This book, edited by C. James Hueng of Western Michigan University, is a collection of six lectures aiming to raise awareness of new economic indicators and the cutting-edge methods involved in their construction. The book also examines how the new indicators compare with and complement existing, traditional economic indicators.
The book has seven chapters, each penned by a different author or set of authors. The first chapter, by Hueng, previews and summarizes the next six chapters. The book has no concluding chapter, so Hueng’s summary is helpful in introducing material the reader has not yet seen. However, the lack of a proper conclusion is noticeable throughout the book, with more than one chapter presenting new information or charts (rather than a summary of the chapter’s thesis) in a subsection titled “Conclusion.” While this variation in styles can be seen as a drawback, it is to be expected from a diverse collection of lectures.
The second chapter, by William A. Barnett and Kun He, is the first to delve deeper into the substance of the book. The chapter, which expands on previous research by Barnett, assumes that the reader is knowledgeable about most terms in the money and banking literature, such as inside versus outside money, liabilities versus assets, shadow banking, and the concept of a Divisia index. The authors propose an alternative measure of the money supply—a Divisia monetary aggregate that includes credit card expenditure data. This measure breaks traditional accounting structure (credit card balances are traditionally listed as liabilities rather than assets) and sets itself in opposition to regular measures of the money supply (for example, M1). Although Barnett and He draw few firm conclusions, their preliminary tests and results do not uncover any obvious issues with the inside-money and total-money Divisia measures augmented with credit card data.
Chapter 3, by Scott A. Brave, focuses on how the Federal Reserve Bank of Chicago produces the Chicago Fed National Activity Index (CFNAI) and the National Financial Conditions Index (NFCI). Both indexes are based on a mix of principal component analysis and dynamic-factor models, whose estimation is now possible because of recent advances in computational efficiency. The CFNAI, which is similar to gross domestic product (GDP), is a monthly measure of the strength of the real economy and has a strong record of predicting recessions with a zero-month lag, approximately 6 to 18 months ahead of official measures. The NFCI is a weekly measure of the health of the U.S. financial sector. It functions as a leading indicator of the stress of the financial sector, and its component indicators can be subdivided into three categories: risk, credit, and leverage (separate indexes for these categories can also be constructed). The chapter lays out why and how these indexes were constructed, and its conclusion clearly recaps relevant information and offers index extensions and ideas for future improvements.
Chapter 4, by Domenico Giannone, Eric Qian, Argia Sbordone, Mihir Trivedi, and Patrick Adams, focuses on the problem of timeliness of official GDP estimates, examining how the Nowcast measure of the Federal Reserve Bank of New York can ameliorate the issue. The Nowcast is based on a dynamic-factor model and aims to provide a real-time estimate of GDP. The authors present two case studies showing that the Nowcast is particularly useful and performs well when economic conditions are shifting rapidly, as in the 2007–09 Great Recession, and when official estimates are unavailable, as in the 2018–19 U.S. federal government shutdown. Although not as accurate as official measures published months later, the Nowcast could still signal the direction of GDP in near realtime in the Great Recession. During the federal government shutdown, the Nowcast had larger errors (because of several missing model inputs), but its accuracy was not substantially hampered. Overall, Giannone et al.’s chapter provides many avenues for evaluating and extending the Nowcast.
Chapter 5, written by Alessandro Barbarino and Chiara Scotti, varies in quality. The chapter seeks to answer the following question: “What is the probability of a recession?” Superbly breaking down the intricacies of the question, Barbarino and Scotti explain that its answer, and the answer’s calculation, depends on one’s time horizons. Predicting a recession 3 months in advance is different from predicting a recession 18 months in advance. The authors present a short but strong literature review of previous predictive attempts, and they also construct their own models of recession probability. Their default model uses the Aruoba-Diebold-Scotti (ADS) business conditions index, and they perform a variable selection process to determine if other variables, alongside the ADS index, can help predict recessions at various forecast horizons. Although Barbarino and Scotti present plenty of excellent information, they do not provide an answer, or even a range of answers, in the text itself. They do include several charts showing probabilities, but they miss the opportunity to offer even a single best-guess estimate.
In chapter 6, Steven J. Davis introduces the economic policy uncertainty (EPU) index. This index, which Davis developed with Scott Baker and Nick Bloom, is constructed from a count of newspaper articles that contain at least three terms the researchers deem indicative of economic policy uncertainty. Because different countries can signal uncertainty at different times, Davis constructs EPU indexes for 20 countries besides the United States, as well as a global EPU index. Although the U.S. EPU index increases with elections, military engagements, and fiscal brinkmanship, the EPU indexes of countries in the Eurozone more strongly respond to local issues and crises. Increases in the global EPU index align well with large international events. Later in the chapter, Davis also introduces the trade policy uncertainty (TPU) index. This index, which is similar to the EPU index but focuses on trade uncertainty, rises during trade negotiations, trade disputes, and trade wars. An increase in the TPU index also signals a more volatile stock market. Overall, Davis’s chapter does a good job of presenting the EPU and TPU alternative economic indicators, some of their potential uses, and their relation to broader economic trends and events.
Of all the chapters, chapter 7 offers the most tantalizing look at the future of economic indicators. Written by Adam Storeygard, this chapter presents a compelling case for the use of satellite data as a complement to current methods. Storeygard suggests that such use is advantageous because satellites can fill in missing data, gather spatial data in hard-to-reach locations, cheaply repeat measurements, provide data across the entire globe, ignore discontinuities at state lines or country borders, and verify other measurements that are more prone to false or biased reporting. The chapter presents concrete examples of each potential use of satellite data, and the author also addresses limitations and possible extensions of the method.
One solid throughline of the book is that the new alternative economic indicators complement rather than compete with the more traditional metrics. The Divisia index combines with M1, M2, and M3 measures to deepen our understanding of the money supply. The CFNAI provides a disaggregated view of why GDP rose or fell at any given time, and the ADS index can show underlying weakness (or strength) in business conditions in times of strong (or weak) economic growth. The Nowcast fills in for GDP between official data releases, and satellite data can close information gaps when GDP data are impossible to collect. The NFCI provides a digestible alternative to the myriad yield curves common in finance circles, and the EPU and TPU indexes capture policy environment sentiments that would otherwise be available only through interviews of experts across disciplines. Another consistent thread across the indicators is that they all would have been impossible to construct before the 21st century, which saw dramatic advances in computation technology.
For those looking to be on the cutting edge of econometric measures, this book is an excellent primer. To get the most out of it, readers must have knowledge of economics and mathematics at the undergraduate level, but the book’s conclusions and overall presentation remain accessible. Finally, the book is licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International license, which allows anyone to legally obtain the book at no cost. Professors also can freely distribute the book to their students. Given its high quality and low barrier to entry, Alternative Economic Indicators may become a touchstone for future economists, if or when the new economic indicators reach the mainstream.
United States Department of Labor
