Panel Data Econometrics

Panel Data Sets

Professor W. Greene
Department of Economics
Office: MEC 7-90
Ph: +1-212-998-0876
E-mail: wgreene@stern.nyu.edu
Home Page: http://people.stern.nyu.edu/wgreene

Return to course home page.

Notes: The following list points to a series of data sets. We will use some of these in our class discussions. A
number of others are provided for students to analyze as part of their study of the topic. The Penn World Tables
are a major cross country data base online that provides a wealth of interesting data. This can be accessed
directly: The Penn World Tables

There are many other sources of data on the web. One that is particularly rich is the archives of the Journal of
Applied Econometrics: (Click here to visit)

Data below are provided in two formats: (1) The 'csv format' is a plain vanilla ascii text file containing the
variable names at the top of the file followed by the variables, arranged neatly in the file and separated by
commas. This is a portable file that is readable by any econometrics package. It will also import directly into
Excel just by double clicking its name. (2) If you are using LIMDEP or NLOGIT, the .lpj project file can be
imported directly into the program, as is. (The .csv file can as well.)

Grunfeld Investment Data, 10 Firms, 20 Years (1935-1954)

Variables in the file are
Firm = Firm ID, 1,...,10
Year = 1935,...,1954
I = Investment
F = Real Value of the Firm
C = Real Value of the Firm's Capital Stock
Data are from the Ph.D. dissertation of Y. Grunfeld (Univ. of Chicago, 1958), See, e.g., Zellner, A.,
"An Efficient Method of Estimating Seemingly Unrelated Regression Equations and Tests for
Aggregation Bias," Journal of the American Statistical Association, 57, 1962, pp. 348-368 for analyses
of these data.
o Excel Spreadsheet (csv)
o LIMDEP Project file (lpj)

Spanish Dairy Farm Production, N = 247, T = 6

Variables in the file are
FARM = Farm ID
YEAR = year, 93, 94, ..., 98
Inputs
COWS, X1 = log of, deviations from means (logs)
LAND, X2 = same
LABOR, X3 = same
FEED, X4 = same
Translog terms = squares and cross products: X11, X22, X33, X44, X12, X13, X14,X23, X24, X34
YEAR93,...,YEAR98 = year dummy variables
Output
 MILK = farm output
YIT = log of MILK production
o Excel Spreadsheet (csv)
o LIMDEP Project file (lpj)

Swiss Railways, N = 49, T(i) = 3 to 13

Variables in the file are
ID = Company ID from 1 to 50 (49 companies, 604 obs) ID = 21 zero count.
RAILROAD = Company ID reduced to remove gap at ID=21, values 1 to 49.
YEAR = Year (1985 to 1997).
NI = Number of year observations for firm, repeated in each observation.
STOPS = Number of stations on the network.
NETWORK = Length of railway network (m).
LABOREXP = Labor expenses in 1000 CHF.
STAFF = Number of employees.
ELECEXP = Electricity expenses in 1000 CHF.
KWH = Total consumed electricity (in 1000 kWh).
TOTCOST = Total cost (in 1000 CHF).
NARROW_T = Dummy for the networks with narrow track (1 m wide).
RACK = Dummy for the networks with RACK RAIL (cremaillere) in at least some part;
(used to maintain a slow movement of the train on high slopes).
TUNNEL = Dummy for networks that have tunnels with an average length of more than 300 meters.
VIRAGE = Dummy for the networks whose minimum radius of curvature is 100 meters or less.
CT = Total costs adjusted for inflation (1000 CHF).
Q1 = Total output in train kilometers.
Q2 = Total passenger output in passenger kilometers.
Q3 = Total goods output in ton kilometers.
PL = Labor price adjusted for inflation (in CHF per person per year).
PK = Capital price using the total number of seats as a proxy for capital stock (CHF per seat).
PE = Price of electricity (CHF per kWh).
LABOR = Quantity of labor.
ELEC = Quantity of energy.
CAPITAL = Quantity of Capital
Capital costs = TOTCOST- (LABOREXP + ELECEXP)
Inflation adjustment is done with respect to 1997 prices.
Logs of costs and prices (lnCT, lnPK, lnPL) are normalized by PE.
LNCT, LNPK, LNPL, LNQ1, LNQ2, LNQ3, LNSTOP, LNCAP, LNNET = logs of variables
o Excel Spreadsheet (csv)
o LIMDEP Project file (lpj)

