The Training Budget and Other Autumn Sacrifices
There is a particular kind of corporate meeting that takes place every September, in which someone presents the budget review and everyone looks at the training line with the expression of a person who already knows what is about to happen. The line does not get cut, exactly. It gets "deferred pending a review of Q4 performance." It gets "paused while we assess the current environment." It gets "held flat for now" while everything around it quietly inflates. The effect is identical to cutting it. The vocabulary is considerably kinder.
The current environment, in September 2025, has provided no shortage of material for this conversation. CPI inflation has been sitting at 3.8% for two months running, which sounds almost manageable until you remember that the Bank of England's target is 2%, that this is the rate after considerable effort to bring it down, and that it is still pushing up the cost of everything organisations need to function. RPI, the measure that tends to follow people around in salary negotiations and contract escalations, is running at 4.5%. The Chancellor's April increase to employer National Insurance contributions, which we wrote about with some foreboding last November, has by now been fully absorbed into payroll costs and is producing more or less the effect that seemed likely at the time: organisations that were already watching their numbers carefully are tighter than they were, and organisations that were already tight are having a genuinely difficult year.
One L&D professional, surveyed earlier this year, summarised it with admirable brevity: "Tough times in the industry, increased costs because of National Insurance hikes, a close eye on resources and costs means that any substantial new spend for training and L&D is on the back burner." This is, in our experience, representative of a fairly wide constituency at the moment, and not only among smaller businesses.
The numbers, which are not encouraging
UK employers are now spending an average of £1,700 per employee on training, down from £1,960 the year before, and the average employee receives 5.7 days of training annually, a figure which, when reported, prompted the observation that this was the lowest ever recorded in the survey's history. Annual training spend in the UK has fallen by around 28% in real terms since 2005, which means it has been declining with some consistency through most of the significant economic events of the past two decades, including several years in which it emphatically should not have been. We are now spending less than half the EU average, a comparison that tends to surface briefly in policy discussions and then disappear without troubling the September budget meeting unduly.
The picture across the economy is not uniformly bleak, which is worth acknowledging. The manufacturing sector has been in contraction for the better part of a year, but services, which accounts for the larger share of GDP, has been growing, and confidence there is at least relatively intact. This split matters because it means that the "difficult environment" cited in training budget discussions is not equally difficult for everyone. Some organisations genuinely cannot protect their development investment this year. Others could but have decided, perhaps not unreasonably, that this is not the moment. The tendency to treat "the economy is uncertain" as a universal instruction to pause all discretionary spending is, in practice, a way of outsourcing the decision to whoever wrote the headline.
The logic and its rather expensive limits
The case for cutting the training budget in a difficult year is not entirely without merit. It is visible, it is discretionary, and it delivers a clean number on a spreadsheet in a way that most cost reductions do not. Unlike salaries or infrastructure, it can be reduced without immediately breaking anything, which makes it attractive to anyone who needs to find savings before the end of a quarter. This is perfectly rational. It is also, as anyone who has had to rebuild capability after a multi-year freeze will confirm, a way of spending money that has not yet been invoiced.
The skills and confidence an organisation needs to respond when conditions improve take time to build, and that time does not compress neatly into the window between conditions improving and the competition noticing. Organisations that maintained development investment through previous difficult periods have generally come out the other side in better shape than those that did not, not because training is magical, but because capability gaps are easier to prevent than to close, and considerably cheaper. The saving achieved by deferring a training programme in the autumn rarely survives contact with the cost of remediation, recruitment, or lost bids twelve months later. We made essentially this argument last November in the context of the NI rise; the passage of time has not made it less true.
There is also a more specific concern for the autumn of 2025. The AI transformation agenda, which is consuming significant executive attention and, in many organisations, significant budget, runs directly into the training freeze in a way that ought to be uncomfortable. Upskilling teams to work effectively with AI tools does not happen by default or by an afternoon of watching tutorials. The organisations pausing their development investment this quarter while simultaneously planning AI transformation programmes are, at some level, planning to build a house without bothering to train the builders. The contradiction is not always noticed before it becomes expensive.
September will pass, as it always does, into Q4 with its own pressures and promises. The deferred training will get deferred again into next year's budget cycle, next year will arrive with its own reasons for caution, and the 28% real-terms decline will tick quietly upward. Or, and this is not impossible, the organisations that treated capability development as essential rather than discretionary will have built something their competitors have not. The data on this particular question has been fairly consistent for about two decades. It has not, it is fair to say, made a great deal of difference to the September budget meeting.