Improvements in the productive capacity of finite resources are likely to be the key to long term economic growth in developed economies such as Sweden. Measuring productivity and decomposing it into the sources of growth is therefore important for assessing an economy's performance as well as targeting policies that can boost productivity. Despite the prevalence in the literature of the growth accounting approach to productivity measurement, this method is based upon some restrictive and often unrealistic assumptions, which precludes accurate identification of the drivers of productivity change.
This thesis will present an econometric method for measuring productivity using the Lowe index, stochastic frontier analysis and a decomposition method that will allow the sources of productivity change to be determined, whilst making none of the restrictive assumptions that accompany growth accounting. Using Bayesian econometric techniques, estimates of total factor productivity change have been obtained for the Swedish economy for the period 1970-2007. The results show that total factor productivity has declined since the early 1990s. As this method allows for a detailed decomposition of the changes, and therefore identification of the drivers, the decline can be explained by changes in the scale and mix of inputs. In particular, strong growth in capital has altered the mix of inputs to be less efficient compared to 1970. Increased investment by firms in capital is likely to have been the result of factors such as lower corporate taxes and interest rates over the 1990s and 2000s, as well as the proliferation of information and communication technology related capital.