本文加入動態調整的考量重新驗證90年代之生產力逆說,探討各製
造業採用資訊科技所產生之動態調整與生產力成長。本文應用動態
的要素需求模型, 假設產業以極小化預期未來生產成本貼現值為目
標, 考慮準固定要素的落後生產效果, 及自動化科技在生產過程之
調整效果。在變數設定方面,將勞動與準固定要素劃分為一般勞動、
自動化相關勞動及一般資本、自動化資本。實證結果推翻生產力逆
說,並得出高科技產業比較於傳統產業,能以較低調整成本、較快速
調整速度達到長期均衡, 顯示台灣高科技廠商之投資決策較佳。再
者, 高科技產業由於使用自動化科技的層面較廣, 因此在自動化科
技的產出貢獻就比傳統產業大; 而自動化勞動在生產力上又比自動
化資本來的高, 這表示相關人員的訓練比不斷投資自動化資本來的
重要。

This paper reconsiders the 1990s productivity paradox for information communication
technology (ICT), controlling for industry heterogeneity, endogeneity
of ICT investments, and associated dynamic adjustments. We apply
a dynamic factor demand model that assumes each firm derives an optimal
plan such that the expected present value of current and future cost streams
is minimized. Labor inputs are divided into ICT labor and non-ICT labor;
and capital inputs are divided into automation capital and physical capital.
The empirical evidence shows the followings. First, the adjustment speed
of quasi-fixed inputs – ICTs and physical capital – are relatively faster in
high-technology industries compared to other industries, showing Taiwan’s
high-technology industries may have performed better in facing environmental
shocks. Second, the contribution of ICTs to high-technology industries
is more significant than to traditional industries due to wide uses of the
ICTs. Third, the productivity of automation labor is higher than ICT capital,
indicating that the investment in ICT capital should be accompanied by
automation personnel training.