Discussion Paper, Zentrum für internationale Entwicklungs- und Umweltforschung 62
Pricing-to-market (PTM) evidence in German sugar confectionery exports is examined, combining the original fixed-effects model of Knetter (1989) and an errorcorrection specification (ECM) at three frequency levels, to assess how neglecting time-series properties and the choice of frequency affect the outcomes. In order to ensure validity of unit values as price proxies, the type of competition with every destination market is evaluated, proving price-driven competition with Canada, Sweden, the UK and the US, but not with Switzerland. Results show that fixed-effect model findings of PTM might be spurious, if time-series properties of the data are not considered. German exporters seem to exploit their market power and adjust their markups to protect market shares in strategically important expanding destinations. Local currency price stabilization (LCPS) was found for Canada and the UK, while LCPS for the exports to US in fixed-effects model turned out to be an erroneous result, as cointegration was rejected. Data of the higher frequency was suggested to be preferable for PTM studies, once measurement error due to heterogeneity is minimized. Finally, using marginal costs estimates from a fixed-effects model as cost proxies in the ECM improves the quality of the model and reveals a higher degree of PTM.
German agri-food exports pricing-to-market sugar confectionery