Market Dynamcs and Productvty n Japanese Retal Industry n the late 1990s Toshyuk Matsuura, Research Insttute of Economy, Trade and Industry Kazuyuk Motohash, Unversty of Tokyo & Research Insttute of Economy, Trade and Industry
Ponts at ssue Motvaton Why s the performance so low? Does market selecton mechansm worked? Objectve To examne whether Japanese retals sector s n process of transformaton. To lnk the aggregated productvty growth wth mcro productvty growth.
Comparson of productvty between Japan and US - Mcknsey Global Insttute -
Reason for the productvty gap between Japan and US Why do unproductve stores reman n market? Entry barrer for Large scale stores Large scale stores low Ext barrer for Small stores Tax reducton and allowance for the owner of Pap and Mum type famly busnesses.
The falure of the Natural Selecton Mechansm Nshmura, Nakajma and Kyota[2002] Comparson of frm-level productvty among entres, contnues, and exts After 1996, productvty of exts are better than others. Has Market selecton mechansm worked well?
The Methodology Overvew of the trends n turnover by entry and ext of establshment Transton of productvty rankng wth longtudnal dataset. Productvty decomposton Regresson analyss
The Turnover n Retal Industry 15 10 5 0-5 -10-15 -20 entry(%) ext (%) net growth (%) 1988 1991 1994 1997 2002 Ext rate has been ncreasng.
Turnover and Gross Job flow Entry and Ext from 1997 to 2002 net growth(%) entry(%) ext(%) Total -19.4% 12.6% -32.0% Department stores -12.9% 8.2% -21.1% Supermarket stores -6.5% 15.5% -22.0% Specalty supermarket stores -0.5% 30.1% -30.5% Convenence stores 1.6% 30.7% -29.2% Specalty Stores (Small stores) -21.0% 11.8% -32.7% Gross Job Flow from 1997 to 2002 net growth(%) net growth by contnue(%) Job creaton entry(%) Job destructon ext(%) Total -12.0% -12.4% 15.7% -15.3% Department stores -38.2% -45.6% 12.0% -4.6% Supermarket stores 8.1% 1.2% 14.6% -7.7% Specalty supermarket stores 22.3% 30.6% 10.6% -19.0% Convenence stores 16.1% 22.7% 9.5% -16.1% Specalty Stores (Small stores) -19.0% -18.0% 10.5% -11.5%
Productvty Measurement Multfactor vs Sngle factor Captal Input s not avalable Output Gross margn s not avalable Sales deflated by CPI
Full tme equvalent labor nput Input full-tme worker and part-tme worker share of part-tmer average hours 1997 2002 worked per day Department store 23.7% 40.0% 7.1 Large superstores 57.1% 80.5% 5.5 Specalty superstores 51.2% 76.8% 5.5 Convenence store 52.6% 83.2% 4.8 Specalty store (small store) 19.6% 37.3% 6.8 Full tme equvalent labor nput =full tme worker+ (total hours worked by part-tmer)/8
Transton matrx of relatve productvty rankng Low Productvty Hgh Low Productvty Hgh Quntle 1 Quntle 2 Quntle 3 Quntle 4 Quntle 5 Ext (2002) (2002) (2002) (2002) (2002) Quntle 1 32.48 14.54 6.56 4.36 4.67 37.39 (1997) 40.35 18.06 8.15 5.42 5.8 23.32 Quntle 2 16.84 28.2 13.13 5.02 2.17 34.64 (1997) 20.92 35.04 16.31 6.24 2.7 21.61 Quntle 3 7.09 18.54 27.83 11.07 3.02 32.46 (1997) 8.81 23.03 34.57 13.75 3.75 20.25 Quntle 4 3.97 6.95 18.6 32.29 9 29.18 (1997) 4.93 8.64 23.11 40.11 11.17 18.2 Quntle 5 3.58 2.52 4.37 16.62 46.26 26.66 (1997) 4.44 3.13 5.42 20.64 57.42 16.62 Entry 26.31 15.5 15.93 17.72 24.53 20.55 12.11 12.44 13.84 19.16
P t = + + + Productvty Decomposton - by Haltwanger et.al (2002)- Productvty by sales form C s C C N X t 1 ( P P s s t t 1 t 1 P t s ( P t t ( P t P t 1 t 1 P t 1 ) s ) P t t 1 ) Wthn Between Cross term Entry Ext Productvty for establshment
Productvty Decomposton -Results- P1997 P2002 wthn between cross entry ext Total 7.5 7.4-13.4% -4.4% 8.5% -12.6% -7.9% 3.1% Department stores 8.6 8.7 5.9% 7.4% 4.7% -7.4% -1.6% 2.9% Supermarket stores 8.2 8.0-23.5% -10.6% 4.7% -5.4% -12.4% 0.1% Specalty superstores 8.0 7.8-20.1% -2.7% 8.0% -9.3% -17.9% 1.9% Convenence stores 7.3 7.3-4.5% 11.7% 12.1% -17.0% -12.7% 1.3% Specalty Stores 6.7 6.6-9.9% -1.2% 10.9% -17.4% -5.9% 3.7% (Small stores) Postve between share; Productve stores ncrease ts share. Negatve cross effect; Downszng has been productvty enhancng for contnuous stores. Postve ext effect and Negatve entry effect P
ln ln Output Emp Cross Secton Regresson Analyss Output Emp = (1 = (1 Observaton n 1997 Cap j j α )ln + β X + γ D ext + Emp [ ] j Observaton n 2002 Cap j j α ) ln + β X + η D entry + Emp [ ] j ε ε X; Controllng factor, e.g. age, the rato of nonretal revenues, ndustry dummy etc. Comparng ext wth contnue, and entry wth contnue.
Regresson Analyss -Results (1)- 1997 level Dependant varable: ln(productvty) all Department Store Large superstore Large specalty store Convenence store Small store dummy -0.247-0.242-0.107-0.105-0.097-0.259 for ext (119.82)** (4.87)** (2.00)* (8.99)** (8.39)** (112.97)** Observatons 1204973 447 1888 31807 36579 974718 R-squared 0.17 0.32 0.16 0.16 0.26 0.16 Absolute value of t statstcs n parentheses * sgnfcant at 5%; ** sgnfcant at 1% Productvty of exts are lower than others even after controllng for varous factors.
Regresson Analyss -Results (2)- 2002 level Dependant varable: ln(productvty) all Department Store Large superstore Large specalty store Convenenc e store Small store dummy -0.233 0.147 0.317 0.136 0.122-0.257 for entry (49.15)** (0.86) (3.40)** (6.91)** (7.97)** (46.53)** Observatons 1004210 328 1588 33883 38803 798355 R-squared 0.13 0.16 0.12 0.12 0.21 0.12 Absolute value of t statstcs n parentheses * sgnfcant at 5%; ** sgnfcant at 1% For large stores, entres have hgher productvty than contnues when other factors controlled.
Concluson Natural Selecton Mechansm has been workng well. The entry and ext have contrbuted to the mprovement of aggregated productvty growth, although ts growth rate s negatve. Further research on the effect of nsttutonal factor s necessary.