Bank Cost Data, 500 Banks, 5 Years:

Variables in the file are
Cit = total cost of transformation of financial and physical resources into loans and
investments = the sum of the five cost items described below;
Y1it = installment loans to individuals for personal and household expenses;
Y2it = real estate loans;
Y3it = business loans;
Y4it = federal funds sold and securities purchased under agreements to resell;
Y5it = other assets;
W1it = price of labor, average wage per employee;
W2it = price of capital = expenses on premises and fixed assets divided by the dollar value of
of premises and fixed assets;
W3it = price of purchased funds = interest expense on money market deposits plus expense of
federal funds purchased and securities sold under agreements to repurchase plus interest
expense on demand notes issued by the U.S. Treasury divided by the dollar value of
purchased funds;
W4it = price of interest-bearing deposits in total transaction accounts = interest expense on
interest-bearing categories of total transaction accounts;
 W5it = price of interest-bearing deposits in total nontransaction accounts = interest expense on
total deposits minus interest expense on money market deposit accounts divided by the
dollar value of interest-bearing deposits in total nontransaction accounts;
T = trend variable, t = 1,2,3,4,5 for years 1996, 1997, 1998, 1999, 2000
The data in the file are for a translog cost function, linearly homogeneous in the input
prices. Specifically,
C = log(Cost/W5), W1,W2,W3,W4 = log(Wj/W5), Q1,...,Q5 = log(Ym), and the squared and cross
product
terms are W11, W12,..., Q11,Q12,..., W1Q1,...,W4Q5, T, T2, TW1,...,TW4, TQ1,...,TQ5.
o Excel Spreadsheet (csv)
o LIMDEP Project file (lpj)
Dahlberg and Johansson Municipal Expenditure Data, 265 Swedish Municipalities, 9 years
Variables in the file are
ID = Identification, 1,..., 265
YEAR = year, 1979,...,1987
EXPEND = Expenditures
REVENUE = Receipts, taxes and Fees
GRANTS = Government grants and shared tax revenues
See Greene (2003, pp. 551 and elsewhere) for analysis of these data. The article on which the analysis
is based is Dahlberg, M. and E. Johannson, E., "An Examination of the Dynamic Behavior of Local
Governments using GMM Bootstrapping Methods," Journal of Applied Econometrics, 15, 2000, pp.
401-416. (These data were downloaded from the JAE data archive.)
o Excel Spreadsheet (csv)
o LIMDEP Project file (lpj)
World Gasoline Demand Data, 18 OECD Countries, 19 years
Variables in the file are
COUNTRY = name of country (Does not appear in the LIMDEP project file)
YEAR = year, 1960-1978
LGASPCAR = log of consumption per car
LINCOMEP = log of per capita income
LRPMG = log of real price of gasoline
LCARPCAP = log of per capita number of cars
See Baltagi (2001, p. 24) for analysis of these data. The article on which the analysis is based is
Baltagi, B. and Griffin, J., "Gasolne Demand in the OECD: An Application of Pooling and Testing
Procedures," European Economic Review, 22, 1983, pp. 117-137. The data were downloaded from the
website for Baltagi's text.
o Excel Spreadsheet (csv)
o LIMDEP Project file (lpj)

