Guest post by Eoin O’Leary, Emeritus Professor of Economics, Cork University Business School, UCC.
There has been vast media coverage on the effects of the COVID 19 pandemic on the Irish economy, most of which has been speculative as the struggles of workers, businesses and government has been ongoing. Yet the recent publication of the monthly Retail Sales Index (https://www.cso.ie/en/statistics/services/retailsalesindex/) and the Monthly Services Index (https://www.cso.ie/en/statistics/services/monthlyservicesindex/) for October provides a fascinating insight into how sales in customer-facing businesses in 21 retail and market services sectors have been affected by the pandemic since the beginning of 2020.
These sectors, which employed 1.1 million workers in the Republic of Ireland in 2019, have been directly affected by the sudden onset of the first lockdown of many shops and services enterprises in March, followed by graduated re-openings with social distancing measures and increasing reliance on remote access for customers.
The period January to October 2020 is ideal for investigating the 1st wave of the pandemic, as it covers the first lockdown in March and the subsequent re-opening during the summer before the 2nd lockdown on 19th October 2020. The sectors being investigated are major contributors to the Irish economy. Together they employed over 1.1 million workers in 2019, which represents 49% of total employment in the state.
The first key date in the timing of the response to the pandemic was the government-imposed lockdown starting on 27th March. Apart from retail enterprises involved in food, fuel and pharmacy (ie Food, Beverages and Tobacco in Supermarkets; Food, Beverages & Tobacco in Specialized Stores; Fuel and Pharmaceuticals, Medical & Cosmetic Articles) which remained open throughout 2020, all other retail sectors closed from this date with some providing remote access for customers. This was followed by a graduated re-opening (with social distancing measures continued) from June to mid-October. There was variation, with for example Bars (not selling food) being closed until September 21st, while mainstream non-food retailing (ie the remaining 8 retail sectors) re-opened from June 8th.
In market services, food and accommodation services (ie Accommodation and Food Services Activities) were closed from 27th March and re-opened on June 28th. There were differences within the remaining sectors. For Wholesale Trade those businesses involved in food, fuel and pharmaceutical distribution would not have been significantly affected during the 1st lockdown, while non-food retailers would have experienced a downturn. For Transport and Storage, passenger transport was severely curtailed while, anecdotally, there is evidence that courier services prospered. The other 4 sectors (ie Information and Communication; Professional Scientific and Technical Activities; Administrative and Support Services Activities and Other Business Services) were influenced by having to serve customers almost exclusively via remote access.
The Data and Method
Table 1 shows detailed descriptions of the sectors covered and their 2019 employment levels. For the Retail Sales Index, sales from a sample of 1,700 enterprises of varying sizes are collected each month across 13 sectors and indices generated. In addition to sales, in relevant sectors, data are collected on the % of sales from online sales of enterprises with a presence in Ireland. This source provides very detailed and comprehensive coverage of businesses of all sizes in retailing, accounting for 253 thousand persons engaged in 2019 (22% of the 1.1 million covered in the paper). One exclusion is that the index does not cover the retail trade of food and non-food via stalls and markets.
For the Monthly Services Index, sales from a sample of 2,250 enterprises with more than €20 million sales and more than 100 persons engaged is collected. In addition to the index only covering these larger enterprises, the 8 sectors are more broadly defined than retail (see Table 1). These sectors covered 876 thousand persons engaged or 78% of the total in 2019. Notable exclusions from this series are financial and insurance activities; public administration; education; human health; creative arts, music and entertainment activities. Together, both indices provide interesting insights into a substantial portion of the Irish economy that has been most affected by the pandemic.
In order to estimate the effect of the pandemic on each of these 21 sectors from January to October 2020, the actual monthly value index is compared to what the index would have been if no pandemic occurred. The latter is estimated using a simple model which first estimates the trend of the value index for each sector, calculated over the previous 5 years, for the 9 months starting in February 2020. This month is taken as the beginning of the pandemic because there is evidence, especially in food retailing, that customer behaviour shifted during that month when it became clear that a lockdown was imminent.
