tiểu luận kinh tế lượng 2 explaining factors that affect the working time (minutes that a person spends at paid labor each week) of people in 1975 and 1981

INTRODUCTION Econometrics, one of the most crucial branches of economics, is the application of statistical techniques in evaluation and testing of economic theories Economists apply econometric modeling in a variety of specific fields (such as labor economics, development economics and finance) to shed light on theoretical questions, namely the relationship between variables of interest They also use this tool to inform public policy debates, make business decisions, and forecast future events Majoring in International Economics, all of us are aware of the importance of econometrics This awareness has been motivating us to learn and some further research in order to have an insight of this field Under the instruction of Mrs Phuong Mai in Econometrics I and PhD Dinh Thi Thanh Binh in Econometrics II, we have built an econometric model explaining factors that affect the working time (minutes that a person spends at paid labor each week) of people in 1975 and 1981 Our research is based on the data set 5_SLP75_81.DTA and accomplished by the usage of a well – known computer software – Stata By accomplishing this assignment, we are greatly grateful to our lecturer, Mrs Binh She has provided us with a large amount of precious knowledge through her lessons, which plays a role in the completion of our report Furthermore, in our process of searching for ideas, she has offered suggestions, given us specific examples and advice Under her detailed instruction, our work has become more oriented and completed Moreover, we expected that we would gain more understanding of econometrics and the appliances of it; all of us will be fluent to use Stata software As this is our first research about econometrics, it may contain some mistakes or misunderstanding Therefore, we are looking forward to hearing your comments, suggestions or even criticism to better our assignment I LITERATURE REVIEW Reason for choosing the topic Working time varies substantially among countries and among generation, but also within countries, e.g due to the prevalence of part-time work and working hours regulations or agreements Due to the direct relation between the number of working hours and the labor productivity, understanding how they interact with each other is an important element of understanding labor demand, and has important implications for the regulation of working hours and company’s management Low productivity is not an individual phenomenon but a universal problem Therefore, improving the labor productivity resulted from change in working time is not only related to the development of the national economy, but also directly affects the sustainable and stable development of the enterprise The increase of labor productivity is the basis of the accumulation of wealth, the increase of real wages and the improvement of living standards, while the increase of labor productivity is helpful to restrain the inflation caused by the rise in nominal wages Therefore, it is urgent to study the problem of working hours, especially laborers in 1975 and 1981, who are considered as the experienced generation contributing significantly to the whole economic status The issue of working time is meaningful in each country around the word, but it is hard for each country to really make clear of important factors to promote it and what are not There are so many factors are affecting the working time as well as labor productivity Recognizing the urgency of analyzing and evaluating the effectiveness of the activity in business, within the scope of this paper, we would like to analyze the striking impact of age, educational, health level, gender, marital status and the amount of weekly relaxing time to minutes spent on working of laborers in 1975 and 1981 Methods of analysis To get a thorough understanding about this issue, we should set up a Fixed-effect model and with