Reflections On Four Decades Of Economic Forecasting
November 24, 2015
by Harald B. Malmgren
Advisor Perspectives welcomes guest contributions. The views presented here do not necessarily represent those of Advisor Perspectives.
Having left U.S. service under four presidents mid-1975 and having spent another couple of years as an advisor to the U.S. Senate, I found myself in demand for private, exclusive advisory roles to the CEOs of several of the world’s biggest banks, asset managers, industrial corporations, insurers and sovereign wealth funds. Some of the tasks involved rethinking business strategies and global deployment of management and production resources, but eventually I was asked to devote greater focus to financial interactions across global markets.
Because of my background in economics and econometrics learned under the tutelage of some of the world’s most prominent economists, including eight Nobel Laureates, it was natural that clients asked me privately to give my personal thoughts on many business, government and academic forecasts. Gradually, a number of central banks, finance ministries and even government leaders around the world asked me to compare notes and help look for what might be blindsiding events in their own outlooks.
From this extraordinary experience, I learned that there were wide variations in how economic activity was measured in practice. Data used not only varied in availability, but — more importantly – it varied in quality. The highest level decision makers rarely questioned the details in reports they were provided. Forecasts invariably focused on national economies with minimal attention to the interaction of economies through world trade, investment and financial flows.
It also became evident that everyone’s evaluation of the world economy was reliant on data typically available from the IMF, World Bank, OECD, WTO and other multilateral organization sources. Few decision makers were aware that such sources lag reality by many months, as each member government collects and polishes national data to present the best possible image of its own economy. When objective IMF national review teams suggested criticisms of policies and data provided by officials representing the big industrial economies, IMF directors who held a majority of votes in that institution quietly sidelined them.
By looking at how national economic data is typically collected over many years, it became increasingly clear that information technology was providing data faster and faster, allowing better insight on real-time performance, but forecasting remained mired in methodologies devised decades earlier. A bifurcation was slowly developing between private decisions based on near real-time, sector-specific information and public policy decisions based on observations that were months, sometimes years, old.
Central banks found themselves driving fast-moving machines through winding roadways with eyes glued to their rearview mirrors. Worse, government deliberations involving legislatures spent months and even years considering policy responses to business or labor complaints. Often, legislation was approved that no longer applied to the circumstances that had originally given rise to business complaints. This became particularly evident in the U.S. congressional debate about how to respond to foreign-government exchange rate and purchasing policies, barriers to U.S. exports and perceived foreign incentives for outsourcing American business activities. The U.S. executive branch and congressional responses to claims of foreign unfair treatment of American corporations, financial institutions and labor unions took months, sometimes years, to devise. The actions ultimately taken were no longer relevant as challenges took new forms or emerged in new geographic locations.