Brief Summary
This paper proposes a strategy to study a firm's proper behavior by merging game theoretic ideas and recent economic improvements. It analyses various kinds of collusive action of businesses on two tactical variables-price and advertising- in a differentiated market dominated by a duopoly. The econometric methodology adopted is fully structural. The strategy involves specification of demand and cost functions and hypotheses about the tactical interactions among players. The parameters of the demand functions and the price functions are predicted under different proper hypotheses.
Prior work have modeled proper interactions (output and costs decisions) in a non-cooperative static method using static conjectural variant models. However, recent improvements in game theoretic work as well as experimental research have shown proof cooperation among players in repeated game contexts even under the assumption on non-cooperative tendencies. Such kind of assistance is known as tacit collusion. Given the intricacy of empirical analysis of collusive tendencies, the authors package with the dark-colored box of proper dynamic connections by selecting a sufficiently rich range of formulations expressing various levels of collusion. With two observable instruments of competition-price and advertising, the writers offer various simple formulations of collusive action and select among them. The merits of every formulation is dependant on the number of possible levels of collusion.
Since the econometric models are nonnested, lab tests for nonnested hypotheses is performed to choose the most satisfactory model. The models are approximated by full information maximum probability methods. This analysis also extends the traditional conjectural way for the empirical analysis of market ability. The proposed technique is then put on the soft drink industry which is dominated by The Coca-Cola Company and Pepsico duopoly. Three types of non-collusive habit and three models of collusive tendencies is estimated. Based on the results, the hypotheses of non-collusive action is declined. The results suggest some tacit collusive action in advertising between your Coca-Cola Company and Pepsico for period included in the sample data. However, collusion on prices does not appear to be well backed by the info.
The methodology permits various moving over regimes specifications because there is a potential change of habit in the middle of the test. The paper projected models with two regimes- before and after 1976. Results show that Coca-Cola is a Stackelberg head in cost and advertising until 1976, and after 1976 there is certainly collusion in advertising and prices. Results also show an increase in market power for both the firms after 1976 based on the Lerner indices computations.
Key Strengths
Simplifies demand and cost specifications by imposing constraints on parameters according to economic theory
On the broader level, there is a formidable process of together estimating demand and cost functions, and determine the most satisfactory collusive hypotheses. Because of this gigantic demand on data, there is a dependence on simple features that limit the amount of estimated variables with the chance of having results that may be strongly afflicted by the implied misspecifications. So a straightforward demand specs is chosen for analytical and empirical tractability. That is done by imposing constraints on parameters predicated on economical theory.
So predicated on monetary theory, constraints are enforced on the guidelines indicators. The given demand specification implies diminishing comes back in advertising and also permits an array of cross-advertising effects. The effect of advertising in addition has been restricted only for the given quarter. This constrained form of advertising effects decreases the complexness of the reduced form. Constraints are also imposed on the parameters of cost functions based on monetary theory. A choice of constant marginal cost is perfect for analytical and empirical tractability.
Takes into consideration model misspecification regarding statistical inference
Model misspecification can occur when the models are simplified as in cases like this so when the models are not correctly specified. Therefore the models that are statistically dominated by another competing model are misspecified. However, statistical inference can be made on the parameters of these models provided White solid t statistics are being used.
Adopts full information maximum possibility way for getting reliable estimates
Limited or full information estimation by 2SLS and 3SLS methods have certain downsides. One such drawback is that it does not provide estimations of some structural variables including the collusion parameter and the coefficients in the cost functions. Another disadvantage is that they produce unreliable estimations. Also, the typical Wald statistics can't be used straight because each set of nonlinear restrictions come in the explicit or parametric form. Instead one must use the generalized Wald reports that requires a nonlinear minimization for each set of restrictions. The last drawback is the fact that selection among the models can only just be done indirectly through these generalized Wald exams. This might lead to undesired final results. To avoid the above mentioned issues, the writers adopt a direct method that estimates by maximum possibility (ML) each model using its defining set of nonlinear constraints. This technique produces in most cases very reliable quotes.
Computes multiple elasticity actions to get a better understanding of the magnitude of parameter estimates
The authors have the ability to estimate own price, cross-price and income elasticities for each and every demand equation. They also determine the own and cross-advertising elasticities. The combination advertising elasticity has been further decomposed into predatory advertising elasticity and global advertising elasticity based on the concepts of predatory and spillover results presented by Roberts and Samuelson (1988). Predatory advertising elasticity provides rate of change of the marketplace share of organization j the effect of a 1% upsurge in the advertising of organization i. Global advertising elasticity gives the rate of change of the total market demand caused by a 1 % upsurge in the advertising of company i. The spillover effect of advertising corresponds to a distribution of the change in total demand due to the advertising of company i compared to the show of company j. This is defined from the decomposition of mix advertising elasticity.
The model permits various moving over regimes specifications
The price of Coca-Cola revealed an unusual increase in fall season 1976 and was immediately followed by a sharp fall season. This period corresponds to the middle-1970s sugar problems. So on the basis of the observation the writers have lengthened their work by formulating and estimating models with two regimes- before and after 1976. Both routine models are approximated using a switching dummy variable which has the role of imposing the correct group of nonlinear constraints on the general linear model within each period. The results show that Coca-Cola is a Stackelberg leader in price and advertising until 1976, which collusion in advertising and competition in price takes place after 1976.
Extends the conjectural deviation strategy and compares it with the collusive models used
The authors stretch the original conjectural variation method of the truth of differentiated products with two control parameters- price and advertising. When contrasting this process to theirs, the creators find evidence that their collusive models cannot be viewed as special circumstances of conjectural approach. The conjectural model and any of the collusive models used in the newspaper are nonnested. Specifically, the conjectural model imposes another set of limitation on the parameters of the overall linear model compared to the restrictions imposed by collusive models.
