Variance Specification in Event Count Models: From Restrictive Assumptions to a Generalized Estimator

King, Gary

doi:10.2307/2111071

Cited by 253 publications

(155 citation statements)

References 46 publications

Supporting

Mentioning

145

Contrasting

Unclassified

Order By: Relevance

“…treated by King (1989). Figure 1 shows the variance-mean ratio as a function of the mean for different values of k. It demonstrates that substantial flexibility can be gained by varying k. Imposing a specific value for k, as is done in the literature, may lead to the wrong model in many data situations with the potential of false conclusions.…”

Section: Poisson Modelmentioning

confidence: 99%

“…Then we introduce the generalized event count model which can account for both over-and underdispersion and allows for any variance-mean relationship of the form (2). 1 We generalize the contribution by King (1989) treating k as an continuous parameter to allow for more flexibility.…”

Section: Poisson Modelmentioning

confidence: 99%

“…We develop the GEC k along the lines of King (1989). The derivation is based on a result in theoretical statistics that binomial, Poisson and negative binomial distributions are members of the Katz family defined by a recursive formula for the probabilities p (y):…”

Section: Generalized Event Count Modelmentioning

confidence: 99%

“…A flexible model allowing for any linear variance-mean relationship, the generalized event count model (GEC), was introduced by King (1989). It contains the Poisson model as a special case and allows for over-as well as for underdispersion.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Count data models for demographic data∗

Winkelmann

Zimmermann

1994

Mathematical Population Studies

View full text Add to dashboard Cite

Key demographic variables, such as the number of children and the number of marriages or divorces, can only take integer values. This papers deals with the estimation of single equation models in which the counts are regressed on a set of observed individual characteristics such as age, gender, or nationality. Most empirical work in population economics has neglected the fact that the dependent variable is a nonnegative integer. In the few cases where this feature was recognized, the authors advocated the use of the Poisson regression model. The Poisson model imposes, however, the equality of conditional mean and variance, a restriction which is often rejected by the data. We propose a generalized event count model to simultaneously allow for a wide class of count data models and account for over-and underdispersion. This model is successfully applied to German data on fertility, divorces and mobility. Abstract Key demographic variables, such as the number of children and the number of marriages or divorces, can only take integer values. This papers deals with the estimation of single equation models in which the counts are regressed on a set of observed individual characteristics such as age, gender, or nationality. Most empirical work in population economics has neglected the fact that the dependent variable is a nonnegative integer. In the few cases where this feature was recognized, the authors advocated the use of the Poisson regression model. The Poisson model imposes, however, the equality of conditional mean and variance, a restriction which is often rejected by the data. We propose a generalized event count model to simultaneously allow for a wide class of count data models and account for over-and underdispersion. This model is successfully applied to German data on fertility, divorces and mobility.

show abstract

Section: Poisson Modelmentioning

confidence: 99%

Section: Poisson Modelmentioning

confidence: 99%

Section: Generalized Event Count Modelmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Count data models for demographic data∗

Winkelmann

Zimmermann

1994

Mathematical Population Studies

View full text Add to dashboard Cite

show abstract

“…Parametric models for underdispersed data include the Katz system, see King (1989), and the generalized Poisson, see Consul and Famoye (1992).…”

Section: Maximum Likelihood Estimationmentioning

confidence: 99%

12 Count data models for financial data

Cameron¹,

Trivedi²

1996

Handbook of Statistics

View full text Add to dashboard Cite

To appear in G.S. Maddala and C.R. Rao ed., Handbook of Statistics: Statistical Methods in Finance, North-Holland.In some financial studies the dependent variable is a count, taking nonnegative integer values. Examples include the number of takeover bids received by a target firm, the number of unpaid credit installments (useful in credit scoring), the number of accidents or accident claims (useful in determining insurance premia) and the number of mortgage loans prepaid (useful in pricing mortgage-backed securities). Models for count data, such as Poisson and negative binomial are presented, with emphasis placed on the underlying count process and links to dual data on durations. A self-contained discussion of regression techniques for the standard models is given, in the context of financial applications.

show abstract

Investments to create bargaining power: the case of franchising

2000

View full text Add to dashboard Cite

Hybrid organizational forms such as franchise systems join two or more independent parties under a contract. The ability of each party to achieve its goals depend upon the relative bargaining power in the relationship established by the contract. Using transaction cost economics and Porter's (1980) characterization of sources of bargaining power, this paper argues that the franchisor can make investments in activities such as tapered integration and buyer selection to increase its bargaining power and decrease conflict and litigation in a franchise system. Specifically, tapered integration (owning some units while franchising others), selecting inexperienced franchisees, and employing a long training program are predicted to increase the franchisor's bargaining power and the franchisee's compliance with franchisor standards. An empirical analysis of litigation in restaurant franchise systems supports the theoretical hypotheses. Copyright © 2000 John Wiley & Sons, Ltd.

show abstract

Variance Specification in Event Count Models: From Restrictive Assumptions to a Generalized Estimator

Cited by 253 publications

References 46 publications

Count data models for demographic data∗

Count data models for demographic data∗

12 Count data models for financial data

Investments to create bargaining power: the case of franchising

Contact Info

Product

Resources

About