* 2009 6 15 2006 10 24 MNP MNP MNP MNP MNP 18% 2.6% Keywords: Mobile Number Portability (MNP), Switching Cost, Nested Logit Model, Choice-based Sampling JEL Classification: D12, L96 1. 2006 10 24 (Mobile Number Portability MNP) MNP * ( ) ( ) ( ) Empirical Micro Forum 2 7 22 1 E-mail: t-kitano@grips.ac.jp 1
MNP 1 2007 3 9600 2 MNP 3 MNP MNP 2 (Nested Logit Model NL ) MNP MNP MNP MNP 1 Klemperer(1987) 2 3 MNP MNP 2004 3 2
MNP Choice-based Choice-based Manski and Lerman(1977) Weighted Exogenous Sampling Maximum Likelihood(WESML) 2 Shy(2002) Kim(2003) 4 Shum(2004) Chen and Hitt(2002) 5 MNP 6 MNP 2000-2300 MNP 300-350 MNP 18 MNP 25-35 2.6% 4 Kim 2007 5 Farrel and Klemperer(2008) Klemperer(1995) 6 MNP Ida and Kuroda(2009) Lee et al.(2006) 3
MNP 2 3 4 5 4 MNP 6 2. MNP 2.1 MNP 2006 10 24 Choice-based 7 3 (NTT au ) 2-1 (i) 10 24 ( 1000) (%) (%) (%) NTT 50 50 10 10 au 30 15 20-50 85 20 35 60 5 (ii) 10 24 ( 500) (%) (%) (%) NTT 20 50 10 10 au 60 15 20-50 85 20 35 60 5 7 500 1000 531 1047 2007 3 MNP 2% 4
2-1 2.2 MNP MNP MNP 2-2 MNP 64% MNP MNP MNP 2-3 3 MNP 5 MNP MNP 6 2-2 MNP MNP 339 64 MNP 192 36 531 100 2-3 MNP 1. MNP 12 2. MNP 3 3. MNP 60 4. 2005 10 11 au MNP 31 5. 30 6. 24 7. PHS 17 8. 15 192 5
MNP MNP 8 3. (NTT au ) MNP 2006 10 24 MNP MNP MNP 3-1 A [ A B C ] 3 B C [MNP MNP ] B C 1 2 MNP NL 9 NL MNP MNP MNP 8 10 24 1 MNP 12 4 au 7 PHS 9 NL 6
3-1 MNP 10 h i h, i h h McFadden(1978) i j U i, j, MNP = = V P S M ( α + x ' α )* p + ( β + x ' β + ( γ + x ' γ )* MNP ) 0 + x' δ 0 i 1 + ε i, j, MNP ( λ) ( θ) + V ( γ) + ε ( λ) i, j, MNP 0 i i, j, MNP 1 0 i 1 * SWITCH (3-1) MNP MNP 1 0 MNP MNP h U ih V V i, P S ( θ) = ( α0 + xi ' α1) * p + ( β0 + xi ' β1) * M j, MNP ( γ) = ( γ 0 + xi ' γ1) * MNP ) SWITCH * SWITCH + x' δ 0 (3-2) MNP 10 NL Independence of Irrespective Alternatives (IIA) MNP 7
x P i x S i x M i x NTT au MNP i j MNP 1 0 SWITCH j 1 0 θ=(α 0, α 1 ', β 0, β 1 ', δ 0 ')' γ=(γ 0 γ 1 ')' MNP ε i,j,mnp (λ) (Generalized Extreme Value GEV) λ λ McFadden(1978) 0 1 11 p i j 11 λ 1 2 Gumbell ( 1 ) 8
12 2006 p 13 3.2 NL i j h MNP MNP P i,mnp j P P i, j, ( θ, γ, λ) = P ( θ, λ) Pi, MNP j ( θ, γ, λ) (3-3) MNP (3-3) exp( V ( θ) + λi ( γ, λ)) P ( θ, γ, λ) = (3-4) exp( V ( θ)) + exp( V ( θ) + λi ( γ, λ)) ih l h il il 12 (discrete-continuous choice model) (Hanemann(1984)) 13 5000 50% 9
P i, MNP j ( exp( V i, j, MNP ( γ) λ) γ, λ) = (3-5) { 0,1 } exp( V,, ( ) ) MNP i j MNP γ λ I MNP 14 I ( γ, λ ) = ln( exp( V ( ) )) { 0,1},, λ MNP i j MNP γ (3-6) = ln(1 + exp(( γ 0 + x M i ' γ 1 )* SWITCH λ )) h exp( Vih ( θ)) Pih ( θ, γ, λ) = (3-7) exp( V ( θ)) + exp( V ( θ) + λi ( γ, λ)) ih l h il il Choice-based SWITCH Manski and Lerman(1977) Weighted Exogenous Sampling Maximum Likelihood (WESML) Choice-based WESML WLL( θ, γ, λ) = [ w y ln P + w y ln P ] (3-8) h i N ih ih ih j h MNP { 0,1} i, j, MNP i, j, MNP h, N h h y ih i N h h 1 0 y i j MNP j MNP 1 0 w i h 14 MNP 0 NL Train(2003) Ben-Akiva and Lerman(1985) 10