the support of Stata software to measure the extend of each factor’s influence on the working time of workers in 1975 and 1981 Stata Stata is a general-purpose statistical software package created in 1985 by StataCorp Most of its users work in research, especially in the fields of economics, sociology, political science, biomedicine and epidemiology Stata's capabilities include data management, statistical analysis, graphics, simulations, regression, and custom programming It also has a system to disseminate user-written programs that lets it grow continuously Fixed-effect model A statistical model of the output of a business or manufacturing process that treats all variables as non-random values Fixed effects models are used to determine optimal values for inputs to business or manufacturing processes when random factors are judged not to be present in the process, or determined not to have an effect on the process output DESCRIPTION OF THE DATA I Overview: The data that we use for the research is a secondary data that was collected into the file 5_SLP75_81.DTA by Ph.D Dinh Thi Thanh Binh The data include the same groups of observation (age, education, marriage, health condition, young kids, gender, ) observed in two differnt time (1975 and 1981) => This is a Panel Data sample II Description of the data: - To describe the data of the file, we use the command des and have the result: des Contains data from C:\Users\Hieu\Desktop\KTL 2\KTL2_CLC2\KTL2_CLC\file data_ful > l\5_SLP75_81.DTA obs: 239 vars: 20 17 Aug 1999 22:56 size: 6,214 variable name age75 educ75 educ81 gdhlth75 gdhlth81 male marr75 marr81 slpnap75 slpnap81 totwrk75 totwrk81 yngkid75 yngkid81 ceduc cgdhlth cmarr cslpnap ctotwrk cyngkid storage type byte byte byte byte byte byte byte byte int int int int byte byte byte byte byte int int byte display format value label %9.0g %9.0g %9.0g %9.0g %9.0g %9.0g %9.0g %9.0g %9.0g %9.0g %9.0g %9.0g %9.0g %9.0g %9.0g %9.0g %9.0g %9.0g %9.0g %9.0g variable label age in 1975 years educ in '75 years educ in '81 = if good hlth in '75 =1 if good hlth in '81 =1 if male = if married in '75 =1 if married in '81 mins slp wk, inc naps, '75 mins slp wk, inc naps, '81 minutes worked per week, '75 minutes worked per week, '81 = if child < 3, '75 =1 if child < 3, '81 change in educ change in gdhlth change in marr change in slpnap change in totwrk change in yngkid Sorted by: We can see that the data shows the information of factors that affect to the total working time per week (minutes per week), based on 239 observations in different years 1975 and 1981 - In order to run the panel model, we have to reshape the original data into the new data that is suitable for installing panel data: ID 1 2 3 4 5 6 7 8 9 10 10 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18 19 19 20 20 21 Year age 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 educ 46 52 39 45 55 61 39 45 54 60 32 38 30 36 27 33 33 39 51 57 23 29 46 52 38 44 30 36 39 45 53 59 59 65 54 60 35 41 46 52 36 gdhlth 16 16 16 16 15 15 16 16 17 17 16 16 16 16 16 16 17 19 14 14 16 16 17 19 12 12 10 10 12 12 12 12 14 14 14 14 12 12 17 18 12 male 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 marr 1 1 1 1 0 1 1 1 0 1 0 0 0 1 0 0 0 1 0 1 1 1 1 1 1 0 1 1 1 1 1 0 1 1 1 1 1 1 totwrk 2050 2430 2713 2610 2493 2778 1787 3118 552 2348 2567 1571 2272 3421 400 1853 2617 1312 2700 3292 2526 2075 2948 3207 2570 2705 2620 2932 3588 4122 958 630 2575 2600 3135 2958 3003 1887 3188 21 22 22 23 23 24 24 25 25 26 26 27 27 28 28 29 29 30 30 31 31 32 32 33 33 34 34 35 35 36 36 37 37 38 38 39 39 40 40 41 41 42 42 43 43 44 44 45 45 46 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 42 25 31 29 35 37 43 31 37 43 49 35 41 37 43 28 34 35 41 56 62 47 53 25 31 35 41 28 34 24 30 34 40 36 42 40 46 64 70 48 54 32 38 36 42 23 29 59 65 28 12 17 18 12 12 12 12 17 17 12 12 12 14 12 14 16 16 12 12 11 12 16 16 16 16 12 12 12 12 13 13 16 18 12 12 13 13 12 12 12 12 12 12 16 18 13 14 12 13 16 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 1 1 0 