Here we show the calculations in detail to prove why the Treasury is working off the wrong figures.
The government’s initial claim that only 27% of farms will be affected by the new IHT policy materially underestimates the true proportion. We find that around 75% of commercial family farms will be above the £1 million threshold.
The analysis also finds that the majority of medium-sized working farms that will be hit by the liability will not be protected by the ten-year payment window because the resulting payments would still be unmanageably large relative to the economic returns they earn.
Put simply, the majority of farms don’t earn enough money to pay the potential Inheritance Tax Bill without selling off some of their land or business, which in turn makes the farm business unviable.
Report summary
- The NFU has consulted with former Treasury and Office for Budget Responsibility economists to inform our analysis of the impacts of the APR reforms on commercial family farms.
- The research finds that rather than the 27% suggested by Government, it is expected that 75% of commercial family farms will be above the £1m threshold.
- This is because once landowner claims on blocks of bare farmland and non-commercial farms are removed from the evidence base used by Government, historical claim values are adjusted to reflect current market conditions and the combined impact of claiming BPR alongside APR is considered, the proportion of farms impacted increases significantly to 75%.
- In addition, the research demonstrates that even if we consider an optimistic £2m threshold before the tax takes effect, for many medium-sized farms inheritance tax bills spread over 10 years would wipe out the majority of their returns, while for many large farms it would reduce returns by a half. At the £1m threshold, which will be the case for some, the tax bill would significantly exceed the average returns of a medium-sized farm and absorb most of a large farm’s earnings.
- Looking at sector impacts, the tax charge resulting from a £1m threshold would wipe out returns for an average cereals farm and around half of returns for average dairy farms. Considering typical historic returns on an average cereals farm and factoring in the reduction in direct payments, a farm making a profit of £34k will be hit with ten annual IHT instalments of £53k, over 1.5 times its profits. Even at a £2m threshold the annual tax payments of £33k would equal farm profits.
- This tax would in turn undermine investment and innovation in the sector at the time it is needed most. Farmers need to act now to mitigate the impacts of the announced reduction to direct payments as well as the inflationary impacts of the budget on key inputs such as labour costs and fertiliser. Instead, farm businesses are being forced to divert profits to prepare for significant future IHT bills while any value adding investment will further increase the tax burden. This policy undermines the ability for business for invest in their long-term resilience, competitiveness and their environmental performance.
- The analysis suggests that the majority of estates protected by the £1 million threshold are too small to be viable working farms, and are likely to include a high proportion of landowners letting blocks of bare farmland rather than farmers, whereas the majority of medium sized working farms that will be hit by the liability will not be protected by the ten-year payment window because the resulting payments would still be unmanageably large relative to the economic returns they earn.
See our report in full below. You can also download a PDF version of this report.
The NFU has consulted with former Treasury and Office for Budget Responsibility economists to inform this analysis of the impacts of the APR reforms on commercial family farms.
In the Autumn Budget, the Chancellor announced that full APR (Agricultural Property Relief) from inheritance tax would cease from April 2026. From that date, full relief would apply up to £1 million and 50% relief thereafter, for both APR and BPR (Business Property Relief) combined. The Budget states that the £1 million threshold is “to help protect family businesses and farms”.
The Government initially claimed that only 27% of farms will be affected, based on HMRC data on APR claims in 2021-22. The industry believes the proportion of genuine commercial family farms affected will be much higher.
This note reviews the available data and quantifies differences between the distributions of APR claims and farms by size to assess whether the different conclusions drawn from each source can be reconciled or whether they reflect genuine differences that mean the industry is correct to expect a larger proportion of farms to be affected. We also assess the ability of typical farms to meet inheritance tax liabilities over the ten-year payment window given evidence that profitability tends to be relatively low across the sector.
We show that the Government initial claims that only 27% of farms will be affected (as above the £1 million threshold) by the new IHT policy materially underestimates the true proportion. We find that around 75% of commercial family farms will be above the £1 million threshold. This is due to several factors:
- A fundamental difference between what constitutes a working farm and the number of APR claims. A significant proportion of smaller holdings that are shielded from the measure are not commercial family farms. In addition, APR claimants include landowner claims as well as owner occupier farmer claims. As around 80% of new FBTs (Farm Business Tenancies) consist of bare farmland averaging 54 acres1, this suggests many of the sub-£1 million claims may relate to landowner claims.
