Identification Problems in the Social Sciences Book

	Preface
	Introduction	1
	Identification	3
	Tolerating Ambiguity	7
1	Extrapolation	10
1.1	Predicting Criminality	10
1.2	Probabilistic Prediction	11
1.3	Inferring Conditional Distributions from Random-Sample Data	13
1.4	Prior Distributional Information	16
1.5	Predicting High School Graduation	18
2	The Selection Problem	21
2.1	The Nature of the Problem	21
2.2	Identification from Censored Samples Alone	23
2.3	Bounding the Probability of Exiting Homelessness	27
2.4	Prior Distributional Information	31
2.5	Identification of Treatment Effects	37
2.6	Information Linking Outcomes across Treatments	43
2.7	Predicting High School Graduation If All Families Were Intact	47
3	The Mixing Problem in Program Evaluation	51
3.1	The Experimental Evaluation of Social Programs	51
3.2	Variation in Treatment	54
3.3	The Perry Preschool Project	58
3.4	Identification of Mixtures Using Only Knowledge of the Marginals	62
3.5	Restrictions on the Outcome Distribution	65
3.6	Restrictions on the Treatment Policy	67
3.7	Identifying Combinations of Assumptions	72
4	Response-Based Sampling	73
4.1	The Odds Ratio and Public Health	74
4.2	Bounds on Relative and Attributable Risk	78
4.3	Information on Marginal Distributions	81
4.4	Sampling from One Response Stratum	82
4.5	General Binary Stratifications	85
5	Predicting Individual Behavior	88
5.1	Revealed Preference Analysis	88
5.2	How Do Youth Infer the Returns to Schooling?	95
5.3	Analysis of Intentions Data	98
6	Simultaneity	110
6.1	"The" Identification Problem in Econometrics	110
6.2	The Linear Market Model	112
6.3	Equilibrium in Games	116
6.4	Simultaneity with Downward-Sloping Demand	119
7	The Reflection Problem	127
7.1	Endogenous, Contextual, and Correlated Effects	127
7.2	A Linear Model	129
7.3	A Pure Endogenous Effects Model	133
7.4	Inferring the Composition of Reference Groups	134
7.5	Dynamic Analysis	135
	Notes	139
	References	155
	Name Index	165
	Subject Index	169