Jeremy J. Nalewaik

Estimating Probabilities of Recession in Real Time Using GDP and GDI

This work estimates Markov switching models on real-time data and shows that the growth rate of gross domestic income (GDI), deflated by the gross domestic product (GDP) deflator, has done a better job recognizing the start of recessions than has the growth rate of real GDP. This result suggests that placing an increased focus on GDI may be useful in assessing the current state of the economy. In addition, the paper shows that the definition of a low-growth phase in the Markov switching models changed considerably from 1978 to 2005. The models increasingly came to define this phase as an extended period of around zero rather than negative growth, diverging somewhat from the traditional definition of a recession.

The Income- and Expenditure-Side Estimates of U.S. Output Growth

The two official measures of U.S. economic output, gross domestic product (GDP) and gross domestic income (GDI), have shown markedly different business cycle fluctuations over the past 25 years, with GDI showing a more pronounced cycle than GDP. This paper reports a broad range of results that indicate that GDI better reflects the business cycle fluctuations in true output growth. Results on revisions to the estimates, and correlations with numerous other cyclically sensitive variables, are particularly favorable to GDI. The most recent GDI data show the 2007-09 downturn to have been considerably worse than is reflected in GDP.

Improving GDP Measurement: A Measurement-Error Perspective

We provide a new measure of historical U.S. GDP growth, obtained by applying optimal signal-extraction techniques to the noisy expenditure-side and income-side GDP estimates. The quarter-by-quarter values of our new measure often differ noticeably from those of the traditional measures. Its dynamic properties differ as well, indicating that the persistence of aggregate output dynamics is stronger than previously thought.

Forecasting Recessions Using Stall Speeds

This paper presents evidence that the economic stall speed concept has some empirical content, and can be moderately useful in forecasting recessions. Specifically, output tends to transition to a slow-growth phase at the end of expansions before falling into a recession, and the paper designs Markov-switching models that behave in that way. While the switching models using output growth alone produce a considerable number of false positive recession signals, adding the slope of the yield curve, the percent change in housing starts, and the change in the unemployment rate to the model reduces false positives and improves recession forecasting. The switching model is particularly good at forecasting at long horizons, outperforming Blue Chip consensus forecasts.

Incorporating vintage differences and forecasts into Markov switching models

This paper incorporates vintage differences and forecasts into the Markov switching models described by Hamilton (1994). The vintage differences and forecasts induce parameter breaks close to the end of the sample, too close for standard maximum likelihood techniques to produce precise parameter estimates. A supplementary procedure estimates the statistical properties of the end-of-sample observations that behave differently from the rest, allowing inferred probabilities to reflect the breaks. Empirical results using real-time data show that these techniques improve the ability of a Markov switching model based on GDP and GDI to recognize the start of the 2001 recession.

Current consumption and future income growth: Synthetic panel evidence

Abstract Using group means computed from 20 years of high quality survey data, I show a strong and robust relation between households’ consumption growth and subsequent realizations of their income growth, including realizations as distant as six years later. The relation appears in multiple types of variation in income growth: in variation across cohort-education groups, in variation over the life cycle, and in variation over the business cycle. While other explanations are explored, the results are likely due to forward-looking households altering their current consumption in response to information they receive about their income years into the future, information that turns out to be accurate.

Improving GDP measurement: A measurement-error perspective

We provide a new measure of historical U.S. GDP growth, obtained by applying optimal signal-extraction techniques to the noisy expenditure-side and income-side GDP estimates. The quarter-by-quarter values of our new measure often differ noticeably from those of the traditional measures. Its dynamic properties differ as well, indicating that the persistence of aggregate output dynamics is stronger than previously thought.

Missing Variation in the Great Moderation: Lack of Signal Error and OLS Regression

This paper studies measurement errors that subtract signal from true variables of interest, labeled lack of signal errors (LoSE). The effect on OLS regression of LoSE is opposite the conventional wisdom about classical measurement errors, with LoSE in the dependent variable, not the explanatory variables, causing attenuation bias under some conditions. The paper provides evidence of LoSE in US GDP growth during the period known as the Great Moderation (roughly the mid-1980s to the mid-2000s), illustrating attenuation bias in regressions of GDP growth on asset prices. These biases may have contributed to conventional macroeconomic analysis missing the severity of the adverse shocks hitting the economy in the Great Recession.

The Response of Capital Goods Shipments to Demand Over the Business Cycle

We study how producers of capital goods set shipments in response to fluctuations in new orders. We find that shipments respond more to orders when new orders fall below a certain level relative to shipments, usually after orders plunge in recessions. This cyclical change in producers’ behavior accounts for a considerable portion of the downturn in equipment investment in the 2001 and 2008–9 recessions. A simple model of production to order suggests that heightened persistence in new orders growth may explain the greater responsiveness of shipments, as may increases in the producers’ target delivery lag.

The Income- and Expenditure-Side Estimates of U.S. Output Growth—: An Update to 2011Q2

In light of recent large revisions to the official measures of U.S. output, this update reviews the evidence in my 2010 Brookings Paper showing that the income-side estimate of output (currently called gross domestic income, or GDI) likely captures business cycle fluctuations in true output better than its better-known expenditure-side counterpart (called gross domestic product, or GDP). Most notably, over the 2007-09 downturn, the revisions moved the expenditure-side estimates closer to the income-side estimates, which showed that the downturn was considerably worse than reported initially by the expenditure-side estimates. The tendency for the expenditure-side estimates to be revised toward the income-side estimates is clearer now, as is a tendency for the smoothed income-side estimates to be revised away from the smoothed expenditure-side estimates.