Statewide Capital Productivity Data, lower 48 states, 17 years

Variables in the file are
STATE = state name
ST_ABB = state abbreviation
YR = year, 1970,...,1986
P_CAP = public capital
HWY = highway capital
WATER = water utility capital
UTIL = utility capital
PC = private capital
GSP = gross state product
EMP = employment
UNEMP = unemployment rate
See Baltagi (2001, p. 25) for analysis of these data. The article on which the analysis is based is
Munell, A., "Why has Productivity Declined? Productivity and Putlic Investment," New England
Economic Review, 1990, pp. 3-22. The data were downloaded from the website for Baltagi's text.
o Excel Spreadsheet (csv)
o LIMDEP Project file (lpj)
Cornwell and Rupert Returns to Schooling Data, 595 Individuals, 7 Years
Variables in the file are
EXP = work experience
 WKS = weeks worked
OCC = occupation, 1 if blue collar,
IND = 1 if manufacturing industry
SOUTH = 1 if resides in south
SMSA = 1 if resides in a city (SMSA)
MS = 1 if married
FEM = 1 if female
UNION = 1 if wage set by unioin contract
ED = years of education
BLK = 1 if individual is black
LWAGE = log of wage
These data were analyzed in Cornwell, C. and Rupert, P., "Efficient Estimation with Panel Data: An
Empirical Comparison of Instrumental Variable Estimators," Journal of Applied Econometrics, 3,
1988, pp. 149-155. See Baltagi, page 122 for further analysis. The data were downloaded from the
website for Baltagi's text.
o Excel Spreadsheet (csv)
o LIMDEP Project file (lpj)
German Manufacturing Innovation Data, 1,270 Firms, 5 years
Variables in the file are
YEAR = year, 1994-1998
FIRM = 1,...,1,270
IP = Product or innovation occurrec, 0/1 variable
EMPL = employment
IM = Imports
IMUM = import share in industry
FDIUM = FDI share in industry
PROD = productivity measure
LOGSALES = log of industry sales
RAWMTL = dummy for firm in raw materials industry
INVGOOD = dummy for firm in investment goods industry
CONSGOOD = dummy for firm in consumer goods industry
FOOD = dummy for firm in food industry
These data were analyzed in Bertschek, I. and M. Lechner, "Convenient Estimators for the Panel Probit
Model," Journal of Econometrics, 87, 2, 1998, pp. 329-372. See, also, Greene, Econometric Analysis,
5th ed., (2003) for various analyses, and Greene, W. "Convenient Estimators for the Panel Probit
Model: Further Results, Empirical Economics, 2004. These data are not publicly available. Extracts
from the data set will be provided in class.
German Health Care Usage Data, 7,293 Individuals, Varying Numbers of Periods
Data downloaded from Journal of Applied Econometrics Archive. This is an unbalanced panel with
7,293 individuals. They can be used for regression, count models, binary choice, ordered choice, and
bivariate binary choice. This is a large data set. There are altogether 27,326 observations. The
number of observations ranges from 1 to 7. (Frequencies are: 1=1525, 2=2158, 3=825, 4=926, 5=1051,
6=1000, 7=987). Note, the variable NUMOBS below tells how many observations there are for each
person. This variable is repeated in each row of the data for the person.
Variables in the file are
ID = person - identification number
FEMALE = female = 1; male = 0
YEAR = calendar year of the observation
AGE = age in years
AGESQ = age squared.
HSAT = health satisfaction, coded 0 (low) - 10 (high) Note, this variable has 40 coding errors.
