So far, the analysis only includes costs for current
maintenance. But it is obvious that spending for reinvestment
purposes is relevant in our context. The costs for
maintaining tracks increase with time and sooner or later,
the annual maintenance cost will be high enough to warrant
a renewal of the facility; rather than spending increasing
sums of money on current maintenance, it is cheaper to
replace tracks-and-sleepers on a longer section of the line, to
change switches, etc. thus saving money in the future.
The timing of the track renewal is related to the traffic load;
the more traffic over a track unit, the more frequent is the
renewal frequency. This is the logic for also including
renewal costs in any comprehensive analysis of infrastructure
spending.
The Finnish, but not the Swedish, material includes
information about this type of spending. However, we only
have observations from three specific years rather than the
long periods of time that would be required to get an
understanding of the cost structure. With sufficiently long
observation periods and stable external factors, including
traffic load, reinvestment could be expected to be more
frequent on track units with more traffic than on those
with less.
The last two columns of Table 5 provide coefficient
estimates when spending on renewal has been added to
current maintenance. A first observation is that the
explanatory power of the model falls to R2 ¼ 50%: Only
two variables are now significant with the electrification
dummy and the squared utilization capturing most of the
effect on the cost. We abstain from drawing any further
conclusions from this 3-year dataset.