Nonresponse in the Norwegian labour force survey (LFS)

Studying data from the Norwegian Labour Force Survey sample linked to administrative register data, we focus on trends in nonresponse and its effects. Analyzing response rates and nonresponse bias in several subgroups, we find relatively stable response patterns. Although the overall response rate is decreasing, it seems to exist some structural differences. Factors such as age, gender, citizenship and employment seem to determine a response pattern that changes little over a relatively long period, when we observe clearly increasing nonresponse over the same period. Using register-based employment status available for the total sample, we have estimated the nonresponse bias for employment. Defining the response structure as the ratio of response rates among employed/notemployed, we compare this indicator with nonresponse rate and nonresponse bias.