Variable NEWHSAT below fixes them.
HANDDUM = handicapped = 1; otherwise = 0
HANDPER = degree of handicap in percent (0 - 100)
HHNINC = household nominal monthly net income in German marks / 10000
HHKIDS = children under age 16 in the household = 1; otherwise = 0
EDUC = years of schooling
MARRIED = married = 1; otherwise = 0
HAUPTS = highest schooling degree is Hauptschul degree = 1; otherwise = 0
REALS = highest schooling degree is Realschul degree = 1; otherwise = 0
 FACHHS = highest schooling degree is Polytechnical degree = 1; otherwise = 0
ABITUR = highest schooling degree is Abitur = 1; otherwise = 0
UNIV = highest schooling degree is university degree = 1; otherwise = 0
WORKING = employed = 1; otherwise = 0
BLUEC = blue collar employee = 1; otherwise = 0
WHITEC = white collar employee = 1; otherwise = 0
SELF = self employed = 1; otherwise = 0
BEAMT = civil servant = 1; otherwise = 0
DOCVIS = number of doctor visits in last three months
HOSPVIS = number of hospital visits in last calendar year
DOCTOR = dummy variable = 1 if DOCVIS > 0, 0 otherwise.
HOSPITAL = dummy variable = 1 if HOSPVIS > 0, 0 otherwise.
PUBLIC = insured in public health insurance = 1; otherwise = 0
ADDON = insured by add-on insurance = 1; otherswise = 0
NUMOBS = number of observations for this person. Repeated in each row of data.
NEWHSAT = recoded value of HSAT with coding errors corrected.
o Excel Spreadsheet (csv)
o LIMDEP Project file (lpj)
World Health Organization Panel Data on Health Care Attainment: 191 Countries, 5 Years
(Some countries fewer)
These data have been used by many researchers to study the Health Care Survey assembled by WHO
as part of the Year 2000 World Health Report. On the course bibliography, see, for example, Greene
(2004a). Note, variables marked * were updated with more recent sources in Greene (2004a). Missing
values for some of the variables in this data set are filled by using fitted values from a linear
regression. To set the proper sample for panel data analysis, use observations for which SMALL =
0. To obtain the balanced panel, then use only observations with GROUPTI = 5.
The variables in the file are
COMP = composite measure of health care attainment; LCOMP = logCOMP
DALE = Disability adjusted life expectancy (other measure); LDALE = logDALE
YEAR = 1993,...,1997; TIME = 1,2,3,4,5; T93, T94, T95, T96, T97 = year dummy variables
HEXP = per capita health expenditure; LHEXP = logHEXP; LHEXP2 = log-squaredHEXP
HC3 = educational attainment; LHC = logHC3; LHC2 = log-squaredHC3; LHEXPHC = logHEXP *
logHC3
SMALL = indicator for states, provinces, etc. SMALL > 0 implies internal political unit, = 0 implies
country observation
COUNTRY = number assigned to country
STRATUM = another country indicator
GROUPTI = number of observations when SMALL = 0. Usually 5, some = 1, one country = 4.
OECD = dummy variable for OECD country (30 countries)
GINI = gini coefficient for income inequality
GEFF = world bank measure of government effectiveness*
VOICE = world bank measure of democratization of the political process*
TROPICS = dummy variable for tropical location
POPDEN = population density*
PUBTHE = proportion of health expenditure paid by bublic authorities
GDPC = normalized per capita GDP; LGDPC = logGDPC; LGDPC2 = log-squaredGDPC
o Excel Spreadsheet (csv)
o LIMDEP Project file (lpj)