1 1 1 1 1 1 1 0 0 1 1 1 0 1 1 1 1 1 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 587 2363 2610 2270 1065 1851 1400 2592 1049 1995 650 2480 2037 2049 2562 1175 2437 1350 2750 3242 1935 2638 3245 875 237 1125 2330 1497 2362 2686 3510 2205 2475 3598 1994 3276 3350 2508 1287 900 900 2108 1775 1905 362 2477 3727 1013 867 2568 46 47 47 48 48 49 49 50 50 51 51 52 52 53 53 54 54 55 55 56 56 57 57 58 58 59 59 60 60 61 61 62 62 63 63 64 64 65 65 66 66 67 67 68 68 69 69 70 70 71 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 34 41 47 27 33 33 39 23 29 56 62 51 57 60 66 52 58 47 53 29 35 44 50 55 61 37 43 46 52 30 36 64 70 25 31 31 37 37 43 39 45 25 31 55 61 60 66 36 42 35 16 17 18 17 17 12 12 12 12 13 13 12 12 17 17 8 12 12 12 12 9 17 18 17 17 11 11 16 16 12 12 17 19 12 12 12 12 11 11 12 12 12 12 12 14 10 11 12 1 1 1 1 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1500 3518 3817 305 905 1553 2390 0 375 3105 3267 1988 2467 1733 1805 2351 2475 2263 1537 2616 2692 2501 2425 2125 825 1738 530 787 0 2150 2200 2325 2570 3148 2515 1775 1062 1350 1350 188 2712 4325 2430 2592 2250 71 72 72 73 73 74 74 75 75 76 76 77 77 78 78 79 79 80 80 81 81 82 82 83 83 84 84 85 85 86 86 87 87 88 88 89 89 90 90 91 91 92 92 93 93 94 94 95 95 96 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 41 53 59 33 39 32 38 35 41 29 35 40 46 48 54 30 36 28 34 36 42 52 58 36 42 33 39 40 46 32 38 51 57 26 32 40 46 33 39 30 36 60 66 53 59 40 46 47 53 54 12 12 13 16 16 16 16 17 17 16 16 12 12 12 12 12 12 17 18 9 17 18 16 17 12 12 17 18 12 12 16 16 12 12 14 14 12 17 12 12 8 12 12 16 16 12 12 14 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 0 1 0 1 1 1 0 1 0 1 1 0 1 0 0 1 1 1 0 0 1 1 1 1 0 1 1 1 1 1 0 1 1 1 1 1 0 1 1 1 1 0 1 1 1162 2506 2591 2563 1762 2763 3047 2048 692 2418 2537 2420 2512 2735 1675 2163 2175 3556 3067 1555 3765 3869 2607 2281 1882 2488 2250 1263 2150 113 2325 1800 388 3230 4195 2755 2287 2475 2662 2975 0 2388 2425 1538 2140 2150 96 97 97 98 98 99 99 100 100 101 101 102 102 103 103 104 104 105 105 106 106 107 107 108 108 109 109 110 110 111 111 112 112 113 113 114 114 115 115 116 116 117 117 118 118 119 119 120 120 121 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 60 33 39 32 38 33 39 35 41 57 63 40 46 28 34 23 29 24 30 26 32 36 42 32 38 35 41 29 35 33 39 48 54 26 32 42 48 55 61 26 32 63 69 26 32 50 56 28 34 37 14 14 14 12 12 12 12 12 12 12 12 12 12 12 12 14 15 15 16 12 12 10 12 12 16 16 16 16 8 11 11 12 12 11 11 14 14 15 15 5 17 17 12 12 14 15 14 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 0 1 0 1 1 1 0 1 0 0 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 0 1 1 1 1 1 1 0 2512 2488 2525 1050 2437 1806 1352 2113 1973 588 2325 2300 2092 2195 1962 1920 2300 2250 930 2896 2630 2276 950 68 2865 1125 2175 2488 2485 2138 375 1880 2250 2946 1662 375 1360 4065 570 363 525 2775 2400 3194 2815 1168 2954 121 122 122 123 123 124 124 125 125 126 126 127 127 128 128 129 129 130 130 131 131 132 132 133 133 134 134 135 135 136 136 137 137 138 138 139 139 140 140 141 141 142 142 143 143 144 144 145 145 146 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 43 53 59 44 50 54 60 36 42 30 36 28 34 25 31 38 44 29 35 27 33 35 41 25 31 29 35 32 38 58 64 26 32 56 62 45 51 25 31 34 40 44 50 27 33 26 32 47 53 43 15 15 15 16 16 9 17 19 12 12 12 12 12 12 9 12 12 15 15 12 12 17 17 12 12 12 12 8 12 12 5 12 12 16 16 11 11 16 16 16 16 13 13 8 12 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 0 1 1 1 0 1 0 0 0 1 0 0 0 1 1 0 0 0 1 1 1 1 1 0 1 1 1 1 1 1 1 1 0 0 1 1 1 1 1 1 1437 1188 2325 2513 875 1208 1637 1913 2485 3146 2347 2791 2312 2250 2920 2681 2635 1800 2312 2238 1280 1588 2637 2820 2030 2313 2970 2965 1629 2263 1212 2506 2960 2880 2326 2437 3075 1000 2463 1075 1105 3180 1745 2711 2337 2388 1162 2325 146 147 147 148 148 149 149 150 150 151 151 152 152 153 153 154 154 155 155 156 156 157 157 158 158 159 159 160 160 161 161 162 162 163 163 164 164 165 165 166 166 167 167 168 168 169 169 170 170 171 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 49 60 66 30 36 