Personellrelatert materiell, varer og tjenester - en analyse til bruk i KOSTMOD

The annual costs for personnel related material, merchandise and services (P–MVT) in the Norwegian Armed Forces amount to approximately NOK 2.6 bn. These costs are directly related to manpower, and will thus in theory change according to the workforce developments in the different parts of the Armed Forces. The Norwegian Defence Research Establishment (FFI) has for many years studied personnel costs in the Armed Forces. The tool for calculating P–MVT has been in need of improvement and automation. With this purpose, a new tool was developed during the fall of 2014, and in this report we document and describe this tool and the assumptions we make. We also analyze the cost development from 2011 to 2014. The data set include accounting vouchers collected from the accounting system SAP in addition to personnel information about each personnel category and division in the Armed Forces. We collect data from 2011 to 2014, which amounts to more than 30 million vouchers all together. The tool is developed in the statistics program R, and it analyzes each voucher and defines it as either personnel or non-personnel related based on words in the description text, specific suppliers and certain other criteria. Next, we use a distribution formula to calculate the P–MVT rates for each personnel category and division. There are however some uncertainties related to the method, both due to accountancy mistakes and to the fact that the tool does not always pick up all relevant costs. The results are nevertheless stable over time, and we conclude that the method is robust and trustworthy. In this report we analyze the years 2011–2014. We find rather large differences between the different personnel categories, but smaller differences within the same category between the divisions. The highest rates are due to a large number of international positions in the relevant divisions, which causes large travel costs. We analyze the development in the different divisions and find that several divisions have reduced their P–MVT rates during the period. This is especially due to reduced travel costs. The tendency is most clear for some of the operational divisions, especially the Army and the Navy. For some of the non-operational divisions we find an opposite development. The total personnel related costs are somewhat reduced after 2011, but we cannot find that time has a significant effect on P–MVT when we do regression analysis. This result contradicts previous personnel related cost analyses at FFI. The reason is probably that the Armed Forces have been required to improve their cost-efficiency by 0.5 percent per annum the last years. We recommend that the Norwegian Armed Forces track the development of personnel related costs for each division to prevent that some costs grow out of proportions, and to learn from the efficiency initiatives that are already accomplished. In addition we recommend that the Armed Forces continue to improve their accountancy and the inventory registration of personal clothing and equipment.