Bachelor of Science N. Carolina State College, 1961. Master of Arts, Doctor of Philosophy Harvard University, 1963, 1966.
Teaching Fellow, Harvard University, 1961, 1963-1964. State Consultant, Harvard Computer Center, 1965. Lector, Senior Lector, Technion, Haifa, Israel,
Senior Lector, Haifa University, 1966-1967. Research Fellow, National Bureau of Economie Research, New York, New York, United States of America, New York City, 1967-1974. Visiting Professor, City College New York, 1968.
Senior Lector, Association Professor, Hebrew University, 1969-1973. Consultant, Bank Israel, since 1969. Research Fellow, Falk Institute, Institution Economics Research in Israel, 1970-1971.
Visiting Research Fellow, University College London 1971,1972-1973. Director, Data Bank Development Division, Israeli Central Bureau Statistics, 1973-1978. Visiting Professor, Centre for Operations Research and Econometrics, Belguim, Catholic University Louvain, Belgium, 1974.
Director, Social Science Data Archive, Hebrew University, 1973-1978. Visiting Chief Scientist, Erasmus University, Rotterdam, 1977. Visiting Professor, University of California, Los Angeles, Calif., United States of America, 1978-1980.
Director, Empirical Research Unit, Hebrew University, 1980-1982. Visiting Scholar, University British Columbia, Canada, 1980. Visiting Scientist, National Research Institute, Institution Mathematics Sciences, CSIR, Pretoria, S. Africa, 1982.
Professor Economics and Statistics, Hebrew University Jerusalem, Jerusalem, Israel, since 1980.
Early work was concerned with the problem of estimating regression models with insufficient data. One type of insufficiency is when only grouped observations are available, but the cross-classifications, necessary for efficient estimation of multiple regression coefficients, are not and hence the available grouped data must be combined in a particular fashion to yield alternative estimates. Another type of insufficiency is when some data, either random or a whole block of, say, time-series data, are missing.
A third type is the so-called ‘error of measurement in regression’ problem, where in my work the errors are not assumed to be uncorrelated with the true quantities, as it is normally assumed. A related problem is that of multicollinearity. Later work was concerned with macroeconometric models, their estimation (Bayesian method) and their use as a forecasting tool and policy evaluation.
More recently, I have worked on the development of the Linear Hierarchical Model — a generalisation of the statistical linear model (or regression model) — and its application in various economic areas such as household demand for electricity, firm’s production and investment function, returns to schooling, etc. Another area of interest is the use of first differencing in econometric analyses and forecasts.