D. How should we think about the size of the adjustment we need to make?
One way of thinking about the size of the long-term policy changes that need to happen is by comparing the spending path that the "Resume Historic Cost Growth" scenario implies and the spending path that would be necessary to achieve net government debt at 20% of GDP as a long-term average, assuming we do not collect more tax. The selection of 20% net government debt as an average over time is not intended to represent a Treasury recommendation.[41] Rather, it is a benchmark level that is within the range of debt levels that past governments have considered "prudent".
Figure 6 below shows two "what if" spending paths:
- The blue "Spending path that maintains 20% net debt", which tracks the average spending path that would allow us to maintain net government debt at an average of 20% of GDP from 2020, assuming our tax take remains constant at 29% of GDP; and
- The orange "Spending path under 'Resume Historic Cost Growth' scenario", which tracks the average spending path that we would see if expense areas grow at the rates we have seen historically, also taking into account current legislative settings and demographic changes. This is the expense line implied by Table 1.
- Figure 6 Two government spending paths - an illustration of the gap we need to close
Both lines on Figure 6 track "primary" expenditure. That is, they do not include debt-financing costs. For the blue "Spending path that maintains 20% net debt", debt-financing costs would be fairly minimal. But for the orange "Spending path under 'Resume Historic Cost Growth' scenario", debt-financing costs would increase over time, assuming no revenue adjustments, eventually becoming quite significant.
The gap between the two lines illustrates the gap governments need to close. In 2060, there is an annual gap of around 5.5% of GDP (excluding debt-financing costs) between the spending path under the "Resume Historic Cost Growth" scenario and the spending path that maintains net government debt at 20% of GDP on average, assuming no tax increases.[42]
The task of governments is to move the two lines closer together. Broadly speaking, there are three ways we could do that:
- we could move the blue line upwards by collecting more tax than 29% of GDP, meaning that we could spend more and still achieve an average net government debt level of 20% of GDP
- we could move the orange line downwards by reducing growth in expenses, or
- we could do a bit of both.
What is happening at the beginning of the blue line?
- Snippet from Figure 6: Two government spending paths - an illustration of the gap we need to close
The blue "Spending path that maintains 20% net debt" line in Figure 6 has an unusual shape. From 2013 to 2020, it essentially reflects the actual spending path set out in the Government's 2013 Fiscal Strategy Report (before 2013 it reflects historic numbers). That is, between 2013 and 2020 it shows the spending path required to achieve the goal of 20% net government debt by 2020 (assuming no tax increases).
However, beyond 2020, the course the blue line takes is driven by our modelling assumptions. One result is that between 2021 and 2023, the line rises again quickly, before flattening out.
The reason for this quick rise is that producing the blue line, we assume that net government debt will be at 20% of GDP on average over time. This assumption means that once the 20% debt target is reached, as it is in 2020, our model needs to make spending rise again (to some extent) to maintain an average debt level of 20% of GDP. Hence the rise in spending from 2021. If the line rose more slowly, the average debt level over time would be lower than 20% of GDP.
So the unusual shape of the blue line is driven by the fact that it puts two things together: an actually planned spending path, and an assumption-driven spending path. The assumption-driven spending path provides a simple view of the widening gap between the spending path implied by the "Resume Historic Cost Growth" scenario and the average spending path necessary to maintain a steady level of government debt (not taking the impact of economic cycles into account).
How might debt-financing costs affect the size of the policy adjustment that must be made?
In some ways Figure 6 under-represents the size of the gap that needs to be closed. As mentioned earlier, neither of the two lines on Figure 6 includes debt-financing costs. But if governments in fact allowed spending to grow along the path indicated by the orange line, and made no compensating changes to increase tax revenue, our expenses would quickly outstrip our revenue, meaning that we would need to borrow increasing amounts. That means that as well as the expenses shown in the orange line above, we would also need to pay interest on the money we have borrowed.
The existence of debt-financing costs - despite the fact that Figure 6 does not show them - has implications for the timing of the adjustment we must make. If we make adjustments - either to spending or to revenue - immediately, significant debt-financing costs need never arise (debt-financing costs at some level will almost always exist, however, as even a prudent level of debt attracts interest).
However, delaying adjustment means that debt-financing costs start to grow and the eventual adjustment path becomes two-pronged. As discussed in the previous section, we will need to address both the deficit (which grows over time) and the debt that accumulated during the period over which we ran deficits. So we might say that Figure 6 represents the size of the gap we need to close reasonably accurately if we make adjustments immediately, but it under-represents the size of the adjustment if we delay.
Why is figure 6 half jagged and half smooth?
Up until 2012, Figure 6 shows our actual spending path between 1997 and 2012. Then to 2015, it shows forecasts (which are similar to actual numbers in that they take into account economic cycles). However, from 2016 onwards, this graph shows projections - what our spending path could be under certain assumptions. Actual data take into account cyclical fluctuations and government reactions to them. For example, governments tend to spend more on welfare benefits when the economy is going through a downturn.
Actual data also reflect ongoing policy decisions, whereas the "Resume Historic Cost Growth" projections do not. Some of the "jaggedness" on the left of Figure 6 represents government decisions to increase or decrease spending in certain areas.
Compounding these impacts, GDP fluctuates over economic cycles. Since government spending is represented here as a percentage of GDP, it appears to fluctuate more than if it were represented in dollars.
Our projections from 2016 are shown as smooth lines. There will be economic cycles in the future, but we cannot predict when they will occur or how big they will be, so our projections do not show them. Rather, they show two average spending paths given certain assumptions. In reality, any future spending path will look just as jagged as our past one.
Notes
- [41]While we have recommended that net government debt be reduced to 20% of GDP or below by 2020, we have not made any recommendations about prudent debt levels beyond that date.
- [42]5.5% of GDP is just the "primary" gap. If the gap is actually allowed to grow over time it would be far greater than 5.5% of GDP, owing to the effects of compounding debt-financing costs.
