Research Narrative

Below I provide narratives of current and prior research organized by topic.  I have had the good fortune of working with incredibly gifted scholars. Whether these individuals are coauthors or not in the works described below, I have benefited from their insights as well as from the comments and constructive critics of many other economists. If you have work closely related to any of the themes below, please make me aware.

Real Exchange Rates

My recent work revisits two key propositions in international macroeconomics and finance — the Law-of-One-Price and Purchasing Power Parity. The main novelty of my approach is to collect micro-data on absolute prices of goods and use these data to gain novel insights about key properties of international relative prices. These projects were supported by NSF grants (SES-0524868, 2005-07, SES-0136979, 2002-04) and all of the non-proprietary micro-data will be made available to researchers on this site or through links from this site.

Almost all existing literature studies of international relative prices focus on relative PPP, the persistence and time series variability of CPI-based real exchange rates. The consensus finding is that the variance of real and nominal bilateral exchange rates are approximately equal and that deviations of real exchange rates from their unconditional sample means have half-lives of between three and five years. Less is known about the absolute size of the deviations themselves or their patterns across goods, bilateral locations or time periods. Collaborative work with Mototsugu Shintani, Chris Telmer and Marios Zachariadis explores the microeconomic structure of real exchange rates and LOP deviations using absolute retail price data across cities, some of which are located across countries, some across locations within a country.

In joint work with Mototsugu Shintani and Takayuki Tsuruga the goal is to elucidate the dynamics of LOP deviations using highly tractable Calvo-style models with both monetary and productivity shocks. This research extends the Calvo models to multiple locations of production and allows for heterogeneity in the frequency of price changes across goods. Most paper use a particular model structure to suit novel the particular micro-price data under examination.

The paper, “Do Sticky Prices Increase Real Exchange Rate Volatility at the Sector Level?” European Economic Review, May 2013, Vol. 62, 58-72, revisits the debate between two icons of international finance, Alan Stockman and Michael Mussa. Stockman argued that real exchange rate variation arises from fluctuations in productivity under flexible prices while Mussa argued that variation arises from fluctuations in nominal exchange rates under sticky prices. This paper develops a time-dependent pricing model with nominal and real shocks to address both viewpoints. In particular, we show that the Calvo time-dependent pricing model predicts a u-shaped volatility curve for the variation of LOP deviations as a function of the observed stickiness of prices. The downward sloping portion of the curve covers relatively flexible price goods with their variability dominated by real shocks, while the upward sloping portion covers relatively sticky price goods with the variability dominated by nominal shocks. Using micro-data generously provided by Virgiliu Midrigan we find the exchange rate volatility curve is downward sloping over much of the range of goods reflecting an apparent dominance of real shocks. However, at short forecast horizons (less than 12 months) nominal shocks account for a substantial minority of the variance of the forecast error variation. Simply put, Mussa’s view matches closely with short-horizon forecast error variance for relatively sticky goods prices while Stockman’s view matches the median good and particularly the variance at business cycles and lower frequency movements in real exchange rates.

In The Law of One Price without the border: the role of distance versus sticky prices, (Economic Journal, 2010), we build a single-country, two-city, model with nominal rigidities and transportation costs which predicts that variation of LOP deviations is lower for goods with infrequent price adjustment after controlling for distance. We adapt the Engel and Rogers (1996) regression to analyse Japanese retail prices at the city level and find that results are consistent with the theoretical predictions. The regression results also suggest that the distance equivalent of nominal rigidities can be as large as the `width’ of the border found in the literature on international LOP, casting doubt on casual interpretations of distance as a metric for market segmentation. The paper Accounting for persistence and volatility of good-level real exchange rates: the role of sticky information, (Journal of International Economics, 2010), builds upon an intriguing paper Sticky prices and sectoral real exchange rates, by Kehoe and Midrigan (Federal Reserve Bank of Minneapolis Working Paper No 656) who showed that, under a standard assumption on nominal price stickiness, empirical frequencies of micro price adjustment cannot replicate the time-series properties of the Law of One Price deviations. We extend their sticky price model by combining good-specific price adjustment with information stickiness in the sense of Mankiw and Reis (2002). Our framework allows for multiple cities within a country. Using a panel of U.S.-Canadian city pairs, we estimate a dynamic price adjustment process for 165 individual goods. Under a reasonable assumption on the money growth process, we show that the model matches the LOP persistence of the median good and accounts for approximately one-half of its volatility when information updates occur every 12 months.

