It’s that time of the year again: the budget. Alongside the regular “Impact on households: distributional analysis” document (which I’ll blog about later), Table 2.1, “Budget policy decisions” is my favourite part of each budget event. It shows, in a fairly consistent format, the fiscal impacts of each of the main decisions taken in each budget; and I now routinely include it in my analysis of “materiality”.
As you can see from the chart above, I’ve been doing this since 2014, and by and large, the vast majority of budget decisions are very small relative to the size of the UK economy. [Note: I’m very aware that this all depends on exactly how HM Treasury has decided to present the measures, as they are sometimes listed as more granular individual items, and sometimes “bucketed” into broader items. But I work on the assumption that the level of granularity is not drastically different from one year to the next.] That is by no means to say the measures are not important, or that they don’t have important non-fiscal implications. However, I find this — and the summary and data in Table 2.1 — a really useful quick take of any specific budget event.
This year is no different. After a highly extraordinary Summer Statement in 2020, the nature of which was shaped by the pandemic, it seems the general pattern of budget announcements has returned to the historical norm. Out of the 52 measures (excluding financial transactions) in the 2021 budget, only 2 were large enough to show impacts in 2025-26 of more than 0.1% of (projected nominal) GDP. These were the change in corporation tax (up from 19% to 25% from April 2023), expected to add to £17 billion to government receipts in 2025–26; and the freezing of income tax personal allowance and rate thresholds (rather than increasing them each year), expected to raise £8 billion.
As always, it is quite difficult to keep one’s head straight on what the “counterfactual” is that is applied to these figures — i.e., what is the assumption of what would have happened to tax and spend policy (and government income and expenditure) if a particular measure had not been announced now. A classic example is fuel duty, which has historically been expected to go up, and a “cancellation” of such a policy or “freezing” of fuel duty, has been interpreted as a reduction in “expected” government revenues. [I tried to google, but must say lost track of whether there is any legal commitment by the current government to “stabilise” fuel prices or “escalate” the fuel duty year-on-year.]
Governments since 2011 have, however, all “cancelled” predicted or expected increases in fuel duty and kept fuel duty constant. As far as I can tell, this has always been scored as an additional expense for the government (and interpreted as some degree of reduced taxation for households). For example, in this latest budget, the freezing of fuel duty is shown as costing the government nearly £1 billion a year by 2025–26. However, if it has become traditional that the expected increases are always scrapped, then should the path of expected increases still be the counterfactual? Probably not.
The fuel duty example illustrates that the devil is always in the detail — and the chart shown in this blog is not meant to get into the detail in any way. The chart simply takes Table 2.1 as given and compares the size of the different measures (in the last year shown) to projected GDP (using OBR figures).