60 66 33 39 48 54 52 58 37 43 32 38 55 61 33 39 33 39 25 31 37 43 50 56 52 58 32 38 27 33 36 42 52 58 36 42 65 71 34 40 26 32 34 40 27 12 12 12 13 13 11 12 17 19 17 19 17 18 12 12 17 18 12 12 12 13 17 17 12 12 12 12 12 12 17 17 12 16 10 10 10 10 12 12 7 15 15 12 12 12 12 14 1 0 1 0 0 1 1 1 1 1 1 1 1 1 0 0 1 1 1 1 1 10 1 1 0 1 1 1 1 1 1 1 0 0 1 1 1 1 1 1 0 1 1 0 1 1 1 0 0 1 0 1 1 0 0 1 0 0 1 0 2322 2013 687 2533 2845 0 353 1575 838 2588 1750 2188 2392 2353 3245 2288 2262 1125 2212 2366 1390 1510 1535 3213 2175 3411 3035 1138 2100 3533 2825 2050 825 2587 2140 455 3225 2167 2143 2437 2808 2627 950 1587 2505 2082 2466 Hausman test To decide between fixed or random effects we run Hausman test where null hypothesis is that the preferred model is random effects vs the alternative the fixed effects We use the following commands: xtreg slpnap d81 age educ gdhlth marr totwrk yngkid,fe est store fe xtreg slpnap d81 age educ gdhlth marr totwrk yngkid,re hausman fe xtreg totwrk y81 age educ gdhlth marr ,fe note: y81 omitted because of collinearity Fixed-effects (within) regression Group variable: ID Number of obs Number of groups = = 478 239 R-sq: Obs per group: = avg = max = 2.0 within = 0.0829 between = 0.0643 overall = 0.0630 corr(u_i, Xb) F(4,235) Prob > F = -0.2925 totwrk Coef y81 age educ gdhlth marr _cons -40.43996 -20.24237 111.3827 87.00893 3863.725 sigma_u sigma_e rho 808.64708 765.40934 52744828 F test that all u_i=0: Std Err t (omitted) 12.49673 9.832893 137.6032 165.1573 590.2765 -3.24 -2.06 0.81 0.53 6.55 P>|t| 0.001 0.041 0.419 0.599 0.000 = = 5.31 0.0004 [95% Conf Interval] -65.05989 -39.61426 -159.7108 -238.3691 2700.815 -15.82003 -.8704933 382.4762 412.387 5026.634 (fraction of variance due to u_i) F(238, 235) = est store fe 17 2.12 Prob > F = 0.0000 xtreg totwrk y81 age educ gdhlth marr,re Random-effects GLS regression Group variable: ID Number of obs Number of groups = = 478 239 R-sq: Obs per group: = avg = max = 2.0 within = 0.0714 between = 0.0900 overall = 0.0835 corr(u_i, X) Wald chi2(5) Prob > chi2 = (assumed) totwrk Coef Std Err z P>|z| y81 age educ gdhlth marr _cons -120.1202 -17.22136 -7.606707 270.9966 113.0862 2631.825 77.55897 4.511235 7.367247 106.4126 101.8562 250.8902 sigma_u sigma_e rho 532.209 765.40934 32590854 (fraction of variance due to u_i) -1.55 -3.82 -1.03 2.55 1.11 10.49 0.121 0.000 0.302 0.011 0.267 0.000 = = 41.27 0.0000 [95% Conf Interval] -272.133 -26.06322 -22.04625 62.43178 -86.54837 2140.089 31.8926 -8.379504 6.832832 479.5614 312.7207 3123.56 hausman fe Coefficients (b) (B) fe age educ gdhlth marr -40.43996 -20.24237 111.3827 87.00893 (b-B) Difference -17.22136 -7.606707 270.9966 113.0862 -23.2186 -12.63567 -159.6139 -26.07723 sqrt(diag(V_b-V_B)) S.E 11.65406 6.512254 87.24111 130.0086 b = consistent under Ho and Ha; obtained from xtreg B = inconsistent under Ha, efficient under Ho; obtained from xtreg Test: Ho: difference in coefficients not systematic chi2(4) = (b-B)'[(V_b-V_B)^(-1)](b-B) = 10.02 Prob>chi2 = 0.0401 We can see that Prob>chi2 = 0.0401 < 0.05→ Reject → Choose fixed effects model (FE) 18 IV FIXED EFFECT MODEL Fixed effects: comparing xtreg (with fe), regress (OLS with dummies) and areg An unobserved effects model for total minutes of sleeping per week is We use the following commands: xtset ID year panel variable: time variable: delta: ID (strongly balanced) year, 1975 to 1981, but with gaps unit xtreg totwrk educ gdhlth marr,fe Fixed-effects (within) regression Group variable: ID Number of obs Number of groups = = 478 239 R-sq: Obs per group: = avg = max = 2.0 within = 0.0421 between = 0.0063 overall = 0.0163 corr(u_i, Xb) F(3,236) Prob > F = -0.0587 totwrk Coef educ gdhlth marr _cons -23.85516 262.6089 103.3492 2081.878 9.963428 131.9965 168.3603 216.9192 sigma_u sigma_e rho 794.63064 780.61827 50889463 (fraction of variance due to u_i) F test that all u_i=0: Std Err t -2.39 1.99 0.61 9.60 F(238, 236) = 19 P>|t| = = 0.017 0.048 0.540 0.000 1.99 3.45 0.0172 [95% Conf Interval] -43.48378 2.566857 -228.3318 1654.532 -4.226539 522.6509 435.0302 2509.223 Prob > F = 0.0000 estimates store fixed xi: reg totwrk educ gdhlth marr i.ID i.ID _IID_1-239 (naturally coded; _IID_1 omitted) Source SS df MS Model Residual 306880281 143810114 241 236 1273362.16 609364.89 Total 450690395 477 944843.596 Std Err t Number of obs F(241, 236) Prob > F R-squared Adj R-squared Root MSE 478 2.09 0.0000 0.6809 0.3551 780.62 totwrk Coef educ gdhlth marr _IID_2 _IID_3 _IID_4 _IID_5 _IID_6 _IID_7 _IID_8 _IID_9 _IID_10 _IID_11 _IID_12 _IID_13 _IID_14 _IID_15 _IID_16 _IID_17 _IID_18 _IID_19 _IID_20 _IID_21 _IID_22 _IID_23 more -23.85516 262.6089 103.3492 421.5 -937.7253 -9.17458 -432.8194 165.8254 -370.1746 -146.521 -8.964263 -1683.385 704.3254 56.53574 690.4048 437.3481 520.2092 1599.209 -1414.08 248.1151 659.4048 189.1082 -499.5952 282.2827 -616.2461 9.963428 131.9965 168.3603 780.6183 787.9864 785.1441 785.2176 785.1441 785.1441 787.9149 785.4175 785.3763 785.1441 785.4175 786.1138 790.3754 788.9674 788.9674 788.1876 785.3763 786.1138 785.3017 786.1138 780.7613 786.196 -2.39 1.99 0.61 0.54 -1.19 -0.01 -0.55 0.21 -0.47 -0.19 -0.01 -2.14 0.90 0.07 0.88 0.55 0.66 2.03 -1.79 0.32 0.84 0.24 -0.64 0.36 -0.78 0.017 0.048 0.540 0.590 0.235 0.991 0.582 0.833 0.638 0.853 0.991 0.033 0.371 0.943 0.381 0.581 0.510 0.044 0.074 0.752 0.402 0.810 0.526 0.718 0.434 -43.48378 2.566857 -228.3318 -1116.37 -2490.111 -1555.961 -1979.751 -1380.961 -1916.961 -1698.766 -1556.289 -3230.629 -842.4609 -1490.789 -858.2919 -1119.744 -1034.109 44.89069 -2966.863 -1299.129 -889.2919 -1357.989 -2048.292 -1255.869 -2165.105 -4.226539 522.6509 435.0302 1959.37 614.6607 1537.612 1114.112 1712.612 1176.612 1405.724 1538.361 -136.1411 2251.112 1603.861 2239.101 1994.44 2074.528 3153.528 138.7019 1795.359 2208.101 1736.205 1049.101 1820.435 932.6126 _IID_231 _IID_232 _IID_233 _IID_234 _IID_235 _IID_236 _IID_237 _IID_238 _IID_239 _cons 152.1448 -15.58741 578.6151 -642.2461 220.9117 -142.0952 -810.2681 -4.17458 230.4195 2307.399 780.6819 788.0346 785.3763 786.196 788.6152 786.1138 802.9182 785.1441 788.1876 596.6605 0.19 -0.02 0.74 -0.82 0.28 -0.18 -1.01 -0.01 0.29 3.87 0.846 0.984 0.462 0.415 0.780 0.857 0.314 0.996 0.770 0.000 -1385.851 -1568.068 -968.6287 -2191.105 -1332.713 -1690.792 -2392.071 -1550.961 -1322.363 1131.938 1690.14 1536.893 2125.859 906.6126 1774.536 1406.601 771.5345 1542.612 1783.202 3482.86 20 P>|t| = = = = = = [95% Conf Interval] estimates store ols areg totwrk educ gdhlth marr, absorb (ID) Linear regression, absorbing indicators totwrk Coef educ gdhlth marr _cons -23.85516 262.6089 103.3492 2081.878 ID Number of obs F( 3, 236) Prob > F R-squared Adj R-squared Root MSE Std Err t 9.963428 131.9965 168.3603 216.9192 F(238, 236) = P>|t| -2.39 1.99 0.61 9.60 0.017 0.048 0.540 0.000 1.994 0.000 -43.48378 2.566857 -228.3318 1654.532 estimates table fixed ols areg, star stats (N r2 r2_a) educ gdhlth marr _IID_2 _IID_3 _IID_4 _IID_5 _IID_6 _IID_7 _IID_8 _IID_9 _IID_10 _IID_11 _IID_12 _IID_13 _IID_14 _IID_15 _IID_16 _IID_17 _IID_18 _IID_19 _IID_20 _IID_21 more fixed -23.855159* 262.60887* 103.34916 ols -23.855159* 262.60887* 103.34916 421.5 -937.7253 -9.1745801 -432.81942 165.82542 -370.17458 -146.52098 -8.9642629 -1683.3849* 704.32542 56.535737 690.40479 437.34806 520.20922 1599.2092* -1414.0805 248.1151 659.40479 189.10816 -499.59521 21 478 3.45 0.0172 0.6809 0.3551 780.6183 [95% Conf Interval] estimates store areg Variable = = = = = = areg -23.855159* 262.60887* 103.34916 -4.226539 522.6509 435.0302 2509.223 (239 categories) _IID_233 _IID_234 _IID_235 _IID_236 _IID_237 _IID_238 _IID_239 _cons N r2 r2_a 2081.8776*** 478 04205411 -.93618724 578.6151 -642.24605 220.91173 -142.09521 -810.26813 -4.1745801 230.41954 2307.3991*** 478 68091152 35506269 2081.8776*** 478 68091152 35506269 legend: * p 0.05 => The independent variable marr is statistically insignificant 22 Checking and correcting problems of the model a Testing for heteroskedasticity * Heteroskedasticity and its consequences One of OLS assumptions is that Var (u | x1, …, xk) = σ2 