Box 1

Examples of difficulties faced by central bank users of official statistics

CV19-related general increase in data uncertainty:

· Unobserved indicators: uncertainty created by mitigation techniques (e.g. for imputed price changes in the CPI)

· Changing economic patterns: e.g. in the consumption basket due to supply constraints and lockdown rules, resulting in bias in price deflators/real indicators

· Postponement of publication dates

· Large and more frequent data revisions

· Alteration to conventional data collection methods

· More difficult adjustment for seasonal patterns in time series

· Uncertainty related to the extent of inherent biases (e.g. time variation, impact of the changing economy and policy actions)

Specific issues related to sample surveying:

· Sample representativeness: for instance, sample survey frames are typically drawn from Statistical Business Registers (SBRs), which in turn often rely on taxation information which was distorted by the CV19-related recession and disrupted tax payments

· Outdated profiling information: for instance, enterprises listed in the South African SBR are generally deemed to be active for approximately 18 months after their last tax payment. Hence, those that had ceased operations due to CV19 impact could still be included in the sample frame for a significant period of time

· Changes in time series trends: risk of increasing bias and over-estimation

· Survey accuracy: reduced importance for respondents to comply to survey requirements because of other, CV19-related priorities

· Data sharing constraints: the supply of auxiliary information (e.g. financial statements) could be distorted because of resources issues/statistical delays resulting from CV19