Unit roots and all that: the impact of time-series methods on macroeconomics

Abstract

Over the past two decades applied macroeconomics has been transformed by the widespread adoption of a set of new statistical techniques: unit-root tests, vector autoregressions, Granger causality and cointegration. Although these techniques were developed to answer statistical questions, they diffused very rapidly through applied economics because they were thought to be able to answer important theoretical questions in macroeconomics. This paper argues that these techniques have not delivered on the early promises; not because they were not useful - they are very useful for many purposes - but because economists expected too much: they wanted to believe that a statistical summary of the data (an estimate or test statistic) could answer an economic question without interpretation. The paper sets out the statistical motivation for the procedures; the economic questions they were supposed to answer; and the issues that arise in trying to answer economic questions about the sources of trends and cycles, ausality and the nature of equilibrium from statistical summaries.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 23 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.