That is, the variance of the error term is constant (Homoskedasticity) If the error terms not have constant variance, they are said to be heteroskedastic Unlike multicollinearity, heteroskedasticity violates assumptions of OLS model Heteroskedasticity has serious consequences for the OLS estimator Although the OLS estimator remains unbiased, the estimated standard error is wrong Because of this, confidence intervals and hypotheses tests cannot be relied on In addition, the OLS estimator is no longer BLUE * Causes of heteroskedasticity  Due to the nature of economic phenomena: If the economic phenomena in space is investigated in subjects with different size or economic phenomenon to be investigated over time through the stages with the different fluctuation, error variance can be uneven  Do not format correctly functional form of the model Maybe omissions appropriate variable or function analysis is wrong  Since the data does not reflect the nature of economic phenomena, such as the appearance of foreign observers The inclusion or exclusion of this observation greatly influenced regression analysis  Because technical collection, preservation and processing of data should be improved, error tends to decrease  Human learned in the past * Diagnosis To test the changes of variance, we have the following hypothesis below: where Ho are null hypothesis and H1 is alternative hypothesis There are different methods to test if the model is suffered from heteroskedasticity or not:  Method 1: White test 23 We use the command imtest, white for testing:  Method 2: Breusch-Pagan‘s testing We have a pair of hypothesis: We use the command: xttest3 xttest3 Modified Wald test for groupwise heteroskedasticity in fixed effect regression model H0: sigma(i)^2 = sigma^2 for all i chi2 (239) = Prob>chi2 = 3.1e+35 0.0000 Prob>chi2 = 0.0000 < 0.05 Reject the null hypothesis The fixed effects model has heteroskedasticity To fix the heteroskedasticity, we use the command robust xtreg totwrk educ gdhlth marr, fe robust Fixed-effects (within) regression Group variable: ID Number of obs Number of groups = = 478 239 R-sq: Obs per group: = avg = max = 2.0 within = 0.0421 between = 0.0063 overall = 0.0163 corr(u_i, Xb) F(3,238) Prob > F = -0.0587 = = 39.27 0.0000 (Std Err adjusted for 239 clusters in ID) Robust Std Err totwrk Coef t educ gdhlth marr _cons -23.85516 262.6089 103.3492 2081.878 2.22757 135.4624 160.3955 164.548 sigma_u sigma_e rho 794.63064 780.61827 50889463 (fraction of variance due to u_i) -10.71 1.94 0.64 12.65 a Testing for multicollinearity * Multicollinearity and its Consequences 24 P>|t| 0.000 0.054 0.520 0.000 [95% Conf Interval] -28.24343 -4.249465 -212.627 1757.721 -19.46689 529.4672 419.3253 2406.034 A good model is the model to achieve the characteristics of BLUE (best, linear, unbiased, most effective) But in fact because of some mistakes, the model cannot fully achieve the above properties One of the problems affecting the model is Multicollinearity Though it violates no assumption of OLS model, it affects the accuracy of estimates Multicollinearity is an error of the regression analysis model, occurs due to high (but not perfect) correlation between two or more independent variables The consequences of multicollinearity is that it can increase the variance of the coefficient estimates and make the estimates very sensitive to minor changes in the model The result is that the coefficient estimates are unstable and difficult to interpret Multicollinearity makes it more difficult to specify the correct model * Causes of multicollinearity It is caused by an inaccurate use of dummy variables It’s caused by the conclusion of variable which is computed from other • • variables in the data set • Multicollinearity can also result from the repetition of the same kind variable • Generally occurs when the variables are highly correlated to each other * Diagnosis There are two popular methods to test if the model suffers from multicollinearity:  Method 1: Use command corr to check for