solution to problem set4

# solution to problem set4 - Eco 572 Research methods in...

This preview shows pages 1–3. Sign up to view the full content.

Eco 572: Research methods in Demography Solutions to Problem Set 4 [1] Marriage in France (a) We compute and plot the period total marriage rate (TMR) and the period mean age at first marriage. (The plot of TMR is shown further below, together with the tempo-adjusted adjusted TMR.) . global DATASETS http://data.princeton.edu/eco572/datasets . infile age15-age50 using \$DATASETS/FranceAfmPeriodAge.dat, clear (33 observations read) . gen year = 1967 + _n . egen tmr = rowtotal(age15-age50) . line tmr year, title(Period Total Marriage Rate) . gen afm = 0 . forvalues age=15/50 { 2. quietly replace afm = afm + `age' * age`age' 3. } . quietly replace afm = afm/tmr . line afm year, title(Period Mean Age at First Marriage) /// > subtitle(France 1968-200) . graph export ps4Fig1.png, replace (file ps4Fig1.png written in PNG format) We see a steady decline in period total marriage rates, reaching a low just below 50% in 1993-95, followed by a sligh increase to reach about 60% at the turn of the century. In the meantime period mean age at marriage increased steadily in the last quarter century, climbing from historic levels of 22-23 years to 28 years of age in 2000. (b) To compute the Bongaarts-Feeney adjustment we estimate r , the rate of change in period mean age at first marriage, using half the change from the year before to the year after each calendar year, and then simply divide the TMR by 1-r . . gen r = (afm[_n+1]-afm[_n-1])/2 http://data.princeton.edu/eco572/ps4sol.html (1 of 12) [2/11/2008 2:18:52 PM]

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Eco 572: Research methods in Demography (2 missing values generated) . gen bftmr = tmr/(1-r) (2 missing values generated) . line tmr bftmr year, lpat(solid dash) /// > title(Observed and Tempo-Adjusted Total Marriage Rate) /// > subtitle(France 1969-1999) /// > legend( ring(0) pos(1) cols(1) order(1 "Observed" 2 "Adjusted")) . graph export ps4Fig2.png, replace (file ps4Fig2.png written in PNG format) Not surprinsingly, the tempo-adjusted total marriage rate exceeds the observed rate for the last quarter of the century, when period mean age at marriage was increasing, with some large fluctuations in recent years. This is a counterfactual estimate of how many women would have married each year if they had not postponed marriage that year, under the assumption that women of all ages delay marriage by the same amount of time in any given year. The implication is that as much as 70% of women will eventually marry. (c) To do this part we turn to a dataset showing marriage frequencies by age and cohort (the diagonals of the current data) . infile coh1968-coh2000 using \$DATASETS/FranceAfmAgeCohort.dat, clear (36 observations read) . drop in 34/36 // always missing (3 observations deleted) . gen age = 14 + _n . forvalues year=1968/1987 { 2. gen cum`year' = sum(coh`year')/10000 3. qui replace cum`year' = . if missing(coh`year') 4. } . gen agep = age+0.5 // at end of year of age . twoway (line cum1968-cum1986 agep ) /// > (line cum1987 agep, lwidth(thick) ) , xtitle(age) ///
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}

### Page1 / 12

solution to problem set4 - Eco 572 Research methods in...

This preview shows document pages 1 - 3. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online