Downloads - empirica

empirica.biz

Downloads - empirica

e-Business in the chemical, rubber and plastics industryThe impact of ICT on labour productivity growthIn existing literature, the chemical industries have not been identified as one of thosesectors where intensifying ICT-capital usage contributed significantly to aggregate labourproductivity growth acceleration in the US (see van Ark, Inklaar, McGuckin 2002).Therefore, this study analyses the particular developments and factors which led in theselected sample of EU Member States to fairly different overall outcomes than those thatcould be expected when the Lisbon Agenda was setup in 2000. At this time, it wasexpected that a similar resurgence of productivity growth due to ICT investments asobserved in the US since the mid-1990s would take place in Europe (as a catching-upprocess to the US as leader in development).The following analysis is based on 16 out of the 27 EU Member States. 122 For thesecountries, a complete dataset running at least over the time period from 1995 until 2004was available from the EU-KLEMS database. They are denoted as the “EU-16” in thissection. Data are available for the following variables: gross production value, totalintermediate inputs, total working hours, ICT-capital stock and non-ICT capital stock inputplus total working hours. The latter is broken down into working hours for three separateskill categories: high, medium and low skills.Based on the secondary intermediate inputs and the two primary input factors (capitalbroken down into ICT- and non-ICT-capital stock) and labour measured by working hours(broken down into three different skill-types), a stochastic possibility frontier (SPF) 123was estimated using a panel data set for the EU-16.As a particular specification the error component model of Battese and Coelli (1992) wasused, which allows for estimating average efficiency levels by country (i.e. 100 is equal tofull-scale efficiency, values below measure the percentage points below the overallefficiency level of an industry production possibility frontier at a certain time period). Toguarantee constant returns to scale for the possibility frontier, the output and inputvariables were normalised by the total working hours. This led to an accordingly restrictedstochastic possibility frontier where the real gross production value per working hour isexplained by six factor intensities using total working hours as the denominator. As anadditional variable, a time trend beside the constant term was included to measure theautonomous technical change. For the econometric estimation the Frontiers 4.1 softwareprogramme was used (Coelli, 1996). The estimation results using a Cobb-Douglasproduction function specification are summarised in Exhibit 4.1-8.The parameter estimates obtained are measures for the respective output elasticity of therespective input factor, i.e. an increase of one unit in the respective input factor increasesthe output variable by the respective output units.122 Austria, Belgium, Czech Republic, Denmark, Finland, France, Germany, Hungary, Italy,Luxembourg, Netherlands, Poland, Slovenia, Spain, Sweden, and the UK.123 A "stochastic possibility frontier" introduces a theoretical benchmark which usually cannot bematched by any actual producer, industry or country. It is a quasi-ideal production frontierwhich, due to all kinds of impediments in the particular situations of each player, cannot bematched completely (at least permanently). This gives sufficient incentive for even the bestpracticeplayer to search for further improvements. For detailed information about themathematical model, see Annex II, and Erber 2005.106

More magazines by this user
Similar magazines