Assignment Document

Interpretation of Independent Variables Discrete Variables

Pages:

Preview:


  • "Answer 1aMean and standard deviations of Model A and Model CMean Std. Dev.SURVEY 2 0Age 37.51296 13.49053Sqage 1589.213 1083.361Schooling 5.193202 4.985803schooling_sq 51.82716 65.93796------------- ----------- ------------log_hhsize 1.646247 0.4737..

Preview Container:


  • "Answer 1aMean and standard deviations of Model A and Model CMean Std. Dev.SURVEY 2 0Age 37.51296 13.49053Sqage 1589.213 1083.361Schooling 5.193202 4.985803schooling_sq 51.82716 65.93796------------- ----------- ------------log_hhsize 1.646247 0.473795duch 0.052373 0.22278dobc 0.403022 0.490509dst 0.088632 0.284215dosgr 0.250191 0.433127------------- ----------- ------------dhindu 0.803032 0.397711dmuslim 0.131766 0.338239share_fde~60 0.023226 0.085925share_mde~60 0 0share_nch~15 0.164899 0.184779------------- ----------- ------------share_nch~_5 0.095541 0.134038dmarr 0.11019 0.31313share_fem~59 0.349624 0.161154rank1 0.274574 0.446303rank2 0.293665 0.455444------------- ----------- ------------rank3 0.363632 0.481049heduc2 0.052645 0.223326heduc3 0.220075 0.4143heduc4 0.219666 0.414024heduc5 0.145565 0.352673------------- ----------- ------------heduc6 1 0psu_sh_wka~g 0.100298 0.12215psu_sh_wka~m 0.173952 0.123672psu_sh_wkb~s 0.129151 0.155762psu_sh_wkf~m 0.201673 0.176357------------- ----------- ------------psu_sh_wkn~g 0.172363 0.144933psu_sh_wkn~a 0.023263 0.045231psu_sh_wks~y 0.199301 0.231395 Variant 1 Variant 2dwork Coef. dF/dx Dwork Coef. dF/dxAge 0.104133*** 0.0396338*** Age 0.133749*** 0.050457***sqage -0.00129*** 0.0004921*** Sqage -0.0016*** -0.0006***schooling -0.09079*** 0.0345554*** Schooling -0.08088*** -0.03051***schooling_sq 0.004037*** 0.0015367*** schooling_sq 0.005235*** 0.001975***log_hhsize -0.25654*** 0.0976407*** log_hhsize -0.28893*** -0.109***duch -0.31681*** 0.1139126*** Duch -0.29721*** -0.106***dobc 0.008466 0.0032231 Dobc -0.09607*** -0.03612***Dst 0.465955*** 0.1830277*** Dst 0.210631*** 0.08135***dosgr -0.10191*** 0.0384599*** Dosgr -0.19436*** -0.07196***dhindu 0.211792*** 0.0787472*** Dhindu 0.074096*** 0.027731***dmuslim -0.3647*** -0.131421*** Dmuslim -0.26243*** -0.09518***share_fdep_60 0.425152*** 0.1618156*** share_fdep_60 0.413048*** 0.155824***share_nchild6_15 0.754875*** 0.2873104*** share_nchild6_15 0.813486*** 0.30689***share_nchild0_5 0.276528*** 0.1052485*** share_nchild0_5 0.269739*** 0.10176***dmarr 0.470407*** 0.1845668*** Dmarr 0.574854*** 0.224883***share_fem15_59 0.389531*** 0.1482582*** share_fem15_59 0.45577*** 0.171941***heduc2 0.24314*** 0.0948914*** rank1 0.074935*** 0.028419***heduc3 0.15464*** 0.0595323*** rank2 -0.00868 -0.00327heduc4 0.026606* 0.010149* rank3 -0.00144 -0.00054heduc5 0.002468 0.0009396 heduc2 0.184395*** 0.071172***_cons -2.0003*** heduc3 0.136104*** 0.051929*** heduc4 -0.01364 -0.00514 heduc5 -0.03207* -0.01205* psu_sh_wkagwag 1.436759*** -0.85831*** psu_sh_wkanim -0.98383*** -1.77149*** psu_sh_wkbsns 0.266143*** -1.29993*** psu_sh_wkfarm 2.66912*** -0.3934*** psu_sh_wknonag 0.330833*** -1.27553*** psu_sh_wknrega 3.711926*** -1.40034*** _cons -3.26304****** 1% LEVEL OF SIGNIFICANCE ** 5% LEVEL OF SIGNIFICANCE* 10% LEVEL OF SIGNIFICANCEINTERPRETATION OF INDEPENDENT VARIABLESDISCRETE VARIABLESAGE : For a year increase in age, the probability of participation in labour force increases by0.050457%.Marginal effects in probit model is defined as:? ? =?