740 D. Lo and. h chan At this point, one serious query can arise regarding the comparability between the trade and production categories. It might well be possible that the production function of the exporting firms is rather different from those producing for the domestic markets-that is, there might exist a dualistic structure of the industries Inder consideration. If this is the case, the fact that the electronics industry (and the broader machinery sector)is capital intensive relative to other manufacturing activity then would not imply that its exports are similarly capital intensive. Unfortunately, as far as we know, there does not exist any direct empirical evidence from which this query can be verified. What we do have in hand is the indirect evidence regarding the exports-to-output ratios of the industries concerned. As can be seen from Table 5, the exports-to-output ratios of the electronics industry and the broader machinery sector have both exceeded those of all industries since the mid 1980s, and the differentials have increased substantially over time. The implication is that, if there does exist a dualistic structure of production where mechanical and electronic exports embody a much lower level of capital-intensity than those sold in the domestic market, the increased exports-to-output ratios(relative to the ratios of al industries) would tend to pull down the relative capital-labour ratios and the relative labour productivity of the industries concerned. Hence, taken together, the data in Tables 3 through 5 indicate that such dualistic structure is unlikely to exist-unless Table 5. Exports-to-output ratios, 1980-95. Sources: All industries figures from ZTN, arious years; mechanical industry figures from ZGN, various years; electronics industry igures from ZDGN, various years. All industries Machinery and electronics Electronics (b)(b)(a)(a) (b)(b)(a) 212.35% 5400 3.63%839 03.58%83 1.61% 2273.91%91 2493.85%1072 3.54%109 4.58% 3154.14% 5.32% 4.12%1882 111946886.15%2056 13 9777.07%2619 11.40% l1.48% 6.54 912.40% 929%4989 53310.68%869 0.96% 6625296711.14%6183 75212.16%1134 20.95% 0.81%8929 10761205%1710 48402 43258.94% 7017687241243%18031275615.28% 39.95% 918941065311.59%218983671 %3373136740.53% Notes: (a)=gross value of industrial output at current prices(100 m yuan); (b)=exports value at an), converted from USS data at official exchange rates. Output data cover enterprises at all vels, i.e. including establishments at and below the village level. and those that are not registered as dependently accounting industri ses. Output data of mechanical and electronics industry ar ed from constant-price figu ng the ex-factory price index of machinery as the deflator. data of the electronics in nterprises both inside and outside the administration of the try of Electronics Industry. dustry cover ie 85 export data of electronics are estimated on the basis of exports by enterprises under the Ministry of Electronics Industry. 01998 John Wiley Sons, Ltd J.mnt.Dev.10,733-74901998)At this point, one serious query can arise regarding the comparability between the trade and production categories. It might well be possible that the production function of the exporting ®rms is rather di€erent from those producing for the domestic markets Ð that is, there might exist a dualistic structure of the industries under consideration. If this is the case, the fact that the electronics industry (and the broader machinery sector) is capital intensive relative to other manufacturing activity then would not imply that its exports are similarly capital intensive. Unfortunately, as far as we know, there does not exist any direct empirical evidence from which this query can be veri®ed. What we do have in hand is the indirect evidence regarding the exports-to-output ratios of the industries concerned. As can be seen from Table 5, the exports-to-output ratios of the electronics industry and the broader machinery sector have both exceeded those of all industries since the mid- 1980s, and the di€erentials have increased substantially over time. The implication is that, if there does exist a dualistic structure of production where mechanical and electronic exports embody a much lower level of capital-intensity than those sold in the domestic market, the increased exports-to-output ratios (relative to the ratios of all industries) would tend to pull down the relative capital-labour ratios and the relative labour productivity of the industries concerned. Hence, taken together, the data in Tables 3 through 5 indicate that such dualistic structure is unlikely to exist Ð unless Table 5. Exports-to-output ratios, 1980±95. Sources: All industries ®gures from ZTN, various years; mechanical industry ®gures from ZJGN, various years; electronics industry ®gures from ZDGN, various years. All industries Machinery and electronics Electronics (a) (b) (b)/(a) (a) (b) (b)/(a) (a) (b) (b)/(a) 1980 5154 135 2.62% 894 21 2.35% 78 Ð Ð 1981 5400 196 3.63% 839 30 3.58% 83 1 1.61% 1982 5811 227 3.91% 917 36 3.93% 84 3 3.73% 1983 6461 249 3.85% 1072 38 3.54% 109 5 4.58% 1984 7617 315 4.14% 1315 49 3.73% 164 9 5.32% 1985 9716 400 4.12% 1882 50 2.66% 246 8 3.20% 1986 11194 688 6.15% 2056 87 4.23% 265 23 8.55% 1987 13813 977 7.07% 2619 144 5.50% 396 45 11.40% 1988 18224 1231 6.75% 3661 229 6.26% 620 71 11.48% 1989 22017 1395 6.34% 4739 310 6.54% 796 99 12.40% 1990 23924 2222 9.29% 4989 533 10.68% 869 182 20.96% 1991 26625 2967 11.14% 6183 752 12.16% 1134 238 20.95% 1992 34599 3740 10.81% 8929 1076 12.05% 1710 378 22.11% 1993 48402 4325 8.94% 13822 1308 9.46% 2149 465 21.65% 1994 70176 8724 12.43% 18031 2756 15.28% 2665 1065 39.95% 1995 91894 10653 11.59% 21898 3671 16.76% 3373 1367 40.53% Notes: (a) ˆ gross value of industrial output at current prices (100 m yuan); (b) ˆ exports value at current pices (100 m yuan), converted from US$ data at ocial exchange rates. Output data cover enterprises at all levels, i.e. including establishments at and below the village level, and those that are not registered as independently accounting industrial enterprises. Output data of mechanical and electronics industry are converted from constant-price ®gures by using the ex-factory price index of machinery as the de¯ator. Output data of the electronics industry cover enterprises both inside and outside the administration of the Ministry of Electronics Industry. The 1981±85 export data of electronics are estimated on the basis of exports by enterprises under the Ministry of Electronics Industry. #1998 John Wiley & Sons, Ltd. J. Int. Dev. 10, 733±749 (1998) 740 D. Lo and T. M. H. Chan