1 QS if j = h 1 H S w = (3-9) QS otherwise H S Q S H S 2007 3 MNP 200 1 2% MNP 2 MNP 200 2 800 MNP MNP 800 8% 15 MNP MNP MNP 8% 6% (3-8) (θ γ λ) (ˆ, θ γˆ, ˆ) λ = arg maxwll( θ, γ, λ) (3-10) ( θ, γ, λ ) 4. 15, HS MNP MNP H S 11
MNP 4.1 4-1 4-1 MNP 4 SWITCH 4-1 (i) (ii) 0.4907 0.3086 0.4612 0.2678 2.9930 3.4679 3.0433 3.4282 112 111 112 111 0.352 0.478 0.310 0.463 0.342 0.474 0.333 0.472 0.098 0.297 0.082 0.275 0.150 0.357 0.148 0.355 1279 912 (iii) (iv) MNP 0.5641 0.3828 0.5853 0.4162 2.8678 3.5663 3.1791 3.8529 112 113 121 113 0.455 0.499 0.459 0.499 0.362 0.481 0.328 0.470 0.136 0.344 0.160 0.368 0.155 0.363 0.097 0.297 439 314 (5000 50% ) 1279( 1537 ) 3-2 3-3 MNP 12
16 3 382 759 MNP γ MNP 3.2 MNP 3-2 MNP 64% MNP γ 4-1 MNP 15% MNP 10% MNP ε i,j, MNP (Hausman(1996) Nevo(2001)) 16 Shum(2004) 13
4.2 4-2 8% 6% 3-18 2 4-2 NL 1 (i) WESML(8%) (ii) WESML(6%) (iii) ML( ) -17.780 2.206 *** -17.599 2.637 *** -18.215 1.322 *** * 0.935 0.731 0.918 0.887 0.923 0.424 ** *( ) 2-0.024 0.038-0.024 0.048-0.023 0.020 switch -3.394 0.572 *** -3.778 0.758 *** -2.062 0.261 *** switch* 0.519 0.374 0.537 0.466 0.500 0.201 ** switch* 0.084 0.362 0.078 0.447 0.086 0.200 switch* 0.524 0.508 0.492 0.627 0.635 0.294 ** au 0.505 0.307 0.526 0.388 0.440 0.149 *** au* 0.836 0.469 * 0.818 0.584 0.881 0.240 *** au* -0.935 0.461 ** -0.949 0.573 * -0.831 0.236 *** au* -0.042 0.684 0.003 0.867-0.223 0.341-1.916 0.322 *** -1.920 0.402 *** -1.899 0.166 *** * 0.532 0.469 0.547 0.578 0.443 0.257 * * -0.839 0.477 * -0.841 0.590-0.825 0.261 *** * 0.077 0.573 0.155 0.685-0.216 0.353 MNP 2 0.422 0.416 0.479 0.572 0.292 0.160 * 0.001 0.001 0.001 0.001 0.001 0.001 * -0.617 0.507-0.681 0.661-0.457 0.234 * λ 0.409 0.368 0.450 0.495 0.310 0.162 * Log likelihood 454.424 367.410 886.609 *** ** * 1 5 10% 14
SWITCH SWITCH (iii) (ii) 6% 1.5 SWITCH SWITCH (iii) au au NTT NTT au 8% au NTT au 8% au au NTT (2006) au LISMO NTT i 17 MNP 4-2 (iv) MNP Manski and Lerman(1977) WESML 17 au NTT 15
(Greene(2008) ) 18 λ McFadden(1978) 0 1 2 MNP MNP MNP MNP 5.1 Klemperer(1995) 19 18 Choice-based Cosslett(1981) Imbens(1992) Choice-based 19 Klemperer(1995) (1) (Need for compatibility with existing equipment) (2) (Transaction costs of switching suppliers) (3) (Costs of learning to use new brands) (4) (Uncertainty about the quality of untested brands) (5) (Discount coupons and similar devices) 5 16