(X ß)ßi i ? ? ? If we use Linear probability model, for every unit increase in the age, the probability increases by a constant0.133749, coefficient of age ß .i We are not using LPM because of the following limitations:1) Heteroscedasticity.2) Non normality of errors.3) Probability may be greater than 1 and it may be negative. 4) Linearity in variables.Age square:The coefficient of age square is negative(-0.0016)i.e. the curve for age and labour forceparticipation is concave and upward sloping. This is because after a certain age the productivity of womandecreases which reduces its probability to participate in labour force.Schooling: For a year increase in schooling, the probability of women participation decreases by0.03051%.This may be because with an increase in time devoted towards schooling, the time left for workingdecreases.Schooling Square: he coefficient of schooling square is positive (0.001975) i.e. the curve for schooling andlabour force participation is convex and downward sloping. This is because after a certain level of educationis achieved, for each year increase in schooling, participation in the labour force increases. Log household size:For a unit increase in the number of members living under the same roof, theprobability of participation in labour force decreases by 0.109% as it increases the responsibility of womentowards the household members.DUMMY VARIABLES:For all the categories interpretations are same with respect to base category. Only one dummy is interpretedunder each head. 1) Caste variables: duch, dobc, dst, dosgr (SC is the base category) duch is dummy for Brahmincategory, dobc is dummy for obc category, dst is dummy for scheduled tribe, dsc is dummy forscheduled caste,dosgr is dummy for other categories.Duch:Cetrisperibus, for a woman belonging to Brahmin category, the labour force participation is 0.106%less than the women belonging to SC category.2) Marital status : with reference category as married. Cetrisperibus, for a women who is unmarried,the labour force participation is 0.22488% more than the women who is married.3) Population (rank1 , rank 2, rank 3): Reference category as metro urban. Rank 1 is other urban areas, rank 2 is more developed village, rank 3 is less developed village.Rank 1:Cetrisperibus, for a person belonging to other urban areas, the labour force participation is0.028419% more than the person belonging to metro urban.4) Education(hhedu2,hhedu3,hhedu4,hhedu5) with reference category as hhedu6,post higher secondaryhhedu2 is primary education, hhedu3 is middle education, hhed4 is secondary education, hhed5 ishigher secondary. hhedu2 :Cetrisperibus, for a adult who has completed primary education, the labour force participation is0.071172% more than the adult who has completed post high