Retail sales data are seasonally adjusted. This means that regular changes in sales at a given point in the year, in particular the large peak during the Christmas shopping season, are accounted for in the data. However, this does not account for effects such as unseasonal or extreme weather.
How sensitive to the weather is the retail sector?
In general, only sustained extreme weather conditions can have a substantial impact upon the UK economy as a whole. The only weather event in recent years to be designated as a statistical special event by ONS was the widespread heavy snow and extremely low temperatures in December 2010. (Statistical special events are identifiable events that do not occur on a regular cycle and have a clear impact upon statistical outputs; other recent examples include the Queen’s Diamond Jubilee, the Olympic and Paralympic Games, and the volcanic ash cloud in 2010).
However, retailers have regularly pointed out, both publically and in feedback to ONS, how their sales were impacted by not only severe weather events, but unusually high or low temperatures or rainfall.
Impact of weather on retail sales
The weather has a complicated effect on retail sales, affecting different goods at different times of year. For example, retail sales increased during warm weather in March 2012 and decreased during cold weather in March 2013, but this is often further complicated by other factors.
Figure 1 compares retail sales since 2003 with the average (mean) UK temperature relative to the average for that month. It also highlights months with severe weather events, including flooding in July 2007 and November 2012.
Figure 1: Retail sales and UK mean temperatures relative to 1971-2000 average, 2003-2013
The broader pattern is of a flatter picture of retail sales since 2008, with more frequent changes from month to month. This possibly reflects an increased variability in spending, which may reflect less secure consumer confidence during and after the 2008/9 economic downturn.
Notable weather events and retail sales
June-September 2006 saw mean temperatures at least 1.9°C above average. However, retail sales remained relatively stable throughout the period. This contrasts with a substantial increase in retail sales during July 2013, with temperatures 2.1°C above the average. In both periods there was a similar pattern of sales increasing in food stores and decreasing in non-food stores.
April 2007 saw mean temperatures 3.2°C above average for the time of year, the second largest of any month since 2003, but the spike in retail sales is actually from March 2007, with April 2007 sales relatively flat.
April 2011 saw mean temperatures 3.7°C higher than the average for the time of year, more than any other month since 2003. However, this month also contained the royal wedding and an associated bank holiday. Feedback from retailers in food stores suggested that both of these factors contributed to their increased sales, but retailers in non-food stores suggested that only the weather was a contributing factor.
March 2012 was very warm and dry for the time of year, and retail sales increased to a similar extent in April 2011.
June-July 2007 saw rainfall close to double of the long-term average for the time of year across most of the UK, and tens of thousands of homes and businesses were flooded towards the end of July. However, retail sales remained resilient.
November 2012 saw a prolonged period of heavy rain leading to widespread flooding, particularly in southern and western areas of the UK. However, retail sales increased in this month whilst decreasing in October and December.
January 2010 contained the largest change in retail sales in a single month, a decrease of 3.5%. However, although this coincides with very low temperatures and widespread snow, these were also present in December 2009 when temperatures were 2.1°C below average and retail sales fell by 0.3%. It is likely that the VAT increase in this month contributed significantly to the decrease.
December 2010 saw severe cold weather and was the only weather event recently designated as a statistical special event. Retail sales fell significantly, and there was a corresponding increase in January 2011 despite another VAT increase.
January 2013: although mean temperatures across the UK were close to average for January, heavy snow caused substantial disruption for much of the month. The fall in retail sales, though smaller than in January 2010 and December 2010, showed similar patterns between sectors, with stores selling automotive fuel most affected.
March 2013: this was the coldest March since 1962, and also saw significant snowfall. Retail sales fell, but unlike in previous snowy months, stores selling automotive fuel did not report a large decrease. Furthermore, sales fell again in April 2013 despite more moderate temperatures.
In terms then of the retail sectors sensitivity to the weather, past periods show a mixed picture with no clear relationship between the two.
Retail Sales reporting periods do not exactly correspond to calendar months, as they consist of either four or five weeks. All quoted temperature and rainfall data correspond with calendar months.
ONS does not produce regional data on retail sales, and hence cannot state if the impact of weather events increases in more severely affected areas.
These statistics were produced by the Retail Sales team in the Business Indicators and Balance of Payments Division. Further enquiries should be directed to Kate Davies (firstname.lastname@example.org) or David Howell (email@example.com). More information on retail sales statistics is available in the Retail Sales release.
Links of interest
Met Office pages on severe weather periods: