Justin L. Tobias and Mingliang Li, "A Finite Sample Hierarchical Analysis of
Wage Variation Across Public High Schools: Evidence from the NLSY and High
School and Beyond", Journal of Applied Econometrics, Vol. 18. No. 3,
2003, pp. 315-336.

All data files, which are in DOS format, are zipped in the file tl-data.zip.
Unix users should use "unzip -a". There are two sets of files, one for the
NLDY data and one for the HSB data.

The NLSY portion of the data is broken into three blocks, corresponding to
the various levels of the hierarchy.

In first_stage.dat, the data (2,493 obs) are recorded as follows:

Column 1: Person Identifier variable (1-481)
Column 2: Log hourly wage
Column 3: Potential Labor Market Experience
Column 4: Potential Labor Market Experience^2

In second_stage.dat (481 obs), the time-invariant characteristics are
presented as follows

Column 1: School Identifier Variable (1-98)
Column 2: Education 
Column 3: Ability (Standardized) 
Column 4: Highest Grade Completed by Mother

In third_stage.dat (98 obs), the school-level characteristics are given.
These are recorded as follows:

Column 1: Constant
Column 2: Pupil-Teacher Ratio
Column 3: Percent of Teachers in School with 
		at Least a Master's Degree
Column 4: Books Per Student in School's Library
Column 5: Average of Mother's Education in School
Column 6: Average of Ability in School 

Note that the observations are ordered across the data sets. That is, the
first row of second_stage.dat corresponds to the time-invariant
characteristics for individual 1 in first_stage.dat. Similarly, the 481'st
row of second-stage.dat corresponds to the time-invariant characteristics of
person 481 in first_stage.dat.

Similarly, the j^th row of the third_stage data set corresponds to set of
observations in second_stage.dat with the school ID variable =j.

*****

The HSB data are stored in the file hsb.dat.

The data file is rectangular with 1096 lines. Each line represents an
observation. For each observation, we provide 20 variables. The names of the
variables are in order:

 (1) school id
 (2) wage
 (3) parent education 
 (4) number of siblings 
 (5) family income
 (6) test score 
 (7) education 
 (8) experience 
 (9) experience^2
 (10) suburban
 (11) rural
 (12) pupils per teacher
 (13) books per pupil
 (14) teacher education
 (15) teacher senior
 (16) district expenditure
 (17) average father education
 (18) average mother education
 (19) average family income 
 (20) average test score 

Overall, we have 1096 students from 226 schools.


Justin Tobias
Department of Economics
University of California-Irvine
3151 Social Science Plaza
Irvine, CA 92697-5100
http://orion.uci.edu/~jtobias/jtobias.html
E-mail: jtobias@uci.edu