Monthly Labor Forces Surveys Portray Smooth Labor Markets and Bias Fixed Effects Estimation: Evidence from Israel’s Transition from Quarterly to Monthly Surveys

Authors: Haggay Etkes

Abstract:

This study provides evidence for the impact of monthly interviews conducted for the Israeli Labor Force Surveys (LFSs) on estimated flows between labor force (LF) statuses and on coefficients in fixed-effects estimations. The study uses the natural experiment of parallel interviews for the quarterly and the monthly LFSs in Israel in 2011 for demonstrating that the Labor Force Participation (LFP) rate of Jewish persons who participated in the monthly LFS increased between interviews, while in the quarterly LFS it decreased. Interestingly, the estimated impact on the LFP rate of self-reporting individuals is 2.6–3.5 percentage points while the impact on the LFP rate of individuals whose data was reported by another member of their household (a proxy), is lower and statistically insignificant. The relative increase of the LFP rate in the monthly survey is a result of a lower rate of exit from the LF and a somewhat higher rate of entry into the LF relative to these flows in the quarterly survey. These differing flows have a bearing on labor search models as the monthly survey portrays a labor market with less friction and a “steady state” LFP rate that is 5.9 percentage points higher than the quarterly survey. The study also demonstrates that monthly interviews affect a specific group (45–64 year-olds); thus the sign of coefficient of age as an explanatory variable in fixed-effects regressions on LFP is negative in the monthly survey and positive in the quarterly survey.

Keywords: measurement error, surveys, search, LFSs