I have been presenting the Toll Index in “raw form” putting off having to do the work to account for structural changes in the MAUT rules (e.g. including trucks of lower tonnage, adding more highways etc). In July of 2024 yet another change took place so I decided to do the econometric work to produce an index without the effect of the various MAUT reforms (see here).
So the newly defined Toll Index is based on border crossing lorries (inbound and outbound) divided by the number of working days in each month. Then we regress the result on:
- time
- reform dummies
- reform dummies interacted with time
- month-of-year fixed effects
- quarter-of-year fixed effects
Taking the residuals after removing the predicted effects of 2, 3, 4 and 5 above and scaling so that the value is 100 in January of 2007 gives us the Toll Index.
In the figure below we jointly plot the Toll Index and the seasonal and calendar adjusted Index of Industrial Production.
The correlation is visible to the naked eye. In the graph we report this month’s value, the overall mean, the previous month and the same month a year ago. Comparing with the mean tells us how unexpected the new value is (the z-score shows the eccentricity of our current state), comparing with the previous month gives the monthly derivative and comparing with the same month last year gives us the annual derivative. In this chart we are mildly below the mean.
The Toll Index was first proposed in IZA DP5522 which was published in the Journal of Forecasting. It has been widely covered in national and international media (selection):
Focus Magazin,
Tim Harford – The undercover economist,
Financial Times,
MoneyWeek,
WirtschaftsWoche,
CNN International,
DRS3 Swiss public radio,
Deutsche Welle.
The German statistical office, in cooperation with the Bundesamt für Güterverkehr, has taken the MAUT data in its portfolio of data products and their efforts can be found here. The Destatis document describing the data is here and here is their publication calendar for 2019.
Citation: “Nowcasting business cycles using toll data.” Journal of Forecasting 32:4 (2013): 299–306(with K. F. Zimmermann).