secondary.5) Religion(dhindu,dmuslim) reference category is other religionsDhindu: Cetrisperibus, for a adult who is hindu, the labour force participation is 0.027731% morethan the adult who belongs to some other religion. PROPORTION VARIABLES:Share of female adults in the household aged 15-59: (share_fem15_59): Cetrisperibus, if the percentage offemale population between 15 to 59 increases by 1% then labor force participation increases by 0.17% as thepopulation of working age is increasing.1) Share of female dependents (aged 60+) in the household: (share_fdep_60): with base categoryas male dependent. Cetrisperibus,Labour force participation of dependent female is 0.155%more than dependent male. 2) Share of Children: (between 0 to 5 years and between 6 to 15 years)Between 6 years to 15 years in the household:cetrisperibus,If the percentage of children between 6 to 15increases by 1% then labour force participation increases by 0.30689%.3) Share of working hours in different occupations:(psu_sh_wkagwagpsu_sh_wkanimpsu_sh_wkbsnspsu_sh_wkfarmpsu_sh_wknonagpsu_sh_wknregapsu_s h_wksalry)psu_sh_farm: Cetrispreribus,If the percentage of working hours in farming increases by 1%then labourforce participation decreases by 0.3934% more than the working hours of salaried person.VARIANT 1 OR VARIANT 2: WHICH IS BETTER ?Variant 1 and variant 2 can be comparedon the basis of following variables:2 2 2 1)Pseudo R :Pseudo R of unrestricted model is greater than restricted model. Thus according to pseudo Rvariant 2 is better i.e. unrestricted model.pseudo R^2(UR)= 17.03 > pseudo R^2(R) =8.94%2)LR Test: In LR test we use the following hypothesis :H :Rß(hat)= ?0H1: Rß(hat)? ? Where R is the matrix in which number of rows is equal to number of restrictions. ? is a matrix ofexplanatory variables.LR test is rejected which means that unrestricted model with population based on region and employmentvariables is better than restricted model.3)Wald test: Wald test is significant in both the models but we take the variant in which our added variablesturn out be significant.Changes in results from variant 1 to variant 2:1) dobc was insignificant in variant1 but significant in variant 2.2) Hhedu4 was significant at 10% level ofsignificance in variant 1 but insignificant in variant 2.3) Rank 1which is dummy for population in other urban is significant in variant 2 but rank 2 and rank3 which represents population in more developed and less developed villages is insignificant.4) Share of work done in farm, non agriculture, business and other persons is significant in unrestrictedmodel.Answer 1bVariant 3dwork coefficiennts dF/dxAge 0.137521*** 0.051825***Sqage -0.00162*** -0.00061***dedu_pri -0.0194 -0.00729dedu_mid -0.14901*** -0.05533***dedu_sec -0.28217*** -0.10217***dedu_hisec -0.1351*** -0.04981***dedu_phisec 0.265034*** 0.10283***log_hhsize -0.23129*** -0.08716***Duch -0.2806*** -0.10038*** Dobc -0.09179*** -0.03447***Dst 0.206089*** 0.079501***Dosgr -0.18675*** -0.06911***Dhindu 0.079198*** 0.02959***Dmuslim -0.27383*** -0.09898***share_fdep_60 0.366443*** 0.138094***share_nchild6_15 0.709169*** 0.26725*** share_nchild0_5 