Postal delivery pricing: econometric analysis

7. Conclusion

Consumer fairness is a priority for the Scottish Government. One of the key issues affecting consumers in Scotland right now is unfair delivery pricing and practices in remote and rural areas of the country. 

The Scottish Government has formulated a detailed action plan aiming to tackle the underlying causes that lead to unfair treatment of consumers. Among the key actions detailed in the plan is the development of an interactive data hub that will allow users to measure the fairness of delivery pricing to improve transparency and drive behaviour change and the development of the Scottish Parcel Delivery Map to understand consumer experiences and target interventions.

As part of the efforts to reduce unjust delivery practices, the Scottish Government commissioned Alma Economics to undertake econometric analysis as a first step towards determining when and where delivery charges are fair and unfair in Scotland. 

We collected 24,364 pricing quotes from six major delivery companies and 6,771 pricing quotes from seven national online retailers across all 1,029 postcode sectors of Scotland.^[59] Shipping quotes were requested for small, medium and large parcel sizes as well a range of bulky consumer products. We then used a linear model with company fixed effects to estimate average prices and delivery availability, controlling for package and geography characteristics that may influence how companies set their delivery policies. 

The model allowed us to estimate the relative impact of each of these characteristics and the resulting postal charge. To determine whether a postal charge is unfair or not we compared the actual postal charges with the postal charges predicted by our model; the higher the difference between the actual and predicted value, the higher the degree of discrimination that consumers are experiencing. 

In geographic terms, our analysis showed that postcodes in the Highlands and Islands face 21% higher postal charges on average compared to South Western Scotland, with certain council areas such as Na h-Eileanan Siar and the Shetland and Orkney Islands facing average surcharges of at least 25% compared to Glasgow. Similarly, postcodes in the Highlands and Islands are 31% less likely to have access to home delivery services from online retailers compared to Eastern Scotland, with the council areas of Na h-Eileanan Siar, the Shetland and Orkney Islands having virtually no access across the range of retailers in our dataset.

Our results suggest that remoteness plays a more important role than rurality in explaining prices and the availability of delivery. However, when remoteness, rurality and council areas are explored together in the model, rurality and remoteness of a postcode sector seem to have a smaller impact on the delivery charges than the wider council area. Finally, based on the predictions of our model, excessive charges are more prevalent in next-day, door-to-door delivery.

Defining “fairness” is a challenging undertaking. For this report, our team sought to provide an answer as to whether postal charges could be deemed unfair by using an econometric model to predict postal charges based on underlying geographic and parcel characteristics. The model allowed us to estimate the relative impact of each of these characteristics and the resulting postal charge. 

The findings from our model provide robust results that strengthened the evidence base on unfair delivery pricing in Scotland and will be used by the Scottish Government for the development of an interactive data hub and the Scottish Parcel Delivery Map. Yet, it should be noted that the differences between the actual postal charges and the postal charges predicted by the model cannot be used as conclusive evidence about whether a specific postal charge is unfair or not. This is because we have not controlled for key factors such as fuel and labour costs and also due to the fact that defining “fairness” encompasses a value judgement.

Future research can further explore the notion of consumer fairness in the context of deliveries in remote and rural areas in Scotland. Moreover, though the sample used for our research was informed by a market research exercise and was representative of a large share of the postal and retail markets in Scotland, a bigger data collection exercise in the future would increase the representativeness of the sample and further improve the accuracy of the results. Additional research into the large heterogeneity within specific areas, as well as investigation into the structure of the postal and retail market would also be critical for informing evidence-based policies to ensure that all Scottish residents have access to fair and transparent delivery charges.