Will the Levy run dry?

17 November 2017

The Office for Budget Responsibility is going to significantly downgrade its forecasts for the potential of the economy to grow – its overall speed limit. This is because growth in productivity (the amount that each worker can produce) has consistently disappointed since the 2008 financial crisis.

The Bank of England gave a taste of what may come in its recent Inflation Report, where it said it now thinks productivity can only grow at around 1.4% per year, compared to the 2-2.25% since before the crash.

This has profound implications for living standards (productivity is one of the key drivers of wages) and public services (if the economy grows more slowly, then so will tax receipts, leaving less money for public services).

It also has implications for the Apprenticeship Levy, a 0.5% tax on employers’ payrolls above £3 million per year which is ringfenced for spending on apprenticeships. The 2016 Autumn Statement downgraded the initial estimates of the annual amount that would be raised by 2020 by £200m (from £3bn to £2.8bn). So will the more depressed outlook for our economy drive this down further?

To estimate that, we need to consider three factors:

  • How many people will be in work, and how many of these will be employed in Levy-paying companies? Employment growth has been consistently stronger than expected in recent years. So it would be surprising if the OBR downgraded its forecasts, it may even raise them. That said, if net migration falls towards the government’s target of a maximum of 100,000 per year, this could reduce levels (but not necessarily rates) of employment. Employment forecasts would need to change significantly to make a big difference to Levy forecasts. For this analysis, we assume the OBR will maintain its employment forecasts;
  • How much will earnings (and therefore payroll bills) grow, and how will this vary for Levy paying employers? The significant downgrades that will be made to productivity growth will surely mean downgrades to earnings forecasts, though of course productivity is far from the only driver of earnings. The question is by how much. For this analysis, we use a range with the most recent Bank of England forecasts as an upper limit and the upper end of average earnings growth seen in recent years as a lower limit; and
  • The introduction of policy such as the Levy may alter employer behaviour. For example, employers could avoid growth opportunities that would take their paybills over £3 million, they could outsource services to take their paybills below this level, or they could pay lower pay rises than planned to keep below the Levy paying threshold. The OBR will have made assumptions about the likely degree of behaviour change in its initial forecasts. With the Levy only six months old, it is likely too early to have evidence of how this is working in practice. For this analysis, we assume no behaviour change beyond that already previously assumed by the OBR and implicit in their Levy forecasts.

Taken together, this gives the scenarios for downward revision in earnings growth shown in Chart 1. The blue sections are the revisions to earnings growth forecasts made by the OBR in Autumn Statement 2016 (AS16) compared to Autumn Statement 2015 (AS15). The red sections are the additional revisions that would results from using the Bank of England’s most recent forecasts. The green sections are the additional revision on top of these that would result from using a possible ‘new norm’ of 2.5%.[1] These revisions are shown cumulatively for each year to 2020.

Chart 1: Scenarios for revisions in earnings growth


We then apply these different rates of earnings growth to the OBR’s original (Autumn Statement 2015) and subsequent (Autumn Statement 2016) projections for the amount the Levy will raise. This gives the potential range of downward revisions shown in Chart 2. This follows the same format and conventions as Chart 1.

Chart 2: Scenarios for revisions to Apprenticeship Levy raised


Ultimately, this could mean a further £100m downward revision to the total Levy funds raised in 2020, meaning a total downward revision of £300m from the forecasts made when the Levy was first announced. That’s equivalent to almost 10% less than originally anticipated. Using the Bank of England’s most recent earnings forecasts, which would represent a significant rise compared to experience since 2008, would mean a £50m reduction in 2020 compared to last year’s OBR forecast (that is, a total downward revision of £250m since the Levy’s announcement).

The data from the two charts are summarized in Table 1.

Table 1: Summary of scenarios


This brings risks to the government’s target of three million apprenticeship starts by 2020, particularly if there is a shift in the type of apprenticeships levy payers take up (for example, from less expensive to more expensive, such as management). It also raises questions of whether sufficient resource will be available for non-Levy payers.

Learning and Work Institute have raised these concerns before, calling for a guaranteed minimum apprenticeship budget over the economic cycle. This is to prevent funding for apprenticeships falling during economic slowdowns, arguably precisely the time when it’s most important to maintain national investment in training.

We also need to ensure that Levy money raised is spent. There has been a slow start with a dip in Apprenticeship numbers (though with a spike before the Levy came in and some expectation of teething troubles). Forthcoming Learning and Work Institute research suggests up to one in seven large employers don’t employ apprentices and have no plans to do so. Engaging this group, along with SMEs too, will be crucial in delivering the skills the economy needs.

The Budget will confirm to what extent the OBR expects lower earnings growth will follow lower productivity growth, and to what extent it believes this will affect the Apprenticeship Levy. Any changes may get lost in the roundings this time.

But even if the scenarios we have calculated don’t come to pass, at some point the economic cycle will turn, and with it employment and earnings growth will weaken. It’s essential we plan for that now to avoid the perverse scenario where we could be cutting investment in training when we need it most.

[1] Historically its been expected that, in steady state, earnings would grow broadly by productivity growth plus inflation. In practice this hasn’t been the case in recent years and it’s not clear what steady state is in the current environment.