0.227985*** 0.085916***Dmarr 0.554454*** 0.216831***share_fem15_59 0.460807*** 0.173655***rank1 0.077239*** 0.029268***rank2 -0.01101 -0.00415rank3 -0.00837 -0.00315heduc2 0.023586 0.008919heduc3 -0.04345** -0.0163** heduc4 -0.18704*** -0.06907***heduc5 -0.22274*** -0.08134***heduc6 -0.3435*** -0.12436***psu_sh_wkagwag 1.376874*** -0.87863***psu_sh_wkanim -1.00706*** -1.77701***psu_sh_wkbsns 0.238183*** -1.30774****** 1% LEVEL OF SIGNIFICANCEpsu_sh_wkfarm 2.652979*** -0.39773***** 5% LEVEL OF SIGNIFICANCEpsu_sh_wknonag 0.272414*** -1.29484**** 10% LEVEL OF SIGNIFICANCE psu_sh_wknrega 3.708376*** -1.3975***2 Variant 3 is better than variant 2 as it pseudo R is_cons -3.29945***2 2 higher(new R = 17.33>R of variant 2=17.03)Probability of a female with primary education being a part of labour forceis 0.00729%less than femalewith no education.Probability of a female with middle education being a part of labour forceis 0.05532% less than female with no education.Probability of a female with secondary education being a part of labourforceis 0.102165%less than female with no education.Probability of a female with higher secondaryeducation being a part of labour forceis 0.04981%less than female with no education.Probability of afemale with post higher secondary education being a part of labour forceis 0.10289%more than femalewith no education.The interpretation of schooling and schooling square won’t be same as when education is taken as a dummyvariable. The interpretation of schooling and dummy education has been stated above.0 5 10 15 schooling metro urban 0 other urban 1 more del vill 2 less dev vill 3 Fitted values .1 .2 .3 .4 .5 .6All these curves are U-shaped curves. At initial level of education, probability of women entering labourforce is more in less developed villages. As the number of years of education increases, the probability ofwomen joining labour force decreases till a threshold level (higher secondary) then it increases (post highersecondary level of education) .This is also consistent with the above sign of estimated coefficients as sign of estimated coefficients ofvariable education years is negative. But if we square education years it becomes positive which implies thatas number of years of schooling increases, squared coefficients starts dominating the negative impact oflinear coefficients.Answer 2aWe have selected variant 2as the benchmark model since in answer 1a we got variant 2 as the best fittedmodel when compared to variant 1Model 2adwork coff dF/dxAge 0.137908*** 0.052018***Sqage -0.00163*** -0.00061***Schooling -0.07218*** -0.02723***schooling_sq 0.005382*** 0.00203***log_hhsize -0.23398*** -0.08825*** Duch -0.27709*** -0.09925***Dobc -0.09093*** -0.03419***Dst 0.20717*** 0.07998***Dosgr -0.18459*** -0.0684***Dhindu 0.072168*** 0.027008***Dmuslim -0.27982*** -0.10116***share_fdep_60 0.361741*** 0.136446***share_nchild6_15 0.710572*** 0.268023*** share_nchild0_5 0.229044*** 0.086394***Dmarr 0.560584*** 0.219299***share_fem15_59 0.460002*** 0.17351***rank1 0.077176*** 0.029269***rank2 -0.0111 -0.00419rank3 -0.0098 -0.00339heduc2 0.06384** 0.024298** heduc3 -0.00758 -0.00286heduc4 -0.18881*** -0.06978***heduc5 -0.2283*** -0.08339***heduc6 -0.30962*** -0.11275***psu_sh_wkagwag 1.389675*** -0.87312***psu_sh_wkanim -1.02171*** -1.78268***psu_sh_wkbsns 0.241302*** -1.30628***psu_sh_wkfarm 2.652288*** -0.39687***psu_sh_wknonag 0.27725*** -1.29272***psu_sh_wknrega 3.704462*** -1.3973***citerc_pcag2 0.00808 0.003049citerc_pcag3 -0.01864 -0.00702 *** 1% LEVEL OF SIGNIFICANCE** 5% LEVEL OF SIGNIFICANCE* 10% LEVEL OF SIGNIFICANCEAnswer 2b Variant 2bDwork Coef. dF/dxAge 0.138506*** 0.052242***Sqage -0.00163*** -0.00062***Schooling -0.07156*** -0.02699***schooling_sq 0.005359*** 0.002021***log_hhsize -0.23322*** -0.08797***Duch -0.2802*** -0.10029***Dobc -0.09037*** -0.03397***Dst 0.212597*** 0.082111***Dosgr -0.18269*** -0.06771***Dhindu 0.06975*** 0.026112***Dmuslim -0.27722*** -0.10026***share_fdep_60 0.366277*** 0.138153***share_nchild6_15 0.706772*** 0.266582***share_nchild0_5 0.223747*** 0.084393***Dmarr 0.562128*** 0.219901***share_fem15_59 0.458013*** 0.172754***rank1 0.088541*** 0.033603***rank2 0.004725 0.001783rank3 0.005928 0.002237heduc2 0.064374** 0.024502**heduc3 -0.00584 -0.0022heduc4 -0.18394*** -0.06802***heduc5 -0.22443*** -0.08202***heduc6 -0.30588*** -0.11144***psu_sh_wkagwag 1.386381*** -0.86287***psu_sh_wkanim -1.0404*** -1.77822***psu_sh_wkbsns 0.205054*** -1.30845***psu_sh_wkfarm 2.626874*** -0.39498***psu_sh_wknonag 0.270753*** -1.28367***psu_sh_wknrega 3.674066*** -1.38579***psu_sh_cig2 0.169458*** 0.063917***psu_sh_cig3 -0.01919 -0.00724_cons -3.35069*** *** 1% LEVEL OF SIGNIFICANCE** 5% LEVEL OF SIGNIFICANCE* 10% LEVEL OF SIGNIFICANCEModel 2(b) is better than 2(a) because 2 2 1) Pseudo R in 2(a) is 0.1721 which is less than pseudo R of model 2(b) which is 0.1726 2) F statistics is significant in both the cases.3) Individual confidence index,citerc_pcag2 is insignificant whereas aggregate confidence psu_sh_cig2is significant.There are very minute changes in Model 2(b) as compared to model 2(a).Answer3:Hypothesis:H : Rß(hat)= ?0 H1: Rß(hat)? ?Where R is the matrix in which number of rows is equal to number of restrictions. ß is a matrix ofexplanatory variables.Number of restrictions (R) :psu_sh_wkagwag=psu_sh_wkanim=psu_sh_wkbsns=psu_sh_wkfarm=psu_sh_wknonag=psu_sh_wknrega= 0Answer4:Dwork coeff. dF/dxAge 0.143636*** 0.054048***Sqage -0.0017*** -0.00064***Schooling -0.07204*** -0.02711***schooling_sq 0.005407*** 0.002035***log_hhsize -0.22624*** -0.08513***Duch -0.31968*** -0.11301***Dobc -0.11439*** -0.04286***Dst 0.176129*** 0.067678***Dosgr -0.18617*** -0.06879***Dhindu -0.00071 -0.00027Dmuslim -0.32724*** -0.1169***share_fdep_60 0.347405*** 0.130724***share_nchild6_15 0.722579*** 0.271897***share_nchild0_5 0.236435*** 0.088967***Dmarr 0.583505*** 0.228069***share_fem15_59 0.445359*** 0.167582***rank1 0.028586 0.01078rank2 0.07615* 0.028799*rank3 0.13409*** 0.05073***heduc2 0.064144** 0.024361**heduc3 -0.06005*** -0.02246***heduc4 -0.22396*** -0.08218***heduc5 -0.28978*** -0.10443***heduc6 -0.36286*** -0.13076***psu_sh_wkagwag 1.200501*** -0.31491***psu_sh_wkanim -1.00125*** -1.1434***psu_sh_wkbsns 