Stated Discrete Choice 400 individuals, 8 choice situations, 4 alternatives:

Data are a simulated discrete choice experiment. 400 individuals choose a brand among 3 alternatives,
or none of the above, the 4th choice. 8 different situations are given to each individual.
Variables in the file are
id = person identification
BRAND = 1,2,3,4
CHOICE = alternative chosen (0=no, 1=yes);
FASH = dummy coded fashion (0=no, 1=yes);
QUAL = dummy coded quality (0=low, 1=high);
PRICE = price, level coded, .04, .08, .12, .16, .20;
PRICESQ = price squared;
 ASC4 = dummy variable for 4th choice (no brand);;
MALE = dummy variable for male (0=no, 1=yes);
AGE25 = age less than 25 (0=no, 1=yes)
AGE39 = age 25 to 39 (0=no, 1=yes),
AGE40 = age greater than 40 (0=no, 1=yes)
o Excel Spreadsheet (csv)
o LIMDEP Project file (lpj)

Stated Discrete Choice Experiment - California Public Utilities Survey Data:

These data were provided generously by Kenneth Train for use in teaching about stated choice
modeling. We gratefully acknowledge his assistance. This is a real stated choice experiment in which
361 customers were surveyed about their preferences for an electricity supplier. Each customer
responded to from 8 to 12 choice situations in which four configurations were offered. There are
altogether 17,232 rows in the data set comprising 4,308 choice tasks.
The variables in the file are
ID = respondent ID number
CHOICE = binary indicator of the preferred alternative.
NTASK = number of tasks ranging from 8 to 12, repeated on each of the 4*NTASK records
PRICE = fixed price in cents per kilowatt hour
CNTLNGTH = contract length
LOCAL = dummy variable for local utility
KNOWN = dummy variable for well known company
TOD = dummy variable for time of day rates, 11 cents in day, 5 cents at night
SEAS = dummy variable for seasonal rates (10 cents summer, 8 cents winter, 6 cents spring/fall)
NUMREC = 4*NTASK
o Excel Spreadsheet (csv)
o LIMDEP Project file (lpj)

Koop and Tobias (2004) Labor Market Experience Data.

(See Koop, G. and J. Tobias, "Learning About Heterogeneity in Returns to Schooling," Journal of
Applied Econometrics, 19, 2004, pp. 827-849.
The data file is in two parts. The first file contains the panel of 17,919 observations on the Person ID
and 4 time-varying variables. The second file contains time invariant variables for the individual or the
2,178 households. The data were downloaded from the Journal of Applied Econometrics archive
website. The two data sets are merged in the .csv and .lpj files noted.
Variables in the file are
Time Varying
PERSONID = Person id (ranging from 1 to 2,178),
EDUC = Education,
LOGWAGE = Log of hourly wage,
POTEXPER = Potential experience,
TIMETRND = Time trend.
Time Invariant
ABILITY = Ability,
MOTHERED = Mother's education,
FATHERED = Father's education,
BRKNHOME = Dummy variable for residence in a broken home,
SIBLINGS = Number of siblings.
o Excel Spreadsheet (csv)
o LIMDEP Project file (lpj)

Data on Industry Spillovers in Production

Used in Bloom, N., M. Schankerman and J. Van Reenen, "Identifying Technology Spillovers and
Product Market Rivalry," Econometrica, vol. 81(4), pages 1347-1393, 07, 2013. Also Burda, M.
and M. Harding, "Panel Probit with Flexible Correlated Effects: Quantifying Technology Spillovers
in the Presence of Latent Heterogeneity," Journal of Applied Econometrics, 28, 6, 2013, pp. 956-981.
The data were downloaded from the Journal of Applied Econometricswebsite. Unbalanced panel, 729
firms, 1 to 21 years of data per firm. 12928 rows of data in total.
 The Variables in the file are:
Firm = 1 to 729
Year = 1981 - 2001
Pat_any = 1 if firm had any patents in that year
LGSPILLT = lagged log of stock of tec weighted R&D (Jaffe distance)
LGSPILLS = lagged log of stock of sic weighted R&D (Jaffe distance)
LGMALSPI = lagged log of stock of tec weighted R&D (Mahalanobis distance)
LGMALSPT = lagged log of stock of sic weighted R&D (Mahalanobis distance)
LGRD1 = lagged log stock of R&D expenditures (coded -1 for missing)
LSALES1 = lagged log sales
LGRD1_DU = dummy variable indicates missing value of LGRD1
TI = Number of observations for firm i
T = year - 1981. 1 to 21
LGSTBAR = Firm mean of LGSPILLT
LGSPBAR = Firm mean of LGSPILLS
LGMSIBAR = Firm mean of LGMALSPI
LGMSTBAR = Firm mean of LGMALSPT
LSALEBAR = Firm mean of LSALES1
LGRD1BAR = Firm mean of LGRD1
o Excel Spreadsheet (csv)
o LIMDEP Project file (lpj)