- APR data from 2021-22, which is where the 27% estimate derives from, is not a reliable representation of the prevailing conditions from April 2026 onwards. Land prices have grown rapidly since 2021, bringing more farms in scope of the measure, even if we assume that price growth subsequently slows.
- Around 40% of those claiming APR also claim BPR, so bringing claims under both reliefs within the same £1 million threshold is much more restrictive than implied purely from the APR data.
- The fact that farms may have multiple owners does explain some of the difference between the Government’s 27% and the higher proportion suggested by Defra’s data on the average net worth of farms, but this only goes a small way to explaining the gap.
We also show that the low levels of profitability in the sector mean that most affected farms will struggle to pay their tax bills, even over the ten years allowed for by the Government, reflecting the asset-rich, cash-poor nature of the farming business.
These results are based on reasonable and central methodologies and assumptions and are therefore robust to changes in the key underlying parameters.
The proportion of commercial family farms that are impacted
Comparing HMRC and Defra data sources
The Defra and HMRC data are not directly comparable, since the former show the net worth of a single farm while the latter show the value of the ‘agricultural’ component of a single estate that has claimed APR. These two concepts can be different in several ways, as explored in this note.
The mean net worth of a farm in 2022-23 was £2.2 million and the median was just below £1.5 million. By contrast, the mean value of an estate claiming APR in 2021-22 was £0.9 million (59% below the mean farm size) and the median was just under £0.5 million (68% smaller). The fact that the difference between the two metrics is greater at the median than the mean shows that tax claims are more heavily skewed to small land holdings. This is illustrated in Figure 1, which fits a smooth distribution through the available data on APR claims from HMRC and farm values from Defra.
The HMRC distribution of APR claims is materially above the Defra distribution of farm values for all values below £1 million and materially below it for all values above £1 million.
It is these distributions that determine the proportion of estates and the proportion of farms that are expected to become liable to inheritance tax as result of the Autumn Budget measure.
Figure 1: Fitted distributions of APR reliefs claimed and the net worth of farms
Source: Analysis using data and information from HMRC, Defra, Advani et al (2024) and the NFU.
It is important to note that the net worth of farms includes both assets that are eligible for APR (land particularly, as well as farm buildings) and assets that are not (such as machinery and livestock, which are eligible for BPR).
On average, around three-quarters of farm assets are made up of land. As the next section shows, this combination of APR and BPR assets is one key reason why using only APR data to estimate the proportion of farms that will be affected by the Budget measure paints a misleading picture.
Relative to the key threshold of £1 million, the farm net worth data point to 66% of farms being larger than that, whereas the APR claims data suggest that only 27% are larger (Table 1). For the Government’s initial claim that only 27% of farms will be affected by the measure to be true, it would need to be possible to reconcile these two figures. If it is not possible, and there is a genuine difference between the two sources, that would mean that using APR claims data to understand the proportion of farms affected by the new APR threshold will give misleading results. The evidence set out below points clearly to the latter: the APR data do give a misleadingly low figure.
Table 1: Distribution of APR reliefs claimed and net worth of farms
Source: HMRC, Defra
Why the two data sources differ and the implications for impacted working farms
The Government’s initial claim that only 27% of farms will be affected by the measure rests on a fundamental misunderstanding of the difference between a working farm and an APR claim.
APR can be claimed by landowners who have owned the land for seven years and rent their land to a farmer or to land owned and used by farmers for at least two years. Given that around 80% of new lettings (Farm Business Tenancies or FBTs) are for bare farmland in blocks averaging 54 acres, this indicates that many of the claims below £1 million are for let land rather than farms. In fact, it is possible to make a claim on a single field that has been let out.
A simple way of illustrating this difference is to convert the average APR claim in different size buckets into acreage based on an average land value in 2021 of around £8,000 per acre.
As shown in Table 2, the 27% of claims on assets of less than £250,000 would have been only 12 acres in size if the claim related only to bare land. This is clearly too small to be a ‘working farm’ in the conventional sense (See Annex on consideration given to horticulture, specialist pigs and poultry sectors). Factoring in land price appreciation since then takes this figure down to just over 10 acres.