0.215918*** -0.6854***psu_sh_wkfarm 2.274738 0.089306psu_sh_wknonag 0.445892*** -0.59886***psu_sh_wknrega 2.037403*** -0.76665*** psu_sh_cig2 0.181856*** 0.06843***psu_sh_cig3 0.118503*** 0.044591***S1 -0.02383*** -0.00897***S2 0.477447*** 0.187338***S3 -0.28477*** -0.10136***S4 -0.10084 -0.03724S5 0.184394*** 0.07117***S6 -0.18588*** -0.06763***S7 -0.13422** -0.04928**S8 0.150959*** 0.057905***S9 -0.21728*** -0.07894***S10 -0.4824*** -0.16303***S11 0.053722 0.020391S12 -0.11737 -0.0432S13 -0.11087 -0.04086S14 -0.58367*** -0.18904***S15 0.42981*** 0.168793***S16 -0.61661*** -0.19769***S17 -0.01504 -0.00564S20 -0.51085*** -0.17041***S21 -0.49859*** -0.16844***S22 0.280372*** 0.10905***S23 0.124674** 0.047675**S24 -0.06487 -0.02415S25 -0.31598** -0.11093**S26 -0.02832 -0.0106S27 -0.00658 -0.00247S29 0.098185* 0.037402*S30 -0.69323*** -0.21632***S32 -0.19171*** -0.06962***S33 0.007957 0.002998S34 -0.14955 -0.05466_cons -3.25519*** 1% LEVEL OF SIGNIFICANCE ** 5% LEVEL OF SIGNIFICANCE* 10% LEVEL OF SIGNIFICANCEWith stateWithout state2 2 1) Pseudo R =0.1945(higher) 1) Pseudo R = 0.17262) Hindu dummy is insignificant2) Hindu dummy is significant3) Rank1 is insignificant. Rank 2 is significant at 3) Rank 1 significant 10% level of significance. Rank 2 and Rank 3 are insignificant.Rank3significant 4) Hhedu3 is significant4) Hhedu3 is insignificant 5) Psu_sh__farm is insignificant5) Psu_sh_wkfarm is significant6) Psu_sh_cig3 is significant 6) Psu_sh_cig3 is insignificant7) S4,S11,S12,S13,S17,S24,S26,S27,S33,S34 areinsignificant LR test of 2b model is given as :Hypothesis:H : Rß(hat)= ?0 H1: Rß(hat)? ?Where R is the matrix in which number of rows is equal to number of restrictions. ß is a matrix ofexplanatory variables.LR test is significant in this case which means that model 2b i.e. model with state variables is better.dwork Coef. St dF/dxAge 0.146882*** 0.055227***Sqage -0.00172*** -0.00065***Schooling -0.06511*** -0.02448***schooling_sq 0.005374*** 0.002021***log_hhsize -0.1799*** -0.06764***Duch -0.27695*** -0.09877***Dobc -0.08844*** -0.03314***Dst 0.161024*** 0.061743***Dosgr -0.14477*** -0.05368***Dhindu -0.01281 -0.00482Dmuslim -0.34174*** -0.12161***share_fdep_60 0.341716*** 0.128483***share_nchild6_15 0.683492*** 0.256988***share_nchild0_5 0.173436*** 0.06521***Dmarr 0.5583*** 0.218163***share_fem15_59 0.470901*** 0.177055***rank1 0.022786 0.008582rank2 0.056913 0.021481rank3 0.099072* 0.037407*heduc2 0.075489*** 0.028693***heduc3 -0.03012 -0.01129heduc4 -0.1672*** -0.06173***heduc5 -0.20807*** -0.07595***heduc6 -0.24547*** -0.0898***psu_sh_wkagwag 1.037421*** -0.32425***psu_sh_wkanim -1.09083*** -1.12446***psu_sh_wkbsns 0.19786*** -0.63992***psu_sh_wkfarm 2.161353 0.09834psu_sh_wknonag 0.327349*** -0.59123***psu_sh_wknrega 1.899806*** -0.71431***psu_sh_cig2 0.178726*** 0.0672***psu_sh_cig3 0.117805*** 0.044294***S1 -0.02743*** -0.01031***S2 0.51426*** 0.201812***S3 -0.1982*** -0.07184***S4 -0.08143 -0.03016S5 0.202398*** 0.078224***S6 -0.13091** -0.04811** S7 -0.09345 -0.03455S8 0.158121*** 0.06066***S9 -0.28074*** -0.10066***S10 -0.56597*** -0.18641***S11 0.009943 0.003745S12 -0.1186 -0.0436S13 -0.0303 -0.01133S14 -0.6564*** -0.20706***S15 0.55806*** 0.219308***S16 -0.67494*** -0.21174***S17 -0.04214 -0.01573S20 -0.57238*** -0.18708***S21 -0.56324*** -0.18661***S22 0.265404*** 0.103057***S23 0.088454* 0.033657*S24 -0.05422 -0.02021S25 -0.25675 -0.09143S26 -0.01153 -0.00433S27 -0.00701 -0.00263S29 0.084184* 0.031994*S30 -0.70135*** -0.21789***S32 -0.15088** -0.0552**S33 0.041556 0.015724S34 -0.07673 -0.02845ai_quin1 0.475612*** 0.184749***ai_quin2 0.334917*** 0.129195***ai_quin3 0.310935*** 0.119661***ai_quin4 0.148747*** 0.056654***_cons -3.63705*** 1% LEVEL OF SIGNIFICANCE** 5% LEVEL OF SIGNIFICANCE* 10% LEVEL OF SIGNIFICANCE2 1) Pseudo R is 0.1987 when we include ai_index which is higher than .1945 (model without ai_index).2) Rank 2 is now significant at all the level of significance whereas earlier it was significant only at10% level of significance.3) Psu_sh_farm is now significant which was insignificant earlier.4) S7 and S25 turned out to be insignificant with ai_index but was significant earlier.5) Individual coefficients turned out to be significant at all the level of significance.6) LR test of 4b model is given as :Hypothesis:H : Rß(hat)= ?0H1: Rß(hat)? ?Where R is the matrix in which number of rows is equal to number of restrictions. ß is a matrix ofexplanatory variables.LR test is significant in this case which means that model 4b is better than any other model. Answer 4cDwork Coef. dF/dxAge 0.143735*** 0.054081***Sqage -0.0017*** -0.00064***schooling -0.0725*** -0.02728***schooling_sq 0.005383*** 0.002025***log_hhsize -0.21494*** -0.08087***Duch -0.32308*** -0.11411***Dobc -0.11434*** -0.04283***Dst 0.186799*** 0.071847***Dosgr -0.18955*** -0.07001***Dhindu 0.001392 0.000524Dmuslim -0.32213*** -0.11517***share_fdep_60 0.368089*** 0.138495***share_nchild6_15 0.771784*** 0.290387***share_nchild0_5 0.28941*** 0.108892***Dmarr 0.594441*** 0.232324***share_fem15_59 0.445125*** 0.16748***rank1 0.034888 0.013161rank2 0.083876** 0.031733**rank3 0.14716*** 0.055695***heduc2 0.062781** 0.023837**heduc3 -0.06558*** -0.02451***heduc4 -0.23768*** -0.08705***heduc5 -0.3073*** -0.11039***heduc6 -0.38864*** -0.13951***psu_sh_wkagwag 1.246085*** -0.32454***psu_sh_wkanim -0.966*** -1.15684***psu_sh_wkbsns 0.229066*** -0.70719***psu_sh_wkfarm 2.32377 0.080948psu_sh_wknonag 0.458544*** -0.62085***psu_sh_wknrega 2.10863*** -0.79338***psu_sh_cig2 0.183422*** 0.069013***psu_sh_cig3 0.111982*** 0.042134***S1 -0.02192*** -0.00825***S2 0.481385*** 0.188889***S3 -0.27867*** -0.09932***S4 -0.08387 -0.03107S5 0.202938*** 0.078476***S6 -0.17557*** -0.06401***S7 -0.11279* -0.04158*S8 0.167492*** 0.064354***S9 -0.16959*** -0.06214***S10 -0.43108*** -0.14775***S11 0.06085 0.02312S12 -0.12244 -0.04501 S13 -0.09814 -0.03626S14 -0.56785*** -0.18486***S15 0.418156*** 0.164155***S16 -0.59693*** -0.19262***S17 -0.01398 -0.00525S20 -0.46003*** -0.15576***S21 -0.44207*** -0.15161***S22 0.332832*** 0.129908***S23 0.170469*** 0.065503***S24 -0.02968 -0.01111S25 -0.29712* -0.10481*S26 0.007494 0.002823S27 0.0167 0.006298S29 0.128111** 0.048959**S30 -0.66032*** -0.20843***S32 -0.17036*** -0.06214***S33 