Comment

Abstract

This is a very timely paper. Thomas Laubach is to be commended for working in real time on this important issue. His analysis starts with an important analytic point: To understand the effect of fiscal conditions on interest rates, one must separate four effects: (1) traditional crowding out, (2) macroeconomic cyclical effects, (3) perceived risk of default, and (4) risk aversion with respect to the risk of default. These have played different roles in different countries and at different times. Laubach notes that only the first two have been relevant in the United States, with (in his judgment) the macrocyclical effects dominating. In the EuropeanMonetary Union (EMU) in recent years, default became more important, with differences in the country-specific risk becoming key since 2009 in explaining large increases in interest rates in several countries. I agree with this important analytic point. Measuring deficits. There are difficult problems in measuring fiscal deficits and future fiscal conditions, the key variables in Laubach’s analysis. He correctly notes that pure econometric forecasts of future deficits are inadequate.He therefore usesCongressional BudgetOffice (CBO) forecasts for the United States and OECD forecasts for the EMU countries. Laubach recognizes that these are quite imperfect as measures of the conceptually relevant variables. The CBO’s “baseline forecast” is required by law to project future deficits on the assumption that the existing lawwill continue in the future. There are also CBO forecasts based on the administration’s budget each year.While there are no forecasts based on consensus judgments about what legislative changes might occur in the future, the CBO’s Long Term Budget Outlook ( June 2010) contains an “Alternative Fiscal Scenario” that the CBO staff believes is themost likely outlook for future budget policies and the resulting deficits. It would be interesting to compare these informed judgmental forecasts with the CBO baseline figures to see howmisleading the latter might be. The data