Download presentation
Presentation is loading. Please wait.
Published byArleen Lambert Modified over 9 years ago
1
NETWORK NEUTRALITY AND DIFFERENCE IN EFFICIENCY AMONG INTERNET APPLICATION SERVICE PROVIDERS : A META-FRONTIER ANALYSIS DAEHO LEE, JUNSECK HWANG 電管碩一 R96001023 白書瑄 Telecommunications Policy 35, 2011 764-772
2
Outline Introduction Econometric methodology Data Meta-frontier function estimation results Estimation of the determinants of differences in efficiency Summary and Conclusion Telecommunications Policy 35, 2011 764-772
3
Introduction With the development of broadband Internet, many applications have been actively developing in the market Among these applications, newly developed Internet application services tend to require high bandwidth and a wide and stable network becomes a very important determinant of the quality of those services. For this reason, network neutrality (NN) proponents assert that ISPs have greater incentives to discriminate against application services, as such services use more bandwidth and are more sensitive to latency Telecommunications Policy 35, 2011 764-772 ISPs Wide bandwidth & Stable network The bargaining power ASPs Degrading Internet access service or increasing the price of the service
4
Introduction The study investigates the issue of NN and tests the arguments of the proponents and opponents of NN on whether the efficiency of ASPs will decrease as their requirements of a wider and more stable bandwidth increase The study categorizes application services into four distinct groups according to the attributes of bandwidth usage and latency sensitivity, and compares the efficiency of those four groups A parametric frontier function model The technology gap ratio (TGR) The meta-frontier function Telecommunications Policy 35, 2011 764-772
5
Econometric methodology This paper estimates the efficiency using stochastic frontier analysis (SFA), a parametric approach to the empirical estimation of production functions For specification and estimation simplicity, the model for the jth group is assumed to be of the form (1) From Battese et al. (2004) notation, the meta-frontier production function model is defined as (2) Β* is the vector of unknown meta-frontier parameters that satisfy (3) Telecommunications Policy 35, 2011 764-772
6
The decomposed into TGR and technical efficiency (TE) components\ parametric frontier function model Use two approaches to estimate the parameters of the meta-frontier function: Linear programming (LP) is to minimize the sum of absolute deviations, Quadratic programming (QP) is to minimize the sum of squared deviations Telecommunications Policy 35, 2011 764-772 Econometric methodology
7
Data This study uses data obtained from four groups distinguished according to the bandwidth usage and latency sensitivity of the ASP services 1) User-created content (UCC) 2) Online games (GAME) 3) Online music (MUSIC) 4) Web services(ex: Weblogs, online communities, and Web portals) Telecommunications Policy 35, 2011 764-772
8
Data The data set used in this study consists of unbalanced panel data from 2000 to 2008. The data were obtained from the Korea Information Service Co. Ltd.-Value (KIS-VALUE) database In order to maintain the homogeneity of the data sample, only firms without any offline business are selected A number of variables are employed to estimate the production function and to measure the efficiency of service providers In this study, sales, which is the sum of gross margin and COGS, is considered the output variable, and capital (K), the number of employees (L), and other costs of goods sold (M) as the input variables Telecommunications Policy 35, 2011 764-772
9
Data The summary statistics of the input and output variables in the data set used in this study to estimate the efficiency of ASPs is reported in Table 2 The correlation between output and inputs is high (range: 0.57–0.98, mean: 0.81), while those between the inputs is low (range: 0.36–0.80, mean: 0.59), indicating the high explanatory power of inputs and a low collinearity problem. Telecommunications Policy 35, 2011 764-772
10
Meta-frontier function estimation results The maximum-likelihood estimates of the parameters in the stochastic frontier models of each service provider group are obtained using the FRONTIER 4.1 program ( Coelli, 1996), while the meta-frontier function is estimated using MATLAB 7.1. This paper reports on the SFA as it has policy implications, and the random effects time-varying production model is used for the SFA estimation. Table 3 shows the estimation results Telecommunications Policy 35, 2011 764-772
11
Meta-frontier function estimation results
12
Estimation of the determinants of differences in efficiency In addition to estimating the technical efficiency of individual firms and time by using stochastic frontier without accounting for technology differences, the study estimated the model accounting for technology difference using the meta-frontier approach. In order to identify the possible determinants or sources of efficiency/inefficiency, the study estimated a Tobit model. The study uses the Tobit regression model to examine variables that explain the TGR differences between the four service provider groups. Table 5 shows the MLE estimates of Tobit regressions with TGR as the dependent variable and firm size, market share, and technology type as explanatory or determinant variables. Telecommunications Policy 35, 2011 764-772
13
Estimation of the determinants of differences in efficiency With reference to the above arguments, this study explains the TGR component with the number of exogenous variables The study uses the Tobit regression model to examine variables that explain the TGR differences between the four service provider groups Table 5 shows the MLE estimates of Tobit regressions with TGR as the dependent variable and firm size, market share, and technology type as explanatory or determinant variables Telecommunications Policy 35, 2011 764-772
14
Estimation of the determinants of differences in efficiency The results: 1. Information technology improves productivity more as the competition within the industry intensifies 2. The efficiency of one group using more bandwidth is not always worse than that of another group that uses less bandwidth, suggesting that there is no significant discrimination by ISPs against ASPs. Telecommunications Policy 35, 2011 764-772
15
The efficiency of the four groups was estimated using both the stochastic frontier production function and meta-frontier analysis. The former ignores the difference in the service providers’ technology, while the latter considers it for estimating the performance of the providers There is no significant relation between the application service characteristics (i.e., bandwidth usage and latency sensitivity) and the TGR of ASPs, thus supporting the NN opponents’ assertion that there is no significant ISP discrimination against ASPs Summary and Conclusion Telecommunications Policy 35, 2011 764-772
16
It does not consider the simultaneity between input choices and unobserved productivity shocks, and the sample selection issues due to the entry/exit of firms over the sample period. In addition, it may not detect the efficiency change due to the discrimination by the origin and destination of the packet. Because almost all Internet firms are venture companies, the data used in this study is sourced from only a small number of firms. It should be precisely analyzed whether the existence of NN, competitiveness of the firm and its size can make a difference as far as efficiency measurement and analysis is concerned. Summary and Conclusion Telecommunications Policy 35, 2011 764-772
17
Thank you for your attention The End Telecommunications Policy 35, 2011 764-772
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.