EU Economic Modelling System

EU-EMS is a dynamic spatial general equilibrium model. Originally developed within the EU Horizon 2020 Research and Innovation Programme by the PBL Netherlands Environmental Assessment Agency, the JRC contributes to the model development both conceptually and empirically to ensure credibility, salience and legitimacy when providing evidence-based scientific advice and support to EU policy. EU-EMS is used by the European Commission, European Investment Bank, European Parliament, European Union Agency for Fundamental Rights, European Institute of Innovation and Technology as well as EU Member States and European Neighbourhood Policy countries.

EU-EMS is useful both for ex-post policy evaluation and for ex-ante policy impact assessment and provides sector-, region- and time-specific results to provide model-based scientific advice and support to EU policy on resilience, preparedness, supply chain, inequality, migration, education and employment effects as well as structural reforms. All direct, indirect spatial and inter-temporal spillover effects of public investments or EU policies are captured and reported. EU-EMS is suited for assessing distributional impact by immigration status, firm heterogeneity in terms of technology and productivity (efficiency, capital deepening, human resources, green technology).

In order to ensure scientific credibility, a particular attention is being paid to a valid representation of the modelled policy area, the model is embedded in a conceptual framework and agreed by the scientific community as well as backed by the scientific literature, and the results are quantifiable, evidence based. Salience is achieved by adjusting the model to be relevant to information needs of decision makers, contributing to raising awareness and motivation to take policy action, results are understandable and the model output is readily understood by policy makers, results are temporally explicit – considers change over time, as well as scalable and transferable – applicable at different geographical and sectorial scales. To assure legitimacy, prior being used for the policy advise all key assumptions are selected through an inclusive process, a wide acceptance and agreement by the involved stakeholders, the model is fair in its treatment of stakeholders’ opposing views and divergent values, unbiased in its representation of preferences and interests, as well as transparent and clear assumptions are provided.

The dynamic spatial general equilibrium model features discrete time, multi-sector, dynamic discrete choice migration decisions, and investment determined as the solution to an intertemporal consumption-investment problem. The economy consists of many locations that differ in productivity, amenities, bilateral trade costs and bilateral migration costs. There are two types of infinitely-lived agents: workers and factor-owners. Workers are endowed with one unit of labour that is supplied inelasticity and are spatially mobile subject to migration costs. Workers do not have access to an investment technology and hence live “hand to mouth”. They make forward-looking migration decisions, taking into account migration costs and the expected continuation value from optimal future location decisions. Factor-owners are spatially immobile and own the capital stock in their location. They make a forward-looking decision over consumption and investment in this local stock of capital to maximise intertemporal utility. Capital is spatially immobile once installed and depreciates gradually at a constant rate. There are two sources of dynamics in the model: forward-looking investment decisions of the immobile factor (capital structures) and forward-looking migration decisions of the mobile factor (labour). To relax the assumption of perfect foresight, the model is linearised, allowing to derive a closed-form solution for the economy’s transition path in terms of impact conditions, which captures the impact of (policy) shocks in the period in which they occur, and transition conditions, which governs the evolution of the state variables from one period to the next in response to these shocks. The policy impact and transition paths depend on observable trade and migration shares. Agglomeration is captured through clustering forces in production (productivity spillovers) and residential decisions (amenity spillovers).

In the endogenous input-sourcing version of the model, production technology is vertically organised. Downstream producers sourcing from upstream suppliers along the supply chain produce a composite good in N production tiers that need to be performed sequentially. For the sake of tractability, in the baseline model we assume two tiers in the production process, though the model is more general and extends to any number of tiers straightforwardly. Firms in the upstream tier only use the local factor. Firms in the second tier combine the local factor with intermediate inputs produced in any location in the first tier. Both inputs – local factor and intermediate components – are required in the tier-two production, no output can be produced with one input only. If either labour or at least one source of intermediate inputs (or both) are not supplied, no production can take place in this location. The optimal input sourcing location at the different production tiers of the supply chain is determined by a cost minimisation of downstream firms. Final goods can be produced using different sets of intermediate inputs that can be sourced from any location. Firm-specific technique - a production function that specifies from which locations and suppliers intermediate inputs to source - can differ in terms of productivity. When choosing a production technique, a firm can adjust the importance of a supplier e.g. by including or dropping that supplier. Production techniques are chosen at the beginning of the period, firms have to produce with the selected input suppliers. These decisions, when aggregated, lead to changes in the supply network along both the intensive and extensive margins, which have implications to the supply chain's robustness. Firms can adjust the input sourcing path at the end of the period. The downstream firm’s problem in the vertically organised supply chain consists of choosing the least-cost path of production to manufacture a final composite good’s variety and deliver to consumers in their location.

EU-EMS requires detailed data across multiple dimensions –. Key data inputs include regional economic variables like GDP, employment, and capital stock by location; trade flows between regions, capturing goods, services, and labour mobility; production and consumption data for different industries; and demographic information such as population, migration patterns, and household preferences. Additionally, infrastructure and transportation costs, regional policies, and environmental factors (e.g., land use, emissions) are crucial. Time series data is needed to capture dynamic elements like investment, technological change, and policy impacts, ensuring the model accurately simulates spatial and temporal interactions.