A researcher interested in examining the relationship between wage and unemployment experience, uses panel data from...

Transcribed Image Text:

A researcher interested in examining the relationship between wage and unemployment experience, uses panel data from the UK over the years 2011- 2015, and estimates the following relationship: lwageit = x₁ + a₂ + Eit i = 1,.., N and t = 1,..,5 (where, lwage=the logarithm of weekly wage; x contains the variables of interest: unwks- the number of weeks unemployed in the previous year), and two control variables: age and age²/100= age* age/100. The results from the estimation of model in equation (1), using two different linear panel data estimators are provided in Columns [1]-[3] of Table 1. Total number of individuals in the sample N = 1,200. (1) VARIABLES age age²/100 unwks Constant Individual time averages mean(age) Table 1 [1] WG 0.130*** (0.007) -0.134*** (0.009) -0.004*** (0.001) 4.266*** (0.146) [2] GLS1 0.115*** (0.005) -0.147*** (0.007) -0.005*** (0.001) 4.514*** (0.104) [3] GLS2 0.130** (0.007) -0.134*** (0.009) -0.004*** (0.001) 4.603*** (0.144) GLS1 -0.000 (0.010) -0.020 Mean(age²/100) (0.013) mean(unwks) -0.053*** (0.010) Notes: (i) Standard errors in parenthesis with robust standard errors in Column [1]; (ii) *** p<0.01, ** p<0.05, * p<0.1; (iii) Columns are: [1]: Within-group; [2] GLS; [3] GLS with additional time averages of time-varying variables. (a) Compare and contrast the estimators used and also the results reported in Columns [1]-[3]. The discussions should include interpretation of the estimates and the importance of the underlying assumptions that are necessary to obtain consistent estimators. Do not derive the estimators. (15 marks) (b) Why do you think the researcher entered the variable age as age/100? (3 marks) (c) Why do you think the researcher included the individual time averages in Column [3]? (5 marks) (d) Comparing columns [1] to [3] in Table 1, (i) is there an interesting hypothesis you think, you could test? You will have to explain in detail what the hypothesis that is being tested and why you think this might be an important hypothesis to test; (ii) what would be the implication if the null is rejected? (7 marks) A researcher interested in examining the relationship between wage and unemployment experience, uses panel data from the UK over the years 2011- 2015, and estimates the following relationship: lwageit = x₁ + a₂ + Eit i = 1,.., N and t = 1,..,5 (where, lwage=the logarithm of weekly wage; x contains the variables of interest: unwks- the number of weeks unemployed in the previous year), and two control variables: age and age²/100= age* age/100. The results from the estimation of model in equation (1), using two different linear panel data estimators are provided in Columns [1]-[3] of Table 1. Total number of individuals in the sample N = 1,200. (1) VARIABLES age age²/100 unwks Constant Individual time averages mean(age) Table 1 [1] WG 0.130*** (0.007) -0.134*** (0.009) -0.004*** (0.001) 4.266*** (0.146) [2] GLS1 0.115*** (0.005) -0.147*** (0.007) -0.005*** (0.001) 4.514*** (0.104) [3] GLS2 0.130** (0.007) -0.134*** (0.009) -0.004*** (0.001) 4.603*** (0.144) GLS1 -0.000 (0.010) -0.020 Mean(age²/100) (0.013) mean(unwks) -0.053*** (0.010) Notes: (i) Standard errors in parenthesis with robust standard errors in Column [1]; (ii) *** p<0.01, ** p<0.05, * p<0.1; (iii) Columns are: [1]: Within-group; [2] GLS; [3] GLS with additional time averages of time-varying variables. (a) Compare and contrast the estimators used and also the results reported in Columns [1]-[3]. The discussions should include interpretation of the estimates and the importance of the underlying assumptions that are necessary to obtain consistent estimators. Do not derive the estimators. (15 marks) (b) Why do you think the researcher entered the variable age as age/100? (3 marks) (c) Why do you think the researcher included the individual time averages in Column [3]? (5 marks) (d) Comparing columns [1] to [3] in Table 1, (i) is there an interesting hypothesis you think, you could test? You will have to explain in detail what the hypothesis that is being tested and why you think this might be an important hypothesis to test; (ii) what would be the implication if the null is rejected? (7 marks)

Answer rating: 100% (QA)

a The estimators used in the analysis are withingroup WG generalized least squares GLS1 and GLS with additional time averages of timevarying variables GLS2 Lets compare and contrast these estimators a... View the full answer