Even claims for assets for the next group up, those between £250,000 and £500,000, would have been less than 50 acres in size on average. This would include very small working farms but also lots of landowner lettings under FBTs. Price inflation since then lowers the figure to 40 acres.
Table 2: Implied acreage of APR claims based on average land prices in 2021
Source: Analysis of data from HMRC, Knight-Frank, Savills and Strutt and Parker.
Quantifying the differences and the proportion of working farms likely to be affected
Having established that the APR data are misleading as a guide to the proportion of farms that will be affected by the new £1 million threshold, just how misleading is the 27% figure likely to be?
Our initial assessment quantifies three main reasons why the Government’s figure is misleadingly low. Figure 2 shows how factoring in each of these raises the proportion of farms that are affected.
Against that, we also consider the main factor that works in the other direction.
The three factors raising the proportion of farms affected relative to the Government figure are2:
- Not all APR claims relate to working farms. This is clear from the implied acreages in Table 2. To quantify the potential impact of this, we use Defra statistics. Close to half of the number of UK holdings are 20 hectares or less (50 acres or less). In most instances this is an impractical size for a viable farming operation. One exception is horticulture (See Annex on consideration given to specialist pigs and poultry sectors), where farms can be small but relatively valuable. Horticulture covers 2% of utilised agricultural land but with an average net worth of £1.1 million. Farms of less than 20 hectares, excluding those in horticulture, provide a proxy for non-working farms. These account for a combined 4% of the total land area of UK holdings, but 44% of all holdings. Removing them from the distribution increases the estimated proportion of affected working farms from 27% to 49%.
- The published APR data are from 2021-22, which means that by the time the policy takes effect from April 2026, higher prices will have brought many more farms into scope. Furthermore, it is not clear that this single year of claims is representative of claims over time. Agricultural land prices, the largest component of a farm’s value, have risen annually by double digit rates since 2021. We estimate that even if prices subsequently return to the average growth rate seen over the past decade, the cumulative increase since 2021 will bring around 25,000 holdings into scope for the measure, raising the proportion of working farms impacted to 70%. With the threshold due to remain fixed at £1 million to at least 2029-30, the impacted proportion would rise further to 72%, even if we allow for land price growth to slow as a result of this measure. Indeed, if the threshold remains flat ten years after its introduction – quite plausible since the main nil rate band has not been increased in fifteen years and counting – then the affected proportion would rise further still, to an estimated 76% by 2036-37.
- Relying purely on past APR claims does not capture the impact from the changes to BPR, which many working farms will currently claim alongside APR. BPR is available for those estates claiming IHT relief on assets that are not eligible for APR. These include agricultural machinery and certain farming buildings, as well as non-farm business assets. BPR is claimed by around 40% of APR claimants, so bringing those claims within the same £1 million tax-free allowance will bring even more farms within scope of the measure3 4. Factoring in the additional impact from the changes to BPR lifts the proportion of affected working farms to 75%.
There is a degree of uncertainty to each of these estimates, but we nonetheless believe that they are central, with upside as well as downside risks. For example, taking each component in turn:
- Viable holdings below 20 hectares will also include some farms with specialist pigs or specialist poultry, but since the data was not readily available, we did not make an adjustment. However, there are likely to be offsetting factors, such as the high likelihood that some farms over 20 hectares are not genuine working farms.
- Similarly, it may be that the high rate of recent land price inflation is temporary and that prices will slow going forward. We have allowed for this by basing our land price forecasts on long- rather than short-term levels. Some argue that bringing in these IHT restrictions will slow or even reverse the growth rate, but others take the view that those impacted by this measure are not those that contribute most to land price inflation.
- Finally, while we know the number of agricultural BPR claims we don’t know their average value. But we do know the ratio of land to non-land assets and this suggests an average claim that is close to the median BPR value is not an unreasonable assumption – this is supported by the data subsequently released alongside the Government’s response to the Treasury Select Committee. Furthermore, we have not adjusted for any interaction between these factors. For example, the BPR impact will bite more if we first account for fiscal drag.
One factor that is likely to reduce the estimated proportion of impacted farms, when comparing the two datasets, is the average number of owners.
