The Pay gap and the tax system

Doug Bamford
Apr 11, 2018 · 4 min read

Recently published pay gap data also proves that a better tax system is possible

UK Employers with over 250 employees have recently announced their pay gap data. Employers were required by legislation to report the mean and median hourly pay difference between male and female employees. Most have complied, and data for each employer is available here.

The existence of a gap doesn’t prove anything — the question is whether the gap reveals discrimination or the genuine choices of male and female employees. Most employers pay men more than women, which indicates the likely occurrence of discrimination — or at least stereotyping. There have already been calls to publish the same information regarding disability, ethnicity and class.

The data gets us talking and thinking about the issue, which is a very good thing. Publishing the information may mean that employers take steps to improve their rating over time. Hopefully the pay gap was already on a downward trajectory and this will help the process. Can we learn anything else from the exercise?

I’m particularly interested in the process because it utilises people’s average hourly pay. I’m sure this has been a challenge for some employers, who employ staff on contracts that are not readily translatable into an hourly amount. Nevertheless, they have been forced to work this out as part of their calculations, and most have clearly managed to do so.

This is of interest to me because hourly income is the key component of my proposed lifetime hourly average taxation system. There are several advantages of lifetime hourly averaging. The main one is that it allows for a highly progressive tax and benefit system without the usual economic disincentives of such progressivity. Hour credits are the input into the calculation and people obtain primarily when they work for money (though there would be additional sources for those in education and caring for loved ones in need). This occurs because people need to work to obtain hour credits for their lifetime calculation and so there is a strong incentive to work full-time up until retirement age.

A further economic advantage to the system is that it avoids severe thresholds. Firstly, because it is a lifetime system and so there is no year-end. Second, because the tax rates are a smooth curve rather than a series of marginal rates. I’ve outlined the system and its advantages in my book Rethinking taxation, an academic paper, a book chapter and on my blog.

One major concern with hourly averaging is the practical one of getting reliable hourly earnings information. Employers will have to undertake an additional process which will be costly to them and which creates a new kind of fraud. The new fraud is claiming that employees have worked more hours than they have. The recent pay gap reporting is important to the hourly averaging proposal for two reasons.

The first is that it proves employers can provide the required information. They have managed it to comply with pay gap legislation, and this same information could be used for hourly averaging calculations. Furthermore, once companies are in the habit of undertaking the calculation it is much easier to repeat the process. The reporting process will be cheaper each time it is repeated.

Most of the data required for hourly averaging was already held on human resource systems and so it is just a case of making sure the information is captured for all employees. Employers now collect the data for pay gap purposes so there is little or no extra cost in providing the same data to the tax authority to enable better tax calculations. The data-cost problem, minor as it always was, has pretty much disappeared.

The second reason that the information is useful is because it can be used for benchmarking. Some types of work are harder to monitor, increasing the temptation to attempt hourly reporting fraud. Having existing hourly income data from a pre-averaging economy makes it much easier to respond to such fraud.

The first way the data is useful is that it becomes easy to notice if hourly incomes suddenly drop upon the imposition of the scheme. Some such changes might be genuine of course, as the economy could change substantially upon the inception of the scheme. However, these changes would give cause to investigate the possibility of fraudulent reporting.

Benchmarking within the system also allows pay standards for distinct types of jobs. Hourly averaging allows for generous hourly subsidies, which is one of the key advantages of the system. However, there are some jobs that people also like doing without pay — such as proselytising for their religion or being professional athletes. If a million-people become unpaid but hard-working “footballers” then this would be very expensive to society, both in the subsidy and in the lost work done in other professions.

One way to limit the scope for hourly reporting fraud is to set minimum hourly wage requirements for certain jobs. This can be applied both to jobs which display a sudden drop in hourly earnings and to jobs that are also popular leisure activities. Benchmarking makes this process easier, as it provides an initial basis for the job-specific minimum hourly wage requirements which can then be altered over time.

Of course, not all employers have submitted the information, but if the penalties for non-compliance were more onerous (including as a final threat losing the ability to trade) then compliance would rise. The practical issues of collecting the data required for an hourly tax scheme are falling away. One of the major barriers to hourly averaging has just disappeared and hourly averaging just got a lot easier.