multicollinearity Using command corr is more popular than the other method If tindependent variables are strongly correlated with each other (r> 0.8), multicollinearity occurs  Method 2: Use command vif (variance inflation factor) Before using command vif, we must use the command reg 25 reg totwrk educ gdhlth marr Source SS df MS Model Residual 17670193.9 433020201 474 5890064.64 913544.729 Total 450690395 477 944843.596 totwrk Coef educ gdhlth marr _cons -2.304156 453.3369 117.7933 1628.844 Std Err 7.616151 108.1659 100.1731 149.7094 t -0.30 4.19 1.18 10.88 Number of obs F( 3, 474) Prob > F R-squared Adj R-squared Root MSE P>|t| 0.762 0.000 0.240 0.000 vif Variable VIF 1/VIF educ gdhlth marr 1.01 1.01 1.00 0.986595 0.987382 0.996602 Mean VIF 1.01 Mean VIF = 1.01 < 10 so the model has no multi-colinearity Conclusion: The model has no multi-colinearity c Testing for time-fixed effects We use the commands: xi: xtreg totwrk educ gdhlth marr i.year,fe testparm _Iyear* 26 = = = = = = 478 6.45 0.0003 0.0392 0.0331 955.8 [95% Conf Interval] -17.26975 240.7929 -79.04508 1334.668 12.66144 665.8808 314.6316 1923.02 xi: xtreg totwrk educ gdhlth marr i.year,fe i.year _Iyear_1975-1981 (naturally coded; _Iyear_1975 omitted) Fixed-effects (within) regression Group variable: ID Number of obs Number of groups = = 478 239 R-sq: Obs per group: = avg = max = 2.0 within = 0.0829 between = 0.0001 overall = 0.0213 corr(u_i, Xb) F(4,235) Prob > F = -0.0571 totwrk Coef educ gdhlth marr _Iyear_1981 _cons -20.24237 111.3827 87.00893 -242.6397 2286.059 9.832893 137.6032 165.1573 74.98037 221.8544 sigma_u sigma_e rho 799.67256 765.40934 5218818 (fraction of variance due to u_i) F test that all u_i=0: Std Err t -2.06 0.81 0.53 -3.24 10.30 F(238, 235) = P>|t| = = 0.041 0.419 0.599 0.001 0.000 2.12 5.31 0.0004 [95% Conf Interval] -39.61426 -159.7108 -238.3691 -390.3593 1848.981 -.8704933 382.4762 412.387 -94.92016 2723.136 Prob > F = 0.0000 testparm _Iyear* ( 1) _Iyear_1981 = F( 1, 235) = Prob > F = 10.47 0.0014 Prob > F = 0.0014 < 0.05 Reject the null hypothesis Time fixed-effect are needed RECOMMENDATIONS Through the analysis process of the data, it is proved that there are many factors that have significant impact on people’s working time as well as rest time such as: age, the amount of time educated, marriage status, health condition In details, the age can affect the capacity of working as well as productivity, which 27 led to the difference in total working time per week The time spent on education as well as health status also have crucial effects on working time Moreover, the marriage life also have impact on the daily schedule so that people can not have the same amount of working time over years It is essential to have a proper working time while managing the family life happily So, how can we that? First of all, we should plan our schedule properly to balance between jobs and family, not waste time and health on unhelpful activities which don’t help your career or personal life Additionally, we should keep learning even when we graduate and have a job Lenin said “Learn, learn, learn”; learning help us have more and more knowledge and experience in not only academic ways but also in our career and marriage Furthermore, keep fit and stay healthy It is easy to understand: If we are sick, we can not work; if we can not work, we can not earn money to feed our family, which leads to an unhappy family So, we should try our best to have a good health, by limit our drinking, smoking, and other bad habits, also by exercising and relaxing whenever we want to have a healthy mind Moreover, we should not get married and settle a family at early age as by doing so we can be in the dillemma: looking for jobs and taking care of young kids, which will consume our time and harm our health Last but not least, respect our time with family on the weekend and on holidays Let’s relieve ourselves after strenous hours of working, not waste too much time spending on working, enjoy and build up great memories whenever