Panel Database and its Statistical Analysis on the Network
|Allocation Type||Single-year Grants|
|Research Institution||OSAKA UNIVERSITY|
BAN Kanemi Osaka University, Faculty of Economics, Professor, 経済学研究科, 教授 (30027578)
YOSHIDA Atsushi Osaka Prefecture University, College of Economics, Assistant Professor, 経済学部, 助教授 (60240272)
SUZUKI Kazuyuki Meiji University, Faculty of Commercial Sciences, Professor, 商学部, 教授 (40226501)
MAEKAWA Koichi Hiroshima University, Faculty of Economics, Professor, 経済学部, 教授 (20033748)
HISAMATSU Hiroyuki Kagawa University, Faculty of Economics, Assistant Professor, 経済学部, 助教授 (90228726)
小川 一夫 大阪大学, 社会経済研究所, 教授 (90160746)
|Project Period (FY)
1996 – 1998
Completed(Fiscal Year 1999)
|Budget Amount *help
¥17,600,000 (Direct Cost : ¥17,600,000)
Fiscal Year 1998 : ¥5,300,000 (Direct Cost : ¥5,300,000)
Fiscal Year 1997 : ¥5,800,000 (Direct Cost : ¥5,800,000)
Fiscal Year 1996 : ¥6,500,000 (Direct Cost : ¥6,500,000)
|Keywords||Pseudo / Panel / Cohort / Repeated Cross Section / Liquidity / 最適階層選択 / 単位根 / 共和分 / 疑似パネル / 企業データ / 全国消費実態調査 / パネルデータ / 疑似パネルデータ / 財務データ / 企業投資 / パネル分析 / 企業財務 / 企業行動 / コ-ホ-ト / ミクロデータ / 計量経済学 / 日経金融行動調査|
Least squares on the pooled cross-section and time series data yields inconsistent estimator if unobserved individual effects and disturbances are correlated. It is well known that an instrumental variable estimator yields a consistent estimator for these repeated cross section. The most important class of instruments is based on functions of time, because the fixed effect is time-invariant and hence must be uncorrelated with functions of time. Micro data usually contain information on exogenous time invariant characteristic such as the year of birth. The special case of this instrumental variable estimator is to compute group means of age-cohort cells, and use them in regression with cohort dummy. This method is called "Pseudo Panel Data".
We constructed pseudo panel data from the Family Income and Expenditure Survey conducted in 1984, 1989 and 1994. We consider the age of household heads and the residential location as the time-invariant variables. We aggregate the sample of household
s by ages and locations, in which age of head was between 20 and 70 years old in 1984, and the 47 prefectures are treated as locations. Consequently, we have the following cohorts.
Then, we estimated the model of consumption and liquidity constraints. In this model, implications for consumption in the presence borrowing constraints are derived and tested using the pseudo panel data. The estimated results support, the hypothesis that an inability to borrow against future labor income affects the consumption expenditures.
We developed a database of corporate firms, which includes financial data of more than 2000 companies in Japan over 40 years. The Japan Development Bank supplies original data. The database is constructed on the WWW server at the Faculty of Economics, Osaka University. The URL address is http://panel.econ.osaka-u.ac.jp/micro/. Using these data, we examine the relationship between uncertainty and corporate fixed investment for Japanese manufacturing industries by estimating the investment function. Less
Research Products (11results)