EXERCISE 2 - ANSWER

Generally speaking (with a normal labor supply curve) people are willing to work more when they get paidmore after taxes. Investigate this relationship by regressing "hours worked last week" on "household net wage." Have you estimated a labor supply function?

lookfor hours worked

41. k_hours   float  %9.0g                  3d:hours per day
74. hours_wo  float  %9.0g                  5 :hours worked last week
89. hours_v   float  %9.0g                  3d:hours worked average day
90. hours_ot  byte   %9.0g                  3e:overtime hours past week

From this we can see that "hours worked last week" is hours_wo.

lookfor household net

   1. hhid      float  %9.0g                  household identification no
  21. q7b       byte   %9.0g                  7b:total household per unit
  44. p_netpay  float  %9.0g                  4d:take home pay
 214. mxhous    float  %9.0g                  household exp.
 227. homewage  float  %9.0g                  household housing subsidy
 238. travwage  float  %9.0g                  household travel subsidy
 249. hhtfexp   float  %9.0g                  household tot food month exp.
 272. hhnwage   float  %9.0g                  household net wage
 273. hhgwage   float  %9.0g                  household gross wage

From this we can see that "household net wage" is hhnwage. Now we try the regression:

regress hours_wo hhnwage

  Source |       SS       df       MS                  Number of obs =     976
---------+------------------------------               F(  1,   974) =   26.68
   Model |  14212.7911     1  14212.7911               Prob > F      =  0.0000
Residual |   518855.36   974  532.705709               R-squared     =  0.0267
---------+------------------------------               Adj R-squared =  0.0257
   Total |  533068.151   975  546.736566               Root MSE      =   23.08
------------------------------------------------------------------------------
hours_wo |      Coef.   Std. Err.       t     P>|t|       [95% Conf. Interval]
---------+--------------------------------------------------------------------
 hhnwage |   .0010839   .0002098      5.165   0.000       .0006721    .0014957
   _cons |   16.19218   .7784633     20.800   0.000       14.66453    17.71984
------------------------------------------------------------------------------

By reading this table we can find some interesting results. The coefficient on hhnwage is .0010839 so this means that if the net wage were to go up by one rand, the number of hours worked in a week would go up by .0010839 (or approximately 3.9 seconds). Better said, if the wage were to go up by 100 rand, the number of hours worked in a week would go up by 1 hour and 5 minutes. Because the t statistic for hhnwage is significant (5.165>2) then we can believe that these results did not just happen by chance. But have we estimated a labor supply curve? Not really. We have investigated the relationship between the equilibrium number of hours worked and the wage, but this could be a supply function, a demand function, or, more likely, a mix of the two. Just because the relationship between hours and wage is positive does not mean that this is a supply function.