This chapter establishes an explicit link between foreign aid inflows and development indicators classified in the multidimensional setting of the SDGs. This linkage is not a black box as it takes advantage of the model’s causal chains describing budget allocations and indicator performance. First, we create counterfactuals by removing aid flows. Hence, we can estimate aid impacts and assess their statistical significance at the indicator or country levels during the first decade of the 21st century. Second, we produce a validation exercise comparing our results with econometric evidence found in a well-known sector-level study (access and sanitation of water) using a subset of our data.

This chapter introduces a model in which a government allocates financial resources across several policy issues (development dimensions), and a set of public servants (or agencies) that, through government programmes, transform public spending into policy outcomes. We start by describing the macro-level dynamics and the relevant equations involved. Then, we introduce a political economy game between the government and its officials (or public servants). First, we describe the public servants’ decision making in an environment of uncertainty through reinforcement learning. Second, we elaborate on the problem of the government (or central authority) and how we can specify its heuristic strategy. Finally, we provide an overview of the entire structure of the model.

This chapter elaborates on the calibration and validation procedures for the model. First, we describe our calibration strategy in which a customised optimisation algorithm makes use of a multi-objective function, preventing the loss of indicator-specific error information. Second, we externally validate our model by replicating two well-known statistical patterns: (1) the skewed distribution of budgetary changes and (2) the negative relationship between development and corruption. Third, we internally validate the model by showing that public servants who receive more positive spillovers tend to be less efficient. Fourth, we analyse the statistical behaviour of the model through different tests: validity of synthetic counterfactuals, parameter recovery, overfitting, and time equivalence. Finally, we make a brief reference to the literature on estimating SDG networks.

This chapter introduces the reader to the problem of policy prioritisation and why quantitative/computational analytic frameworks are much needed. We explain the various academic- and policy-oriented motivations for developing the Policy Priority Inference research programme. We apply this computational framework in the study of the SDGs and the feasibility of the 2030 Agenda of sustainable development.

This chapter formulates an analytical toolkit that incorporates an intricate – yet realistic – chain of causal mechanisms to explain the expenditure–development relationship. First, we explain several reasons why we take a complexity perspective for modelling the expenditure–development link and why we choose agent-based modelling as a suitable tool for assessing policy impacts in sustainable development. Second, we introduce the concept of social mechanisms and explain how we apply them to measure the impact of budgetary allocations when systemic effects are relevant. Third, we compare different concepts of causality and explain the advantages of an account that simulates counterfactual scenarios where policy interventions are absent.

This chapter provides a comprehensive framework to understand and quantify structural bottlenecks in a setting of multidimensional sustainable development. First, we formalise the idea of an idiosyncratic bottleneck when thinking in a hypothetical situation where a government has all the necessary resources to guarantee the success of its existing programmes (i.e., the budgetary frontier). Second, we compare the development gaps between the baseline and counterfactual outputs to assess how sensitive are the different indicators when they operate at the budgetary frontier. Third, we combine this information with the historical performance of indicators to develop a methodology that identifies idiosyncratic bottlenecks. Finally, we elaborate on a flagging system to differentiate between idiosyncratic bottlenecks according to the ‘urgency’ to unblock them.

This chapter studies the feasibility of the SDGs to improve our understanding of the empirical link between government expenditure and development outcomes. First, we explain the strategy to produce prospective (counterfactual or otherwise) analyses with the computational model and two metrics to evaluate advances in development gaps. Second, we present simulation results showing the development gaps by 2030 when the historical budget, in real terms, is preserved during the remaining years of the current decade. Third, we conduct sensitivity analyses that involve changes in the overall budget size that modify the value observed at the historical period used for calibration. Fourth, we present some reflections on the results.

This chapter analyses the connections between public funding, the rule of law, and multidimensional development. First, via simulation, we document a negative relationship between the budget size and the proportion of embezzled resources (or wasted resources due to inefficiencies). Second, our result suggests that reallocating public funds from other issues to programmes associated with the rule of law can mitigate corruption up to a certain point. Third, we find that the worse the country’s performance, the easier to remain in a development trap, as it becomes more cumbersome to realise a successful allocation profile (i.e., to decipher the proper mix of rule-of-law funding and overall budget size).

This chapter provides the reader with three reflections about the Policy Priority Inference research programme and its potential to make a difference in the real world. First, we synthesise the results found throughout the book and their implications for sustainable development. Second, we elaborate on systematic guidelines for deriving policies from the various analyses presented throughout the book. Third, we discuss the technical capabilities needed to adopt this toolkit and advocate for the training of computational social scientists.