, MNP MNP MNP MNP j exp( V ( θ)) P ( θ ) = (5-1) exp( V ( θ)) l s s il P ( p s, SWITCH = 1) = P ( p, SWITCH = 0) (5-2) p (4-2) 1 s ( 0 x S = β + i ' β) α i (5-3) α i = α 0 + x P i 'α 1 (5-4) (5-3) MNP (3-4) MNP j λi MNP MNP MNP s MNP j MNP j MNP MNP ~ MNP P ( p s, SWITCH = 1) = P ( p, SWITCH = 0) (5-5) 17
~ P ( p s MNP, SWITCH = 1) = exp( V exp( V ( θ, SWITCH ( θ, SWITCH = 1) + λi = 1) + λi ) ) + l j exp( V il ( θ)) (5-6) i h l h,j (5-5) (5-6) MNP MNP 1 s [ λ ln(1 exp(( γ 0 x M s = + + i ' γ1) λ))] (5-7) α i MNP 5-1 4-2 (i)-(iv) (5-3) (5-7) (s ) MNP (s - s MNP ) 1100-2300 MNP 25% 1 8% 6-8% MNP 2057-2328 2000-2300 20 MNP 5-1 5-1 8% 5-1 MNP (i) WESML(8%) (ii) WESML(6%) (iii) ML( ) ( ) 2057 367 2328 397 1145 273 MNP ( ) 311 137 350 154 218 96 (%) 17.8-17.7-23.5-20 2057 18
6% MNP 3 MNP MNP MNP MNP MNP MNP MNP 19
NL 2 MNP MNP 8% 6% 298 409 119 5.2 MNP MNP MNP MNP MNP NL MNP 21 [ ln( exp( Vih ( θ) ) + exp( Vil ( θ) + λi il ( γ, λ) )) ln( exp( V ( )))] 1 E( CSi ) = l h j θ α i (5-8) 8% 6% 5-2 21 Small and Rosen(1981) (6-9) McFadden(1999) Herriges and Kling(1999) (Train(2003)) 70 5-8 20
(i) WESML(8%) (ii) WESML(6%) (iii) ML( ) 35 70 25 55 67 93 MNP 6-8% 25 35 MNP 5-2 8% 6% 5-2 8% 6% 0-30 MNP 21
MNP MNP 3 MNP MNP MNP MNP MNP MNP 5.2 MNP MNP MNP 5.1 MNP 5-3 MNP 8% 6% 3 MNP 8% 6% 10.25% 8.56% MNP MNP 8% 6% 2.77% 2.42% MNP 2.6% 8% (i) WESML(8%) (ii) WESML(6%) (iv) ML( ) 13.02 10.25 2.77 8.56 6.14 2.42 32.09 29.59 2.50 22
6% 87% 91% MNP 3 2 8% 6% 25 22 27 6. MNP MNP 2000-2300 MNP 18% MNP 25-35 2.6% MNP 2 MNP MNP MNP MNP MNP MNP 23
(state dependence) (taste heterogeneity) Heckman(1981) (spurious state dependence) MNP Ben-Akiva, M., and S. R. Lerman, 1985, Discrete Choice Analysis: Theory and Application to Travel Demand, MIT Press. Chen, P., and L. M. Hitt, 2002, Measuring Switching Cost and the Determinants of Customer Retention in Internet-Enabled Businesses: A study of the Online Brokerage Industry, Information Systems Research, 13(3), 255-274. Cosslett, S. R., 1981, Efficient Estimation of Discrete Choice Models, in Structural Analysis of Discrete Data with Applications, ed. By C. F. Manski and D. McFadden, Cambridge, MA: MIT Press, 51-111. Farrell, J., and P. Klemperer, 2008, Coordination and Lock-In: Competition with Switching Costs and Network Effects, in Handbook of Industrial Organization, vol. III, North-Holland, Amsterdam. Greene, W. H., 2008, Econometric Analysis, Prentice Hall, 6 th Edition. Hanemann, M. W., 1984, Discrete/Continuous Models of Consumer Demand, Econometrica 52(3), 541-563. Hausman, J., 1996, Valuation of New Goods under Perfect and Imperfect Competition, in T. 24