you can It is crucial to have a balanced working life It allows us to keep our control in our career and personal life, it helps us improve our work morale When we are interested in our work and find motivation to complete the tasks, we are inclined to work harder and more efficiently We are also eager to remain our position and develop ourselves Futhermore, balancing working life also makes us become healthier When we sleep and rest enough and work appropriately, we will experience fewer health problems, prevent dangerous diseases To sum up, balancing our work and life is extremely essential, especially in the fast-paced life nowadays to bring us more comfortable and meaningful life, and enhance the living standard for not only us but also our children, our grandchildren, and so on As result of this, we should try to find the proper ways to balance our life so as to enjoy it 28 CONCLUSION Our research shows how minutes spent on working per week is influenced by other determinants These include: age, number of years spent on education, health condition, marriage status The variable gdhlth is the most important variable that affects the dependent variable totwrk We started with descriptive summary of all above variables then all these relations are born clearly in the panel data analysis By accomplishing this report, all members in our group have a precious opportunity to put book knowledge into practice, which gives us a thorough understanding of how to make use of Stata – an useful computer software to run an econometric analysis about social phenomena Namely, in our report, we study the determinants of people’s working time in 1975 and 1981 Our analysis really lives up to our expectation because it fits the common sense and gives employers and managers some hints to recruit the right employees Nevertheless, there are still some drawbacks that need improving to have a more efficient research such as other unobservable factors We would like to express our gratitude to Mrs Binh for her patient and enthusiastic instructions Our report could not be complete and oriented without her assistance Due to the limitation of time and professional knowledge, it is difficult for us to make a perfect report Hence, we really appreciate all your feedbacks to elevate the quality of our assignment and properly apply to other further scientific research 29 REFERENCES The source of Data 5_SLP75_81.DTA collected and given by PhD Dinh Thi Thanh Binh STATA – Data Analysis and Statistical Software for Professionals https://www.stata.com/training/ http://www.businessdictionary.com/definition/fixed-effects-model.html “Introductory Econometrics A Modern Approach” – the 5th edition by Jeffrey M.Wooldridge Introduction to Regression Models for Panel Data Analysis by Professor Patricia A McManus http://www.indiana.edu/~wim/docs/10_7_2011_slides.pdf Panel Data Analysis – Fixed & Random Effects (using Stata 10.x) (ver 4.1) by Oscar Torres-Reyna (Data Consultant) https://www.princeton.edu/~otorres/Panel101.pdf STATA – Data Analysis and Statistical Software for Professionals https://www.stata.com/training/ Bruce E Hansen (University of Wisconsin - Department of economics, ECONOMETRICS, 2017 (https://www.ssc.wisc.edu/~bhansen/econometrics/Econometrics.pdf) “FIXED-EFFECTS METHODS FOR THE ANALYSIS OF NONREPEATED EVENTS” by Paul D.Allison and Nicholas A Christakis https://statisticalhorizons.com/wpcontent/uploads/AllisonChristakisSM06.pdf 30 ... 29 05 25 77 3080 27 22 2815 325 0 24 33 26 50 21 63 25 32 940 329 7 26 01 23 00 22 06 1457 650 20 20 1675 23 87 28 80 22 37 1563 837 26 06 1505 25 38 26 37 20 26 1100 26 98 22 1 22 2 22 2 22 3 22 3 22 4 22 4 22 5 22 5 22 6 22 6... 22 6 22 7 22 7 22 8 22 8 22 9 22 9 23 0 23 0 23 1 23 1 23 2 23 2 23 3 23 3 23 4 23 4 23 5 23 5 23 6 23 6 23 7 23 7 23 8 23 8 23 9 23 9 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975. .. 21 4 21 4 21 5 21 5 21 6 21 6 21 7 21 7 21 8 21 8 21 9 21 9 22 0 22 0 22 1 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981 1975 1981