently, demographic indexes may have scores higher or lower than 100 during the official index base period, as these values are all calculated relative to U.S. adults overall.

Characteristics of different groups naturally lead to persistent gaps in unemployment levels, and our chosen approach to calculating demographic-level indexes is intended to preserve our ability to compare these differences. For instance, younger adults transitioning into the workforce for the first time tend to have higher rates of unemployment as they have less work experience (and, if living with parents, less motivating financial pressure) and may take longer than a mid-career adult supporting a family to become and remain employed. Therefore, it is important to remember that the “neutral” unemployment rate for various subgroups is not necessarily equal to 100. Rather, interpretation of demographic index scores should always be viewed as relative to the “neutral rate” of unemployment for US adults overall.