Near-integrated data and the analysis of political relationships

Suzanna DeBoef, Jim Granato

Research output: Contribution to journalArticlepeer-review

69 Scopus citations

Abstract

Theory: In finite samples, near-integrated data, widely thought to be stationary, mimic the same nonstationary data properties as integrated data. Hypothesis: Regressing two independent and near-integrated series results in high false rejection rates of the null hypothesis (spurious regressions). Method: Analytical derivations and numerical (Monte Carlo) analysis. We also extend the spurious regression test to actual data used in political science - macropartisanship - and a simulated near-integrated series. Results: False rejection of the null hypothesis is comparable to the integrated case (Granger and Newbold 1974). In addition, solutions to the spurious regression problem apply with equal force to the near-integrated situation.

Original languageEnglish (US)
Pages (from-to)619-640
Number of pages22
JournalAmerican Journal of Political Science
Volume41
Issue number2
DOIs
StatePublished - Apr 1997

All Science Journal Classification (ASJC) codes

  • Sociology and Political Science
  • Political Science and International Relations

Fingerprint Dive into the research topics of 'Near-integrated data and the analysis of political relationships'. Together they form a unique fingerprint.

Cite this