If a farm has multiple owners, then its net worth will exceed the value of each agricultural estate of its owners (so long as they do not own stakes in other farms).
Conversely, for someone that owns several farms, the value of their agricultural estate will exceed the net worth of any one of their farms in isolation. It is not possible to estimate the average number of owners precisely from publicly available data, but for the Government’s estimate to be correct we believe it would require an implausibly high number.
While we do not have data on the average number of owners, we are able to use suitable proxies to draw reasonable and well-informed conclusions. In 2021/22, there were around 102,000 farm holdings in England that made a combined 84,000 claims under the Government’s Basic Payment Scheme, a close to 1-to-1 ratio.
Similarly, data on the average employed labour units per farm shows that for small and medium-sized farms, annual ‘farmer and spouse’ labour units are either very close to one or only slightly above5.
Farmer and spouse arrangements typically lead to the extensive use of IHT’s spousal exemption whereby the farm effectively remains under continuous single-person ownership. A section of farms are owned within a partnership structure, but these typically fall within the farmer and spouse arrangement.
We would expect the 50% of farms that are owner-occupied to largely follow this ‘family ownership’ model too.
There will be other partnership structures, and multi-ownership might be more common in larger farms. Large farms employ an average five labour units each, which could conceivably be indicative of some multi-ownership, but is more likely to tell us that large farms typically have a small number of employees.
For the Government’s initial assertion that only 27% of farms will be impacted by the measure to be true, requires 2.8 owners per farm. We are not aware of any evidence that supports such a high number.
While we have chosen central estimates for each of the three components in Figure 2, this result is robust to taking a more conservative view. For example, halving the impact of the three factors, would still require an average of close to 2 owners per farm, which is not supported by the data.
Figure 2: Assessing the Treasury’s underestimate of the proportion of working farms valued at more than £1 million
Source: Analysis using data and information from HMRC, Defra, Savills, Knight-Frank, ONS and Advani et al (2024).
Farms’ profitability and their ability to pay the IHT bill
Even with the Government’s proposed ten-year payment window for IHT, the tax burden could place significant financial strain on farms. An IHT bill of 20% of the value of assets over £1 million gives an effective tax rate on assets that rises from 10% for a £2 million farm to 16% at £5 million and 18% at £10 million. Spread over ten years, this gives a tax bill equivalent to a charge of 1%, 1.6% or 1.8% of assets per year, respectively.
These figures can be compared to Defra’s balance sheet analysis of farms in England to show the potential impact of such calls on farms’ income.
Defra’s statistics calculate a ‘return on capital employed (ROCE)’. The denominator – ‘capital employed’ – will differ from the value of agricultural assets subject to the new £1 million threshold for combined APR and BPR relief (for example, in its treatment of liabilities), but it is likely to be a sufficiently reasonable proxy for these purposes.
The ‘return’ is a statistical measure that aims to capture economic returns. Relative to what farms would see in terms of net income, it adds back tax and interest expenses (raising the ‘return’ relative to net income) but it subtracts an imputed wage for unpaid labour (lowering the ‘return’ relative to net income). ROCE estimates are volatile from year to year but have been higher over the past five years (2018-19 to 2022-23) than the preceding five (2013-14 to 2017-18), so we use the most recent five-year average in the following analysis.
On this basis, the median ROCE for small farms was negative at -0.6%, for medium-sized farms it was positive, but only just, at +0.4%, while for large farms it was +2.0%. By farm type, the median for cereals farms was +1.5%, for general cropping it was also +1.5%, for dairy it was +1.8% and for specialist pigs and poultry farms it was +0.3%. Across all farms, the median was just +0.2%, reflecting negative returns for livestock farms (-1.4% on lowland and -0.9% on less-favoured land). This would suggest that for most medium-sized farms, inheritance tax bills spread over ten years would wipe out their entire returns, while for large farms it would materially reduce those returns.
Figure 3 illustrates the interaction between farm value, profitability and the impact of paying off an inheritance tax liability over ten years. Most strikingly, at a 0.5% return on assets (just above the average for medium-sized farms over the past five years), an estate would only need to be valued at £1.33 million for all economic returns to be wiped out and just £1.14 million for returns to be halved. At a 1% return on assets, those cut-off points rise to £2 million and £1.33 million respectively, while at a 2% return on assets (the average for large farms over the past five years), returns are halved for farm estates valued at more than £2 million.
