See for example SinkeyJoseph F.Jr., “A Multivariate Statistical Analysis of the Characteristics of Problem Banks,”Journal of Finance, March, 1975; SinkeyJoseph F.Jr. and WalkerDavid A., “Problem Banks: Identification and Characteristics,”Journal of Bank Research, Winter, 1975; SantomeroA. and VinsoJ.D., “Establishing the Probability of Failure for Commercial Banks and the Banking System,”Journal of Banking and Finance, October, 1979.
2.
LevensteinCharles and EisenEllen, “Certified Environmental Audits: A Proposal,”Journal of Public Health Policy, August, 1987.
3.
WegmanDavid H.EisenEllen A., and FroinesJohn R., “Hazard Surveillance: A Proposal,”Journal of Public Health Policy, Autumn, 1988, p. 335.
4.
BodenLeslie I., “OSHA Inspection Policy: A Fresh Start,”Journal of Public Health Policy, Autumn, 1988, p. 321.
5.
LakshmananT. R. and RatickSam, “Integrated Models for Economic-Energy-Environmental Impact Analysis,”Economic- Environmental-Energy Interactions, Edited by LakshmananT. R. and NijkampP. (Boston, MA: Martinus Nijoff Publishing, 1980); “An Overview of the Strategic Environmental Assessment System,”Systems Models for Energy and Environmental Analysis, Edited by NijkampP. and LakshmananT. R. (London: Gower Press, 1983).
6.
Levenstein and Boden, Op. Cit.
7.
KriebelD., “Occupational Injuries: Factors Associated with Frequency and Severity,”International Archives of Occupational and Environmental Health, Spring-Summer, 1982.
8.
CorsiThomas M.FanaraPhilipJr. and JarrellJudith L., “Safety Performance of Pre-MCA Motor Carriers: 1977 versus 1984,” unpublished manuscript.
9.
For example, although firms report the number of new hires and energy consumed per employee to the Bureau of Labor Statistics, they do not usually include that data in their annual or quarterly financial statements. Also, a predictive model needs to explain more than 6 percent of the variance.
10.
See for example EisenbeisRobert A., “Pitfalls of Discriminant Analysis in Business, Finance, and Economics,”Journal of Finance, June, 1977, pp. 890–891; FosterGeorge, Financial Statement Analysis (Englewood Cliffs: Prentice-Hall, Inc., 1982), pp. 476–480; OhlsonJames A., “Financial Ratios and Probabilistic Prediction of Bankruptcy,”Journal of Accounting Research, Spring, 1980, p. 112.
11.
Motor Carriers are classified by the size of their gross operating revenues. Class I carriers have annual operating revenues of over $3 Million, Class II firms have operating revenues of $500,000 to $3 Million, and Class III have operating revenues less than $500,000.
12.
The names of 26 failed carriers were obtained from lists compiled by Dun and Bradstreet and the American Trucking associations. The remaining names were obtained from a list compiled by the Interstate Commerce Commission (ICC) or from the business and financial press. The list compiled by the ICC was found to contain errors in the date of default which in some cases was off by more than one year. Consequently, default date verification was necessary before a failed firm was included in the study. In nine cases verification of the true default date could not be obtained.
13.
Foster, Op. Cit., p. 478.
14.
ChenKung H. and ShimerdaThomas A., “An Empirical Analysis of Useful Financial Ratios,”Financial Management, Spring, 1981.
15.
BoisjolyRussell P. and CorsiThomas M., “Motor Carrier Failures: A Predictive Model,” unpublished manuscript.
16.
AltmanEdward I., “Financial Ratios, Discriminant Analysis and the Prediction of Corporate Bankruptcy,”Journal of Finance, September, 1968.
17.
See Ohlson, Op. Cit., p. 109–131 for an example of conditional logit analysis.
18.
The reader is referred to WhiteKenneth J., “Consumer Choice and the Use of Bank Credit Cards: A Model and Cross Section Results,”Journal of Consumer Research, June, 1975 for an example of the use of probit analysis.
19.
LachenbruchPeter A., “An Almost Unbiased Method of Obtaining Confidence Intervals for the Probability of Misclassification in Discriminant Analysis,”Biometrics, December, 1967; “On Expected Probabilities of Misclassification in Discriminant Analysis, Necessary Sample Size, and a Relation with the Multiple Correlation Coefficient,”Biometrics, December, 1968; LachenbruchPeter A. and MickeyM. R., “Estimation of Error in Discriminant Analysis,”Technometrics, February, 1968.
20.
The interested reader is referred to Boisjoly and Corsi, Op. Cit. for a more complete discussion of the results.
21.
The signs of the coefficients are interpreted as follows: A positive sign indicates that a lower value for that variable would increase the chance of default, whereas a negative sign informs us that a higher value would increase the likelihood of default.
22.
AltmanEdward and GrittaRichard, “Airline Bankruptcy Propensities: A Zeta Analysis,”Transportation Research Forum, October, 1984.
23.
I wish to thank LohmanSandraPractackarumWirachCastiglioneJoanne, and MacDonaldKathryn for assistance in collecting data for this study.