Bringing Hope, Building Futures: Tackling child poverty delivery plan 2026-2031 – annex 5: Cumulative Impact Assessment

8. Calibration

The child poverty rates implied by our model can differ from the official poverty statistics, which are used to measure the child poverty targets. These differences can occur for a number of reasons, including the well-known problem of survey respondents underreporting their benefit income. We therefore adjust (or ‘calibrate’) the outputs of the model so that they are consistent with the official statistics.^[18]

Until now, our calibration method has involved comparing the statistics with the model outputs in a single year. The rationale for using a single year was that the most recent data was likely to provide the most reliable guide to the future in the context of significant economic and policy change, including the Covid-19 pandemic and the roll-out of the Scottish Child Payment (SCP). However, now that three years of post-pandemic data are available, with SCP in place for most of the period, we have modified our calibration method so that it is now based on three years of data. This should reduce the volatility of our projections going forward.

The overall calibration method has not changed; the only difference is that we are now comparing average child poverty rates over three years, in line with the pooled three year dataset used in the modelling (2021-22, 2022-23, and 2023-24).The differences in these calibration methods are demonstrated in Tables 12 and 13, which set out the results of both the single and multi-year calibration methods.

Notes: n/a indicates that year was not modelled.

Source: SG analysis using UKMOD

Notes: n/a indicates that year was not modelled.

Source: SG analysis using UKMOD

Neither the single-year calibration method nor the multi-year method shown here affect the estimated impacts of policies in percentage-point terms, because the baseline and counterfactual scenarios are adjusted by the same increments in each year, as shown in Table 14. This is by design, since we can generally be more confident when isolating the impacts of policies than when projecting outcomes such as child poverty rates.^[19] However, other calibration methods could produce different results, both in terms of the projections and in terms of the estimated impacts.

Source: SG analysis using UKMOD