Section 1
Why headcount shape decides your bill
Employer NI is charged per person, on their earnings above the threshold. Drop the threshold and you have not just raised the tax. You have pulled a whole band of workers who used to generate zero employer NI into generating some. Every part-timer earning between 5,000 and 9,100 pounds a year was, until April 2025, invisible to the employer NI system. Now each one carries a charge. Watch it on two businesses with an identical 300,000 pound wage bill. Restaurant A spread its wage bill over fifteen people, most earning well above both the old and new thresholds. The threshold cut moves their NI base by 4,100 pounds each, fifteen times. Restaurant B spread the same money over forty people, many of them part-time and clustered near the old threshold. For those workers the threshold drop is not a marginal adjustment. It is the difference between zero employer NI and a live charge. Same wage bill, same street, materially different tax, because the tax was aimed at head count, not at pounds paid. Take one worker to make it concrete. Under the old rules an employer paid NI on earnings above 9,100 pounds. On a 20,000 pound salary that is 13.8 percent of 10,900, or 1,504 pounds. Under the new rules it is 15 percent of 15,000, or 2,250 pounds, a 50 percent jump for the same person in the same job (The Access Group hospitality analysis). Now a part-timer on 6,000 pounds who used to cost zero in employer NI costs 150 pounds. Multiply the second effect across a rota full of part-timers and you have the reason two similar businesses feel a different April.
Section 2
See your own exposure in ten minutes
You do not need a model. You need three numbers. 1. Count the heads earning between 5,000 and 9,100 pounds a year. These are the people the threshold cut switched on. Before April they generated no employer NI. Now each generates 15 percent of everything they earn above 5,000. This is your pure exposure to the threshold change, and it is invisible in any "our costs went up X percent" summary. 2. Count the heads earning between 9,100 and, say, 15,000. These workers already generated some NI, but the threshold drop widened their base by up to 4,100 pounds each. Their charge rose on two fronts at once, the higher rate and the lower floor. 3. Divide your total employer NI bill by your head count. That is your NI per person. Compare it to the same figure for a business built on fewer, fuller roles. The wider the gap, the more the threshold cut, not the rate rise, is what is hurting you. If most of your exposure sits in groups one and two, your problem is not that wages are high. It is that your labour is sliced thin, and the government just made thin slicing expensive. That is a specific, fixable diagnosis, and it points at a specific response: the shape of the rota, not the size of the wage bill. The mechanics of reshaping it are their own piece. The point of this one is to stop you reaching for the wrong lever. One relief is worth naming before you panic. The Employment Allowance rose to 10,500 pounds in April 2025 and the old 100,000 pound eligibility cap was removed, so a large share of smaller employers offset a meaningful chunk of employer NI, and some pay none at all (The Access Group). Check whether you qualify before you model the damage, because for the smallest operators it changes the picture.
Section 3
The fitness test
You understand your exposure if you can say, in one sentence, whether the April 2025 change hit you mainly through the rate or mainly through the threshold, and back it with the count of staff earning between 5,000 and 9,100 pounds. If you can, you know which lever to pull and you know why the restaurant down the road got a different bill. You do not understand it yet if your only number is "costs went up." That number is real, but it is an average that hides the mechanism, and you cannot fix a mechanism you have averaged away. The operators who adapt to this are not the ones who cut hardest. They are the ones who read the shape of their own payroll and matched the response to the part the policy actually targeted.