beta This part of GOV.UK is being rebuilt – find out what beta means

HMRC internal manual

National Insurance Manual

Aggregation of Earnings: The ‘not reasonably practicable’ test

There is no definition of the phrase “not reasonably practicable” in NICs legislation. We are reliant upon the ordinary meaning and case law, the latter arising from Health and Safety legislation. The following arguments also take into account an unreported determination by the Secretary of State for Social Security on whether the earnings of ‘Bank Nurses’ should be aggregated.

The onus is on the employer to show that aggregation is not reasonably practicable because it is the employer making the judgement. It is not a once and for all decision because the duty to aggregate is an ongoing duty. Also, factors may change - such as the composition of the labour force - sufficiently to affect the judgement. The employer will need to take into account the costs, resources, and the effects on running the business. Cost is a material pointer but not decisive. The context is important so the employer will also need to be aware of the effect on the National Insurance Fund (NIF) and the benefit or pension entitlement of the employee.

The reported judgements have to be filtered on the basis of their specific legislation but Mailer v Austin Rover Group [1989] (2 All ER 1087) agreed on three principles:

  • “reasonably practicable” is narrower than “physically possible”
  • risk has to be measured against the cost of removing it, and
  • account has to be taken of the likelihood of the risk arising.

Taking the principles in order:

  • It is always possible to aggregate but that is not the test. And the “separately calculated” provision is not failed merely because it is the same employer or payroll system.
  • Costs to the employer are not just financial. Time, effort and the effect on the business have to be considered because the weight of the cost of compliance should not be disproportionate to the loss of National Insurance Contributions and benefit entitlement.
  • Mailier and other cases are generally considering whether there is a duty to guard against unknown and unexpected events. It is the Department’s view that employers have a very limited argument on the 3rd bullet point because aggregation is a known and recurring event. However, that is not enough to negate any informed judgement by an employer that aggregation is not reasonably practicable.

The cases consider the balance between risk on one hand and the sacrifices necessary for averting the risk on the other. Basically, an employer needs to balance his employee’s interests against his own costs. It is very important that the consequences for the primary contributor are considered, especially the low paid, because of the potential loss of benefits and pension rights.

The employer can only make an informed decision if all the facts are established because ‘reasonably practicable’ is related to the individual circumstances. The evidence is that employers with computerised payrolls are the ones who find difficulty in complying with the need to aggregate. However, the existence of such a payroll is not enough evidence that aggregation is not practicable. Manual or other fixes will have to be considered and costed especially when there will be similar risks in future years if aggregation is not achieved and the employer will have to revisit the issue.

The Department takes account of the following points when comparing the costs of aggregation against the risks to contributors:

  • Is it a fact, rather than an assumption, that payroll software cannot aggregate earnings?
  • Is the payroll software an outside package, tailored package, provided by an internal IT section or able to be upgraded by internal resources? Has the payroll system been changed? If so, why was an aggregation requirement not part of the new specification?
  • Does the provider of an outside or tailored package give an update service that includes aggregation? Is it possible to upgrade or would the employer have to buy a new system? What are the costs of upgrading the software? Is there a dedicated internal IT team that might be able to provide it cheaply subject to competing claims for their services?
  • If the work has to be carried out manually what are the costs? Does the employer already have a manual support resource for payroll glitches, urgent payments and so on? Would new staff be required? What is entailed in staff years in manually calculating the NICs, which takes into account that experience will reduce the need for the initial, possibly untrained, resource?
  • How many employees are potentially affected? What is the total number of employees on the payroll? Do employees have similar pay periods? Is aggregation a continuing requirement or a one-off consideration because of a particular project or task? What are the amounts of NICs at stake and the effect on the NIF and primary contributors? How does this compare to the costs of compliance?
  • Has there been a material change in the labour force since the decision not to aggregate was taken? Or a material change in the state benefits that aggregation could bring (State Second Pension may be relevant here).