The distribution of farms by net worth shown in Figure 1 suggests that around 60% are larger than £1.33 million (the point at which median returns for a medium-sized farm are wiped out), while around 40% are larger than £2 million (the point at which median returns for large farms are halved).
Combined with the previous analysis of what is likely to constitute a working farm, this suggests that the majority of estates protected by the £1 million threshold are too small to be viable commercial family farms, whereas the majority of medium-sized working farms that will be hit by the liability will not be protected by the ten-year payment window because the resulting payments would still be unmanageably large relative to the economic returns they earn.
This suggests that there is a much more significant risk of farms being broken up to pay IHT bills than has been apparent in statements from the Chancellor and others since the measure was announced6.
Figure 3: Low profitability makes it challenging for farms to pay their IHT bill over 10 years
Source: Analysis using data and information from HMRC, Defra, Savills, Knight-Frank, ONS and Advani et al (2024).
In practice, taxpayers will also make use of other available reliefs, notably the £325,000 nil rate band and the ‘spousal exemption’, which allows for the transfer of assets between spouses or civil partners without incurring an IHT charge. Therefore, it is also useful to consider the effect of higher effective thresholds from the combined use of these reliefs on a given farm estate.
This is illustrated in Figure 4, which uses the same assumptions about returns but with an effective threshold of £2 million. It shows that, on this basis, low profitability still makes it challenging to pay the IHT liability. At a 0.5% return on assets, an estate would see all economic returns wiped out for 10 years at a value of £2.67 million, and half would be wiped out at £2.29 million.
At a 1% return on assets, these cut-off points would be £4 million and £2.67 million, respectively, while at a 2% return on assets, returns are halved for farm estates valued at more than £4 million.
Figure 4: Low profitability is still challenging if the first £2 million of a farm estate is tax-free
Source: Analysis using data and information from HMRC, Defra, Savills, Knight-Frank, ONS and Advani et al (2024).
The behavioural response to the measure
The changes to APR and BPR are estimated to raise £520 million a year by 2029-30.
The split between revenue raised from agricultural estates and other business assets has not been published, but if it were to match the roughly one-third APR, two-thirds BPR split of the combined tax cost of the two reliefs in 2021-22, around £180 million would come from agricultural estates.
This figure reflects the Government’s OBR-approved estimate that 36% of the amount that would be raised if those affected did nothing in response is lost to behavioural changes. The OBR’s documentation states that the figures allow for “a significant increase in the use of spousal exemption by married estates” (a response that largely defers rather than reduces the liability, since it would still be due when the second partner dies) and “an attrition assumption” for other estates (an OBR term that means a percentage reduction in yield to allow for a variety of potential responses without specifying and quantifying each of them individually).
The behavioural offset is on a rising path, starting at 21% in 2026-27 and rising steadily to reach 36% in 2029-30. That reflects the time it will take for those affected to undertake whatever planning of their tax affairs is possible. It would not, for example, be possible to utilise the gift route – at least not in full – for any estates where the individual dies within the five-year forecast. Over the longer term, gifting farm assets earlier in life may be a more widespread response.
The OBR document does not discuss the behavioural responses of those farms that have to pay the new tax liability. These might be modest at the aggregate level, from the perspective of the OBR’s fiscal forecasts, but this can mask some significant differences between lower- and higher-valued estates.
Our analysis shows that the impact could be very significant for the farms affected. For those hit in the near term, there may be little option other than to sell part, or all, of a farm to meet the tax liability and the sale of part of a farm could impact the potential future viability of the rest of the business. This result will apply more acutely to smaller and less profitable farms, a group that probably has less scope and is less well-prepared to quickly activate tax planning strategies, compared to much larger estates.
While the behavioural response assumed in the costing lowers the expected yield by around 40% overall, a disproportionate amount of that comes from those with more valuable estates. There is proportionately less ‘attrition’ of the yield from owners of lower value farms.
In the longer term, it might be possible for some of those affected to take out insurance, but the cost of that insurance would largely bring forward and (for those young enough) further spread out the cost of the tax liability.