In joint work with Mototsugu Shintani (“Persistence in Law-of-One-Price Deviations: Evidence from Micro-data“, (Journal of Monetary Economics, 2008), we study the dynamics of good-by-good real exchange rates using a micro-panel of 270 goods prices drawn from major cities in 63 countries and 258 goods prices drawn from 13 major U.S. cities. We find the half-life of deviations from the Law-of-One-Price for the average good is about 19 months. The average half-life is very similar for cross-border pairs in OECD to what we see across cities within the U.S., suggesting little in the way of nominal exchange rate regime influences. The average non-traded good has a half-life of 24 month compared to 18 months for traded-goods, for the OECD, with modest differences elsewhere. Aggregating the micro-data has little impact on persistence, except within the U.S. where the half-life increases from 18 months (median good) to 26 months (CPI-weighted aggregate). The official CPI data give consistently short half-lives over the period 1990 to 2005, consistent with what we find in the micro-data over the same period. It therefore appears that persistence of price deviations is lower in the recent historical period compared to the entire postwar period where half-lives of 36 to 60 months are more typical.

In “Understanding European Real Exchange Rates,” (with Chris I. Telmer, and Marios Zachariadis), American Economic Review, June 2005, Vol. 95, No 23, 724-738″ explores good-by-good deviations from the Law-of-One-Price for over 5,000 goods and services between European Union countries for the years 1975, 1980, 1985 and 1990. We find that between most countries there are roughly as many overpriced goods as there are underpriced goods so that PPP holds to a good approximation, particularly after controlling for wealth differences. These findings are robust across years, in spite of relatively large movements in nominal exchange rates. Variation around the averages is large but is found to be related to economically meaningful characteristics of goods, namely: international tradeability and the cost share of non-traded inputs into production. We measure both at highly disaggregated levels. An earlier version of the paper used data on product brands and found that product heterogeneity is at least as important as geography in explaining relative price dispersion. Overall, our data provide strong evidence that international goods markets are segmented, but (i) the evidence relies on absolute deviations from the Law-of-One-Price, not deviations from PPP, (ii) some markets are much more segmented than others, with the distinctions being consistent with economic theory. To move to the site where the data and data appendix are posted, click here. The companion paper, “Price Dispersion: the Role of Borders, Distance and Location” (with Chris I. Telmer, and Marios Zachariadis) uses the Economist Intelligence micro-price survey to examine the relationship between the price dispersion and income levels, distance and borders.

The Terms of Trade

Some earlier work in a similar vein focused on the terms of trade, which while often correlated with the real exchange rate is a very different object (both in terms of its empirical properties and its theoretical role). In “Oil Prices and the Terms of Trade” (Journal of International Economics, 2000) David K. Backus and I developed a three region, three good dynamic model and use it to argue that much of the substantial variability of the terms of trade and unstable correlation patterns of trade prices with output and trade volumes can be accounted for by variation in oil prices and their impact on world business cycles. In “Commodity Prices and the Terms of Trade” (Review of International Economics, 2000). Prasad Bidarkota and I combine data on individual commodity prices (e.g., petroleum, wheat, etc.) and the national terms of trade and show that a few key international commodity prices can account for much of the variation in the terms of trade of a typical developing country.

I have not published work on the relationship between monetary variables and exchange rates (real or nominal), but I view the linkages as important. I did have the pleasure of learning something about “dollarization” at a conference on the topic organized by the Federal Reserve Bank of Cleveland (the papers and comments will appear in the Journal of Money, Credit and Banking this spring). At this conference, I commented on a paper by Del Negro and Obiols-Homs titled: “Has Monetary Policy Been So Bad That It Is Better to Get Rid of It? The Case of Mexico?

