Development of estimation method for an index of complementarity/substitutability without price variables
| Project/Area Number |
18530171
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Applied economics
|
|---|
| Research Institution | Yokohama National University |
|---|
Principal Investigator |
TORII Akio Yokohama National University, Faculty of Business Administration, Professor (40164066)
|
|---|
| Project Period (FY) |
2006 – 2007
|
| Project Status |
Completed (Fiscal Year 2007)
|
| Budget Amount *help |
¥2,880,000 (Direct Cost: ¥2,700,000、Indirect Cost: ¥180,000)
Fiscal Year 2007: ¥780,000 (Direct Cost: ¥600,000、Indirect Cost: ¥180,000)
Fiscal Year 2006: ¥2,100,000 (Direct Cost: ¥2,100,000)
|
|---|
| Keywords | complementarity / substitutability / discrete choice model / ロジット分析 / 交差弾力性 |
|---|
| Research Abstract |
An index of complementarity/substitutability is introduced which is calculable even when explicit price variables are not observable. The index Prob(Y=1 | X=1)-Prob(Y=1 | X=0) is proved to have the appropriate properties for an index using a probabilistic utility model. Also methods to estimate the index empirically are shown and their statistical properties of estimators are examined by Monte-Carlo simulation methods. The summary of results are: the index shows expected appropriate properties as an index of complemantarity/substitutability, that is, the index shows monotonistic association with the degree of complementarity/substitutability; under certain construction, the index shows the same value with its cross-elasticity normalized by their own price elasticity; simultaneous logit regression can be employed for estimation of the index proposed in Schmidt and Strauss(1975), on the other hand simple multinominal logit regression provides almost the same estimates of the index; although estimates have a significant bias from the true values expected by theory, the bias does not deteriorate tests of complementarity/substitutability; robust estimators are provided even without price variables.
As a case study, the index is applied to analyze viewers' choice behavior with respect to terrestrial TV broadcasting services in Japan. The public broadcasting channel seems to provide programs that are substitutes to those provided by private channels supported by advertising, while those private channels seem to provide complements each other.
|
Report
(3 results)
Research Products
(3 results)
