Embedded Files
Publications
Cross-border shopping of alcoholic beverages: Evidence from a natural experiment. With Kyrre Rickertsen. Food Policy. January 2026. [replication scripts]
Cross-border shopping of alcoholic beverages: Evidence from a natural experiment. With Kyrre Rickertsen. Food Policy. January 2026. [replication scripts]
Working papers
From forecast to frontline: Optimizing staffing for an alcohol retail monopolist
with Ritvana Rrukaj. [download]
From forecast to frontline: Optimizing staffing for an alcohol retail monopolist
with Ritvana Rrukaj. [download]
Abstract
In companies where success mainly depends on balancing staffing efficiency with service quality, accurate forecasts of future demand are crucial. We study an example of this and develop seasonal ARIMA models with exogenous regressors (SARIMAX) to inform planning and staffing decisions in Vinmonopolet, the Norwegian retail monopolist for off-premises sales of wine and hard liquor. Our out-of-sample forecasts missed the aggregated national sales in 2024 by only 0.14%, outperforming Vinmonopolet’s own estimate which missed by 1%. On the store level, our forecasts have an accuracy of around 1%. Staffing based on these forecasts would bring Vinmonopolet closer to optimal staffing levels, balancing cost efficiency against desired customer service levels, compared with realized staffing in 2024.
Figure: Staffing numbers for Store 1
Notes: Optimal staffing is based on a liter per hour (LPH) measure used by Vinmonopolet. The forecasted hours is based on the out-of-sample liter forecast and the LPH measure. Dashed lines represent 80% prediction intervals (PI).
Tax expenditures and behavioural responses: The case of duty-free exemptions
with Odd E. Nygård. [download]. Revise and resubmit in Fiscal Studies.
Tax expenditures and behavioural responses: The case of duty-free exemptions
with Odd E. Nygård. [download]. Revise and resubmit in Fiscal Studies.
Abstract
Tax expenditures are the losses in revenue from exempting parts of the tax base from taxation. The conventional method for computing tax expenditures disregards behavioural effects. Using an empirically based demand model, we simulate two reform scenarios that repeal current tax expenditures related to on-arrival duty-free sales of alcohol and tobacco in Norway. The model includes both recorded and unrecorded consumption, tracking all demand responses and their impacts on the tax base. Our results suggest that incorporating these behavioural effects reduces tax expenditures calculated by the conventional approach by more than one third.
Presented at:
Figure: Consumtion of pure alcohol (1,000 liter)
Notes: The figure illustrates the estimated consumption of alcohol from recorded (taxed) and unrecored (untaxed) sources pre-reform and in the two reform scenarios. Reform 1 repeals duty-free sales upon arrival at Norwegian airports. Reform 2 sets the personal allowances for alcohol and tobacco to zero and repeals duty-free sales upon arrival at Norwegian airports. Pure alcohol figures are calculated based on the average alcohol by volume observed in recorded consumption.
Page updated
Google Sites
Report abuse