Data on Crime in North Carolina

Used in Cornwell, C. and W. Trumbull, "Estimating the Economic Model of Crime with Panel
Data," Review of Economics and Statistics, 76, 1994, pp. 360-366 and Baltagi, B., "Estimating an
Economic Model of Crime Using Panel Data from North Carolina," Journal of Applied Econometrics,
21, 2006, pp. 543-547
Balanced panel, 90 counties and 7 years, 630 rows of data in total..
The last eight (constructed) variables are missing for 1981. Missing variables are denoted by a "." in
the text file.
The 59 Variables in the file are:
1. county = county identifier
2. year = 81 to 87
3. crmrte = crimes committed per person
4. prbarr = 'probability' of arrest
5. prbconv = 'probability' of conviction
6. prbpris = 'probability' of prison sentenc
7. avgsen = avg. sentence, days
8. polpc = police per capita
9. density = people per sq. mile
10. taxpc = tax revenue per capita
11. west = 1 if in western N.C.
12. central = 1 if in central N.C.
13. urban = 1 if in SMSA
14. pctmin80 = perc. minority, 1980
15. wcon = weekly wage, construction
16. wtuc = wkly wge, trns, util, commun
17. wtrd = wkly wge, whlesle, retail trade
18. wfir = wkly wge, fin, ins, real est
19. wser = wkly wge, service industry
20. wmfg = wkly wge, manufacturing
21. wfed = wkly wge, fed employees
22. wsta = wkly wge, state employees
23. wloc = wkly wge, local gov emps
24. mix = offense mix: face-to-face/other
25. pctymle = percent young male
26. d82 = 1 if year = 82
27. d83 = 1 if year = 83
28. d84 = 1 if year = 84
29. d85 = 1 if year = 85
30. d86 = 1 if year = 86
 31. d87 = 1 if year = 87
32. lcrmrte = log(crmrte)
33. lprbarr = log(prbarr)
34. lprbconv = log(prbconv)
35. lprbpris = log(prbpris)
36. lavgsen = log(avgsen)
37. lpolpc = log(polpc)
38. ldensity = log(density)
39. ltaxpc = log(taxpc)
40. lwcon = log(wcon)
41. lwtuc = log(wtuc)
42. lwtrd = log(wtrd)
43. lwfir = log(wfir)
44. lwser = log(wser)
45. lwmfg = log(wmfg)
46. lwfed = log(wfed)
47. lwsta = log(wsta)
48. lwloc = log(wloc)
49. lmix = log(mix)
50. lpctymle = log(pctymle)
51. lpctmin = log(pctmin)
52. clcrmrte = lcrmrte - lcrmrte[t-1]
53. clprbarr = lprbarr - lprbarr[t-1]
54. clprbcon = lprbconv - lprbconv[t-1]
55. clprbpri = lprbpri - lprbpri[t-1]
56. clavgsen = lavgsen - lavgsen[t-1]
57. clpolpc = lpolpc - lpolpc[t-1]
58. cltaxpc = ltaxpc - ltaxpc[t-1]
59. clmix = lmix - lmix[t-1]
o Excel Spreadsheet (csv)
o LIMDEP Project file (lpj)

Arellano-Bond: Dynamic Panel Data Modeling. UK Industry Data

Data on 140 firms observed in 7, 8 or 9 years. Unbalanced panel. 1031 observations in total.
The variables in the file are
IND = industry code
YEAR = year, 1977 to 1984
EMP = firm employment
WAGE = wage
CAP = capital
INDOUTPT = industry output
NI = log EMP
W = log WAGE
K = log CAP
YS = log IDOUTPT
REC = line number 1-1031
YEARM1 = lag of YEAR.
ID = firm id number
NL1 = lag 1 of N
NL2 = lag 2 of N
WL1 = lag 1 of W
WL2 = lag 2 of W
KL1 = lag 1 of K
KL2 = lag 2 of K
YSL1 = lag 1 of YS
YSL2 = lag 2 of YS
YR1976 to YR1984 are year dummy variables
o Excel Spreadsheet (csv)
o LIMDEP Project file (lpj)
Return to course home page.