Gross domestic product (GDP) measures the size and health of a country’s economy. Without it we would lack crucial information about the lockdowns’ effects, necessary for the formation of future policies. Volatility is used to describe unpredictable and quick changes. Coronavirus fits this description. Unprecedented measures were rolled out rapidly in the UK, with unpredictable economic consequences.

I look at a report by the Office for National Statistics (ONS, May 2020) identifying which conceptual and practical challenges the first lockdown posed for measuring GDP.


Practical challenges

The state of the economy is partially based on forecasts, but the Coronavirus was unprecedented. Therefore, there are more than usual inaccuracies in the  May 2020 estimates. Researchers made “informed judgements” (ONS, 2020) based on available information and proxy indicators. Forecasting methods will be continuously reviewed as new sources are established. For example, the new ‘Business Impact of Coronavirus Survey’, analyzed, instead of relying on past data, in the hope it correlates more to the truth. Due to less accuracy, more, and significant, revisions of GDP are likely. Hindsight will provide more clarity.

It was impossible to decipher which businesses involved in surveys stopped functioning during or after data collection. The subsequent errors must be accepted. The remaining survey questions were transferred online for greater accessibility, hopefully maximizing the continuation of information collection and reducing potential inaccuracies.

Furthermore, the ONS usually uses proxy answers to tackle survey non-respondents. These are based on past reports, or similar businesses’ answers. Nevertheless, with the drastic lockdown measures, this method was unreliable in establishing whether a business continued to function, or had ceased permanently or temporarily. Researchers looked, instead, specifically at first-time non-respondents and discovered if this was due to the virus, or for standard reasons. Online research helped understand whether or not trade continued, or if businesses had permanently or temporarily closed. Thus, they could conclude whether new non-respondents need to be further investigated.


Conceptual challenges

There are three ways of measuring GDP: production, income and expenditure. Usually, they are equivalent. I will look at how the pandemic’s effects appear in each approach.

The Production approach calculates the ‘value added’ to the economy; counting the value of a product, minus the good and services involved in the production process. For example, the flour that is used to make bread, would not count towards GDP.

The pre-existing issue of home production’s role in GDP were exacerbated. In ‘normal’ times it does not contribute towards GDP. However, with more time at home, a greater number of people were cooking, cleaning etc. themselves, when they may not have otherwise. This increase emphasizes the flaw in GDP measurement. Much of the production that was taken out of the economy during lockdowns, was replaced with work at home, which has the potential to add to output, but is unacknowledged in GDP measurement.

For the Expenditure approach, spending via consumption, government spending, investment and net exports (exports-imports), are put together to get GDP.

The government’s maintenance of many peoples’ income via various schemes, means that consumers still had some money to spend. Nevertheless, with fewer channels through which to spend money, there was an increase in both consumer and business savings. However, as government spending increased significantly, especially on health and benefits the question remains whether this increase undermined the decrease in other parts of the economy, influencing total expenditure?

The Income approach measures GDP by summing up income across all sectors. In lockdown there was less opportunity to earn an income, because of reduced labour demand. However, such impacts may not have appeared plainly in the data due to policies such as the Coronavirus Job Retention Scheme (CJRS). The CJRS was considered a production subsidy as its purpose was to support companies to return to pre-pandemic levels of production quickly. Therefore, money subsequently received by employees was considered a wage, was subject to all standard taxes. Furloughed workers’ income would still count towards GDP, yet their hours worked would be 0, with no contribution to output.

Nevertheless, as there was a cap on CJRS payment, furloughing could still reduce some wages, if the cap was lower relative to regular earnings, allowing for income changes to loosely reflect decreased productivity.

The rate of employment remained completely skewed, but this is not included in GDP.

While the effects of abnormal government spending may have presented themselves in income and expenditure, they did not add much to production. Labour hours were reduced, regardless. The subsidies were more an investment into an efficient return to pre-pandemic production in the future. Because of this, the three approaches seem unequal, which is an unusual outcome.

Non-market Output is the free services supplied by the government or non-profit organizations. This includes health and education in the UK. The pandemic particularly effected these areas, making their measurement necessary, but more challenging.

Health activity is usually calculated by weighting the estimate number of procedures by their cost. Expensive ones, such as surgeries, add more to total healthcare output, than cheaper procedures, such as consultations.

On one hand it seems healthcare output would increase with Coronavirus. Hospitalized patients received high-cost care and contact services saw an upsurge in demand. Nevertheless, to prevent the virus spreading and to provide more beds for those infected, other practices ceased altogether. This included dental and eye care, outpatient treatments and elective procedures which make up 6%, 13% and 19% of output in ordinary times.

This unbalanced effect on healthcare made estimates harder to form.

Furthermore, only the number of patients counts towards output, while healthcare facilities do not. The Nightingale hospitals did not guarantee more patients due to necessary social distancing. Therefore, their construction had a limited impact on output, despite requiring much productivity.

Education is measured by the number of fulltime students, multiplied by their education costs. Regional governments attempted to continue measuring school attendance, but this overstated the reduction as only key workers’ children attended.

Instead, the ONS could examine online schooling. However, firstly, while for older students this is an adequate substitute, for younger children school is a form of childcare which cannot be realized virtually. Secondly, although home schooling is excluded from GDP, deciphering between a teacher’s direct impact and learning support from parents is more difficult when education is done from home. Thirdly, students’ participation in online lessons is incalculable. There is no measurement of how engaged students are in ordinary times, either. However, the probable lower participation rate due to online schooling would have negatively affected education output, without appearing in the data, highlighting this shortcoming. Due to all this, analyzing just online class attendance would overstate education output.

Thus, the ONS adapted by measuring school attendance as normal (with the available data), as well as remote learning, but “discounted first by the change in average teacher input and then by the proportion of instruction in the home that is dependent on parental support” (ONS, 2020). Although imperfect due to insurmountable challenges, this method should obtain more accurate results.


Overall, what the ONS’s experience of the first lockdown tells us about research in volatile times is that there will be new, unchallenged issues. The time and effort it takes to identify what these are, their effects, thinking of and executing alternatives, means that inaccuracies must be accepted. The information should be continuously reassessed. Volatility can also emphasize already established flaws, such as the role of home production in GDP, bringing change of the traditional into consideration. Nevertheless, these obstacles were relatively small due to the vast amount of information and research tools already online. Notwithstanding the GDP data, itself, being volatile during the pandemic, the research methods to obtain it was not significantly derailed, and was adapted quickly.






“Coronavirus and the effects on UK GDP”, Office for National Statistics, May 2020


Print Friendly, PDF & Email