0.028113 0.010623S34 -0.14279 -0.05227pcinq1 0.158487*** 0.060482***pcinq2 0.263106*** 0.101103***pcinq3 0.308415*** 0.118788***pcinq4 0.328883*** 0.126849***pcinq5 0.340363*** 0.1315***_cons -3.61596*** 1% LEVEL OF SIGNIFICANCE ** 5% LEVEL OF SIGNIFICANCE* 10% LEVEL OF SIGNIFICANCEThe reason why authors have not used income variables in the model since the proxy for index variable ismonthly expenditure which is endogenous in the model and thus cannot be used as a explanatory variable. The asset index variable captures the economic status and standard of living of the households and per capitaincome captures the income of households which further determines the expenditure incurred by them.Hence ai index percapita income captures the effect of income and thus can be used to counter the argumentsgiven in the paper.Answer 4d:dwork Coef. St dF/dxAge 0.145094*** 0.054533***Sqage -0.00171*** -0.00064***Schooling -0.06927*** -0.02603***schooling_sq 0.005378*** 0.002021***log_hhsize -0.26996*** -0.10146***Duch -0.29817*** -0.10576***Dobc -0.1146*** -0.04288***Dst 0.164978*** 0.063264***Dosgr -0.19588*** -0.0722*** Dhindu 0.003566 0.00134Dmuslim -0.30432*** -0.10901***share_fdep_60 0.4924838*** 0.185097***share_nchild6_15 0.80462*** 0.302412***share_nchild0_5 0.348019*** 0.130801***Dmarr 0.584609*** 0.228412***share_fem15_59 0.513075*** 0.192836***rank1 0.032669 0.012309rank2 0.086738** 0.032788**rank3 0.137274*** 0.051883***heduc2 0.059548** 0.022577**heduc3 -0.0593*** -0.02215***heduc4 -0.21473*** -0.07877***heduc5 -0.26784*** -0.09675***heduc6 -0.32201*** -0.11654***psu_sh_wkagwag 0.868196*** -0.32992***psu_sh_wkanim -1.13376*** -1.08234***psu_sh_wkbsns 0.156681*** -0.59734***psu_sh_wkfarm 2.004788 0.097265psu_sh_wknonag 0.381771*** -0.51274***psu_sh_wknrega 1.745998*** -0.65622***psu_sh_cig2 0.210114*** 0.07897***psu_sh_cig3 0.138944*** 0.052221***S1 -0.02696*** -0.01013***S2 0.506748*** 0.19883***S3 -0.33109*** -0.11635***S4 -0.14815 -0.05409S5 0.152494** 0.058591**S6 -0.23537*** -0.08462***S7 -0.18148*** -0.06587***S8 0.121072** 0.046241**S9 -0.26913*** -0.09667***S10 -0.5147*** -0.17201***S11 0.018231 0.006873S12 -0.17848 -0.06473S13 -0.13174 -0.04828S14 -0.57844*** -0.18728***S15 0.427551*** 0.167822***S16 -0.64797*** -0.20506***S17 -0.07631 -0.02828S20 -0.54556*** -0.1797***S21 -0.54909*** -0.18254***S22 0.222466*** 0.086035***S23 0.07279 0.027633S24 -0.12534** -0.04611**S25 -0.35767** -0.12393**S26 -0.06657 -0.02472S27 -0.06945 -0.02583 "

Related Documents

Start searching more documents, lectures and notes - A complete study guide!
More than 25,19,89,788+ documents are uploaded!

Why US?

Because we aim to spread high-quality education or digital products, thus our services are used worldwide.
Few Reasons to Build Trust with Students.

128+

Countries

24x7

Hours of Working

89.2 %

Customer Retention

9521+

Experts Team

7+

Years of Business

9,67,789 +

Solved Problems

Search Solved Classroom Assignments & Textbook Solutions

A huge collection of quality study resources. More than 18,98,789 solved problems, classroom assignments, textbooks solutions.

Scroll to Top