Concluding remarks
This note shows that the Government’s initial claim that only 27% of farms will be affected by the new IHT policy materially underestimates the true proportion.
We also show that low profitability in the agricultural sector will make it challenging for many of those impacted to pay their IHT bill, even over the ten years allowed for by the Government.
While there is a degree of uncertainty given the lack of available granular data, we have made central estimates, with both upside and downside risks, and therefore believe our key findings are robust.
Annex A: Number of affected farms
In this Annex, we set out the detail underpinning the analysis in Figure 2 – the ‘waterfall’ chart. Each of the three quantified components is estimated using publicly available data.
This is less detailed than the granular data that underlies the Defra and HMRC statistics, which means there is a greater degree of uncertainty around the precise estimates compared to generating them using the disaggregated data.
Nonetheless, the main result – that a greater number of commercial family farms are impacted than implied by the Government’s 27% estimate using APR data from 2021/22 – is robust to variations in the underlying parameters and assumptions.
The number of commercial family farms. This element relies on Defra’s published statistics on the number of holdings by size and the number of holdings by type7. We assume that farms below 20 hectares are unlikely to be commercially viable and therefore remove them entirely from what we classify as ‘commercial family farms’. The one exception is horticulture, which is added back to the total. There are other farming types below 20 hectares that can be commercially viable, such as specialist pigs or specialist poultry. Arguably, these could be added back too, but the data was not readily available to do so. This is a downside risk to our estimate, though its impact would be small. We have also not made an adjustment for the likelihood that some farms over 20 hectares are not commercial family farms. This is an upside risk, and arguably larger than the downside risk. Nevertheless, in the absence of compelling evidence either way, we take the view that these two effects broadly offset and are content that our estimate is central.
Accounting for asset value inflation and its impact on fiscal drag. This element uses three different published sources of data on land value appreciation: from Knight Frank, Savills and Strutt and Parker8. We assume that land value growth is a proxy for asset value growth, on the basis that the majority of the latter is made up of the former. The three sources show rapid cumulative price growth between 2021 and 2024 and we use the average annual growth rate across the three sources to grow asset values across the period. This inflation has already occurred, so even if prices subsequently decline, whether in response to this measure or otherwise, values have risen since the 2021-22 data on APR reliefs. From 2024 to 2026, we assume that asset values grow at the ten-year average. Beyond 2026 we assume a growth rate that is lower than the ten-year average (but still positive). We think this balances the competing views that the current growth rate is unsustainable (or that values might decline as tax advantages are reduced) with others that take the view that those estates impacted by this measure are not necessarily those that contribute most to land price inflation.
We apply this asset value growth to the mean net worth of farms (which we split into price buckets) to estimate how many additional farms will be impacted due to fiscal drag9. Once again, while there is significant uncertainty around the precise estimate, the supporting assumptions are deemed to be central, and the key result is robust.
Allowing for the impact of the changes to BPR. The key judgement is the assumed value of a BPR claim from an affected agricultural estate, which we take to be the median value of all BPR claims. This is broadly saying that the ratio of the median BPR claim to the median APR claim is similar to the ratio of land values to non-land values in an agricultural estate. This is verified by the information released in the Government’s response to the Treasury Select Committee that shows that our assumed 40% ratio (of the median value of agricultural sector BPR claims to APR claims) is within the plausible range. By contrast, the assumption that 40% of APR claimants also claim BPR comes from the findings of Advani et al and is based on HMRC administrative data covering several years, so is therefore relatively robust.
The BPR restriction is effectively a lowering of the £1 million threshold and that is how it is modelled. The number of additional farms impacted as a result again relies on Defra net worth statistics.
Each component’s estimate uses the best available public data that we are aware of and assumptions that we believe are reasonable and central. Therefore, while the precise estimates are subject to considerable uncertainty, the direction and broad magnitude of the three main results are robust.
The proportion of impacted commercial family farms is higher than the Government’s initial claim that 27% of farms would be affected, fiscal drag will lead to more farms impacted by the time the measure is in effect and over time, and the concurrent BPR changes make the measure more restrictive than is implied by looking only at APR claims. Interactions between these factors would be likely to push the number of farms affected somewhat higher still.