Table A4

The effect of college education on smoking behavior at different ages. Panel data with fixed effects, marginal effects of conditional logit. Females

Birth cohort	Dependent variable: smoked at age 17 and smoked at ages 25 to 60
Birth cohort	1910–19	1920–29	1930–39	1940–49	1950–59	1960–69	1970–79
	At age 25
College	−0.0001 [0.0002]	0.018 [0.0269]	−0.020 [0.0269]	−0.001 [0.0154]	−0.054^*** [0.0160]	−0.066^*** [0.0230]	−0.013 [0.0411]
Observations	4,408	8,274	8,466	11,528	12,538	6,824	1,536
	At age 30
College Observations	−0.010 [0.0520]	0.104 [0.256]	−0.070^** [0.0325]	−0.024 [0.0225]	−0.086^*** [0.0195]	−0.096^*** [0.0273]
Observations	5,342	9,420	9,306	12,230	11,232	5,344
	At age 35
College	0.009 [0.0320]	−0.010 [0.02603]	−0.132^*** [0.0329]	−0.056^*** [0.0175]	−0.094^*** [0.0208]	−0.134^*** [0.0362]
Observations	5,630	9,690	9,442	10,958	9,032	2,508
	At age 40
College	0.003 [0.0280]	−0.006 [0.0269]	−0.134^*** [0.0292]	−0.096^*** [0.0210]	−0.085^*** [0.0242]
Observations	5,868	9,774	9,344	9,070	6,216
	At age 45
College	−0.029 [0.0387]	−0.010 [0.0238]	−0.169^*** [0.0279]	−0.117^*** [0.0233]	−0.118^*** [0.0386]
Observations	5,850	9,586	8,122	7,186	2,336
	At age 50
College	−0.041 [0.0341]	−0.001 [0.0287]	−0.124^*** [0.0260]	−0.127^*** [0.0264]
Observations	5,774	9,258	6,506	4,910
	At age 55
College	−0.034 [0.0377]	−0.002 [0.0310]	−0.135^*** [0.0305]	−0.169^*** [0.0417]
Observations	5,470	7,896	5,066	1,820
	At age 60
College	−0.056^* [0.0328]	−0.007 [0.0308	−0.112^*** [0.0375]
Observations	5,120	6,062	3,394

↵Analysis from smoking histories from the 1978, 1979, 1980, 1987, 1988, 1992, 1995, 1997, 1998, 1999, and 2000 National Health Interview Surveys for individuals age 25 and older at the time of the survey. Smoking histories are reconstructed using the age at smoking initiation and cessation. See also Note 5. Panel data regressions with individual fixed effects: the baseline smoking behavior is always taken at age College always takes the value 0 at age 17. At ages 25 and older, it takes the value 1 for college graduates and the value 0 otherwise. Additional controls include age, gender, year of birth, race, and survey year dummies. Robust standard errors in square brackets. ***, **, * denote statistical significance at the 1 percent, 5 percent, and 10 percent confidence levels, respectively.