The trend is computed using the average annual percentage change in the seasonally adjusted value index from January 2015 to January 2020 which is applied to the February to October 2020 indices. It then adjusts the resultant series for each sector by monthly seasonal factors. These are derived by dividing the seasonally adjusted index by the unadjusted value index for February 2015 to October 2019 and calculating the average seasonal factor for each of the 9 months over these 5 years (see https://www.cso.ie/en/statistics/services/). The resultant seasonal factors are then applied to the estimated trend from February to October 2020.
The effect of the pandemic is assumed to be the difference between the estimated trend with seasonal variation and the actual unadjusted value index for each of the 9 months being analysed. It is assumed that the pandemic is the chief irregular component effecting the sales indices in these 21 sectors. There can be little doubt that it was by far the most significant abnormality that prevailed during 2020.
Overall, of the 21 sectors investigated, actual monthly sales are estimated to have been down in 16 sectors, with 5 either unchanged or doing better than they would have if there was no pandemic.
- By far the most negatively affected were Accommodation Services, which relates to hotels and other kinds of accommodation, down a massive 72%, and Bars down 58%.
- Other very large declines were in Transport and Storage (-37%); Books, Newspapers and Stationery (-32%); Clothing Footwear and Textiles (-30%); Fuel (-25%); Food Services, which includes restaurants and cafes (-24%) and Department Stores (-23%).
- 7 sectors were down between 3 and 18%, namely: Furniture and Lighting (-18%); Motor Trades (-12%); Administrative and Support Services (-11%); Professional, Scientific and Technical Services (-10%); Wholesale Trade (-8%); Other Retail (-7%) and Electrical Goods (-3%).
- 2 sectors were unaffected, namely Information and Communication and Pharmacies.
- Notably, 3 sectors experienced higher sales than they would have experienced without the pandemic. These were predictably Specialized Food Stores (+11%), and Supermarkets (+7%) and, perhaps less predictably, Hardware, Paints & Glass stores (+9%).
Figures 1–21 present detailed graphs of the actual value index for each sector from January to September 2020 and the estimated trend and seasonal variation, using the simple model described above. The base is January 2020 = 100 to facilitate comparison between sectors. The average monthly difference in percentage terms between the estimated and the actual is presented under the heading of each graph, to signify the magnitude of the effect of the pandemic. This magnitude is used to determine the order in which sectors are presented, beginning with those whose sales are down the most due to the pandemic and ending with 3 sectors whose sales improved in the pandemic. The commentaries beside each Figure bring out the highlights with suggestions of possible explanations. Where available the trends in the % of sales from online sales are included in the commentaries for retail sectors.
These findings provide fascinating insights on the effect of the COVID 19 pandemic on the sales performance of Irish customer-facing businesses in 21 sectors during the 1st lockdown and subsequent re-opening in 2020. Three points are worth making in conclusion:
- There was enormous variation in sectoral effects, ranging from -72% in Accommodation Services to +11% in Specialized Food Stores. This suggests that headlines heralding a crisis in all of retailing (see for example https://www.irishtimes.com/special-reports/business-ireland-magazine/is-irish-retail-in-crisis-1.3708364) are misplaced.
- The switch to online sales during the pandemic is understated in the CSO sources used here as no account is taken of consumer purchases from enterprises located abroad. There is evidence of considerable rises in these sales with over 50% of consumers increasingly their online shopping due to Covid-19 (https://www.independent.ie/business/irish/pandemic-sparks-increase-in-online-shopping-39694966.html)
- Due to an absence of data, this paper does not consider all customer-facing market services sectors. Examples of excluded activities are businesses involved in music and entertainment, which have been negatively affected (https://www.irishtimes.com/news/politics/musicians-and-artists-can-earn-up-to-960-for-gigs-and-keep-pandemic-payment-1.4432278) and outdoor markets, which may have prospered (https://www.irishtimes.com/news/ireland/irish-news/to-be-honest-i-ll-buy-anything-dundrum-farmers-market-reopens-to-eager-shoppers-1.4261114).
- The enormous variations evident here point to the huge complexity of the effects of the pandemic on Irish businesses providing livelihoods for very large numbers of owners and workers. It suggests that the government’s on-going response to this unprecedented public-health crisis should be informed by analysis such as this, as it comes under increasing pressure to withdraw supports and subsidies due to limits on its borrowing.
Contact email: firstname.lastname@example.org. The author would like to thank Eleanor Doyle and Owen O’Brien for discussions that inspired this work.