This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: Eco 572: Research methods in Demography Solutions to Problem Set 2 [1] U.S. Mortality (a) Calculate agespecific period mortality rates for 1992 for blacks and whites . insheet using http://data.princeton.edu/eco572/datasets/us1992a.dat (5 vars, 19 obs) . gen mw = deathsw/expow . gen mb = deathsb/expob . list age mw mb ++  age mw mb   1.  0 .007809 .019579  2.  1 .000426 .000776  3.  5 .000213 .000375  4.  10 .000282 .000449  5.  15 .00106 .002184   6.  20 .001354 .00321  7.  25 .001533 .003617  8.  30 .001958 .004644  9.  35 .002455 .006096  10.  40 .003122 .008032   11.  45 .004325 .010657  12.  50 .006634 .014193  13.  55 .010715 .021036  14.  60 .017297 .029243  15.  65 .026885 .040291   16.  70 .040124 .0572491  17.  75 .061488 .07502  18.  80 .097005 .1096976  19.  85 .179562 .1671715  ++ (b) Plot the rates against age on a log scale. Exclude the last group, which is actually 85+. . gen agem = (age+age[_n+1])/2 // excludes 85+ (1 missing value generated) . line mw mb agem, title(U.S. Males 1992) /// > yscale(log) ytitle(asmr) xtitle(age) /// > legend(order(1 "White" 2 "Black") ring(0) pos(5) cols(1)) http://data.princeton.edu/eco572/ps2sol.html (1 of 14) [2/11/2008 2:16:28 PM] Eco 572: Research methods in Demography . graph export ps2fig1.png, replace (file ps2fig1.png written in PNG format) (c) Fit a Gompertz model over an appropriate age range. Seems clear that black mortality rates are Gompertz over a wider range than those for whites, which seeem to show a change of slope around age 50. To avoid biases I decided to start at 4044. To simplify interpretation of the intercept I subtract 40 from age. (You could plot the fits but this was not required.) . gen lmw=log(mw) . gen lmb=log(mb) . gen lb40 = agem40 (1 missing value generated) . reg lmw lb40 if age >= 40 Source  SS df MS Number of obs = 9+ F( 1, 7) = 6841.82 Model  11.4650463 1 11.4650463 Prob > F = 0.0000 Residual  .011730118 7 .001675731 Rsquared = 0.9990+ Adj Rsquared = 0.9988 Total  11.4767764 8 1.43459705 Root MSE = .04094 http://data.princeton.edu/eco572/ps2sol.html (2 of 14) [2/11/2008 2:16:28 PM] Eco 572: Research methods in Demography lmw  Coef. Std. Err. t P>t [95% Conf. Interval]+ lb40  .0874263 .001057 82.72 0.000 .084927 .0899256 _cons  6.053246 .0274181 220.78 0.000 6.118079 5.988412 . di exp(_b[_cons]) ", " exp(_b[lb40]) .00235022, 1.0913619 . reg lmb lb40 if age >= 40 Source  SS df MS Number of obs = 9+...
View
Full
Document
This homework help was uploaded on 02/12/2008 for the course ECON 572 taught by Professor Rodriguez during the Spring '06 term at Princeton.
 Spring '06
 Rodriguez

Click to edit the document details