Tackling child poverty delivery plan 2022-2026 - annex 4: cumulative impact assessment

3. Methodology

Our modelling results are based on comparing two scenarios in each year of analysis:

1. The counterfactual scenario, which estimates the child poverty rate in the absence of the policy package;

2. The policy scenario, which estimates the child poverty rate with the policy package included.

The policy scenario represents our best estimate of the child poverty rate on the current trajectory, starting with the official statistics in 2019/20. We then assess the contribution of the policy package by comparing this scenario to the counterfactual scenario in each year of analysis. Note that the two scenarios already diverged in 2019/20, because many of the policies in the package were already in place that year. The two scenarios would have only converged some years prior to 2019/20, before any of the policies in the package had been implemented.

While the counterfactual allows us to assess the impacts of the policy package in each year of analysis, it is impossible to fully isolate these impacts from external factors such as macroeconomic changes and UK Government policies. This is not just a methodological issue: in reality, most devolved benefits are linked to reserved benefits, and both are sensitive to wider changes in the economy. In addition, reserved and devolved policies interact in how they affect a household's poverty status in the same way that the policies within the package interact. Due to the interactions between the policies in the package, it is also impossible to disaggregate the impacts of the package into the impacts of individual policies.

The final section of this report explains these interdependencies in more detail, but the upshot is that changes in poverty over time cannot be disaggregated into the impacts of Scottish Government policies and the impact of other factors. Rather, when it comes to poverty rates, our methodology allows us to estimate strictly three pieces of information:

1. The impact of the policy package in each year of analysis, including any impacts on household incomes which work through the policy package;

2. The impacts in each year which are independent of the policy package; and

3. How these two sets of impacts, when combined, translate into actual changes in the number of children living in poverty.

In general, the nature of our methodology means that we can be more confident in assessing the impacts of specific policies (i.e. 1 and 2) than in projecting actual outcomes (i.e. 3), because most sources of error will be held constant when isolating the difference between two scenarios.^[4] However, all results are modelled estimates and as such are subject to a degree of uncertainty.