International Business Cycles

My publications in open economy macroeconomics have challenged a number of conventional views of the extent — and national impact — of international capital market integration. One of the most stable empirical regularities observed in international macroeconomic data is the fact that national savings and investment rates are highly positively correlated. These observations were interpreted by many economists as indicating a severe lack of international capital mobility while at about the same time a consensus had emerged in the international finance literature that the world was characterized by a high and increasing degree of international capital market integration. The juxtaposition of these views — and the observations that motivated them — was labeled the Feldstein-Horioka puzzle. My co-authored publication “Explaining Saving/Investment Correlations,” (with Marianne Baxter, American Economic Review, 1993) was instrumental in resolving this puzzle by demonstrating that positive savings and investment correlations are a robust prediction of a model that assumed perfect financial market integration.

Despite almost universal skepticism about the assumption that individuals can hedge all idiosyncratic risks at zero cost, it remains the predominant maintained assumption in general equilibrium theory. Marianne Baxter and I examined the quantitative impact of alternative assumptions about asset market structure in “Business Cycles and the Asset Structure of Foreign Trade,” (International Economic Review, 1995). We show that in theory the asset market structure matters most when shocks to the economic environment are very persistent (or contain unit roots). Given the difficulty in distinguishing unit root from near unit root productivity it is not obvious which modeling assumption is to be preferred, at least from the perspective of business cycle statistics. The incomplete markets model does allow one to study a broader set of issues than the complete markets model such as the evolution of relative consumption across countries and the dynamics of debt accumulation.

The manuscript “On International and National Dimensions of Risk Sharing,” develops a metric on which to base statements about the extent of risk sharing and applies the metric in one of the first empirical studies to combine regional and national consumption and production data. The results provide evidence that financial market integration is imperfect at both the national and international levels but the national capital market is more perfectly integrated than the international capital market.

My publication “Country Size and Economic Fluctuations,” (Review of International Economics, 1997) presents evidence that business cycles are more volatile in smaller economies than in larger economies and develops a model to explain these differences. The model shows that in a world of highly mobile capital business cycle volatility will be highest in the smallest countries even when population size is their only distinguishing economic feature. Thus greater variability is intrinsic to small open economies even when the underlying uncertainty that gives rise to the fluctuations is no different than that in large open economies.

Economic History of the Interwar Period

I have four publications related to economic history and tariffs. The paper “Tariffs and Aggregate Economic Activity: Lessons From the Great Depression,” (with James Kahn, Journal of Monetary Economics, 1997) evaluates the contribution of the tariff war to the international economic depression of the 1930’s. The policy implications drawn from this article were that U.S. tariff increases, combined with foreign retaliation contributed to — but were not the main causes of — the global economic depression. We also argue against the prevailing view that the macroeconomic effects of commercial policy are small by allowing for trade in intermediate inputs and taking into account the dynamics of capital accumulation and labor supply that persistent changes in tariffs entail. Jim and I co-authored a paper, “Tariffs and the Great Depression Revisited,” in the Minneapolis Federal Reserve Volume on “Great Depressions of the Twentieth Century” (edited by Edward Prescott and Timothy Kehoe). In this paper we express the tariff distortions of the multi-country, mulit-sector trade model used in our earlier JME paper in terms of three wedges (labor tax wedge, capital tax wedge and efficiency/productivty wedge) using the Chari, Kehoe and McGratten methodology. We show that tariffs in our model show up almost entirely as efficiency wedges (productivity shocks) in a symmetric tariff war scenario. We simulate the proto-type stochastic growth model using these aggregate wedges and replicate the aggregate implications of our trade model. Macroeconomics should find this exercise useful in understanding the tariff war in a macro-context. I believe the CKM methodology will become an important part of the pedagogy and method in future quantitative research.

I have one paper and one book review that focus more on economic history. The article “Sources of Variation in Real Tariff Rates: The United States, 1900 – 1940,” (American Economic Review, 1994) shows that much of the variation in U.S. tariff rates during the interwar period was induced by the impact of price level fluctuations on nominally rigid specific duties rather than legislative changes such as those attributed to the Smoot-Hawley Tariff Act of 1930. Readers interested in the intellectual history of the commercial policy date will find the book: “Against the Tide: An Intellectual History of Free Trade,” by Douglas Irwin (the link is to my Southern Economic Journal review of his book).

Back Home