Bresnahan, T. and R. Gordon eds., The Economics of New Goods, Studies in Income and Wealth Vol.58, Chicago: National Bureau of Economic Research. Heckman, J., 1981, Heterogeneity and State Dependence, in S. Rosen ed., Studies in Labor Markets, Chicago: University of Chicago Press, 91-139. Herriges, J. A., and C. L. Kling, 1999, Nonlinear Income Effects in Random Utility Models, Review of Economics and Statistics, 81(1), 62-72. Ida, T. and T. Kuroda, 2009, Discrete Choice Model Analysis of Demand for Mobile Telephone Service in Japan, Empirical Economics 36(1), 65-80. Imbens, G. W., 1992, An Efficient Method of Moments Estimator for Discrete Choice Models with Choice-Based Sampling, Econometrica 60(5), 1187-1214. Kim, J., 2006, Consumer s Dynamic Switching Decision in the Cellular Service Industry, Manuscript, University of Wisconsin. Kim, M., D. Kliger, B. Vale, 2003, Estimating Switching Costs: the Case of Banking, Journal of Financial Intermediation, 12, 25-56. Klemperer, P., 1987, The Competitiveness of Markets with Switching Costs, RAND Journal of Economics, 18(1), 138-150. Klemperer, P., 1995, Competition When Consumers Have Switching Costs: An Overview with Applications to Industrial Organization, Macroeconomics, and International Trade, Review of Economic Studies, 62(4), 515-539. Lee, J., Y. Kim, J. Lee, and Y. Park, 2006, Estimating the Extent of Potential Competition in the Korean Mobile Telecommunications Market: Switching Costs and Number Portability, International Journal of Industrial Organization, 24, 107-124. 25
Manski, C. F., and S. R. Lerman, 1977, The Estimation of Choice Probabilities from Choice Based Samples, Econometrica 45(8), 1977-1988. McFadden, D., 1978, Modeling the Choice of Residential Location, in A. Karlqvist, L. Lundqvist, F. Snickars, and J. Weibull eds., Spatial Interaction Theory and Residential Location, North-Holland, Amsterdam. McFadden, D., 1999, Computing Willingness-to-Pay in Random Utility Models, in J. Moore, R. Riezman, and J. Melvin eds., Trade, Theory and Econometrics: Essays in Honor of John S. Chipman, Routledge: London. Nevo, A., 2001, Measuring Market Power in the Ready-To-Eat Cereal Industry, Econometrica 69(2), 307-342. Shum, M., 2004, Does Advertising Overcome Brand Loyalty? Evidence from the Breakfast-Cereal Market, Journal of Economics and Management Strategy, 13(2), 241-272. Shy, O., 2002, A Quick-and-Easy Method for Estimating Switching Costs, International Journal of Industrial Organization, 20(1), 71-87. Small, A., K., and H. S. Rosen, 1981, Applied Welfare Economics with Discrete Choice Models, Econometrica, 49(1), 105-130. Train, K., 2003, Discrete Choice Methods with Simulation, Cambridge University Press. (2008) Web2.0 (2004) http://www.soumu.go.jp/s-news/2004/pdf/040427_4_bt1.pdf (2006) 2006 26