Europe seen from the stars: evaluating EU cohesion funding using satellite data
For public auditors, auditing policy performance relates by and large to the effectiveness of any given policy. It involves identifying and measuring the impact of policy implementation, which most often takes place through specific programmes and projects. In cohesion policy and other policy areas, many factors come into play, which makes it challenging to attribute effects to a certain policy action. Availability of data and new research techniques are crucial for performance auditing, and new venues — where they can be used — are particularly relevant. Satellite data are becoming increasingly available, and new insights on how to use remote sensing information are evolving quickly. Hannes Taubenböck is leading the ‘City & Society’ team at the German Remote Sensing Data Center (DFD) of the German Aerospace Center (DLR), and is also a lecturer at the University of Würzburg. Katharina Gnath is Senior Project Manager at the Bertelsmann Stiftung, and an expert on European and international economic governance. They explain below how geospatial data can be used as a proxy for economic and social changes not only at country level but also at local level to capture the local impact of place-based policies.
By Katharina Gnath, Bertelsmann Stiftung, and Hannes Taubenböck, German Remote Sensing Data Center (DFD)
Substantial resources allocated to reducing disparities
In December 2020, the Council of the European Union adopted the EU’s long-term budget for the years 2021 to 2027. With a share of 31% of the total EU budget (around €330 billion over a period of seven years), cohesion policy remains one of the EU’s most significant spending areas. In 2020, another €47 billion (via REACT-EU) was made available specifically for cohesion policy through the NextGenerationEu instrument to promote recovery from the pandemic and economic resilience.
The funding is intended to give the EU’s structurally weaker regions the opportunity to develop, for example through better infrastructure networks, and to catch up with economically stronger regions. Given the amount of resources dedicated to reducing economic and social disparities between European regions and countries, it is essential to learn more about the local impact of the EU’s funding instruments.
EU cohesion funding on the ground: easy to see with the naked eye, but hard to measure
Most Europeans have seen a building, road or bridge being built as part of EU-funded projects, and festooned with banners or plaques bearing the European flag. One such project is the 8-kilometre bypass in Myszków, a municipality in the Polish region of Silesia, that was built between March 2016 and September 2018. About 83% (or €17.4 million) of the construction costs came from the EU through the European Regional Development Fund (ERDF).
Prior to construction, it took a disproportionate amount of time to reach the nearest large city, Katowice, because there were no trunk roads. The new bypass improved the municipality’s links to the country’s long-distance transport network. The road also opened up a new industrial and commercial area, where regional and supra-regional companies with about 200 employees had already settled. Satellite images of the road under construction, as shown in Figure 1, highlight how the regional infrastructure has expanded over time.
Figure 1 — Aerial view of bypass and industrial area in Myszków, Poland
Ideally, EU funding of such infrastructure should lead to more economic activity in the respective city, municipality, or even the entire region. However, effects that are partially visible with the naked eye and known to local residents are difficult to measure and evaluate economically within a larger context. Previous literature has largely studied the growth effects of EU funding at the level of NUTS-2 or NUTS-31 regions, where it is hard to disentangle the impact of EU funds from other regional trends. There has been no consistent, EU-wide monitoring of regional funding at municipal level to date.
In a recent project, the Bertelsmann Stiftung, together with the German Aerospace Center (DLR), the ifo Institute in Munich, and the Austrian Institute for Economic Research (WIFO), developed a novel approach to evaluate the local economic effects of spending from two of the most important EU funds — the European Regional Development Fund and the Cohesion Fund — using satellite data as a proxy (see Box 1). The dataset for the two most recent EU funding periods (2007–2013 and 2014–2020) includes information on 119,116 EU-funded projects in 6,571 municipalities (so-called local administrative units, or LAUs). The process of geo-locating EU-funded projects allows a significantly more precise localisation and subsequent impact assessment of EU funding.
Satellite imagery as a basis for assessing EU funding
After local EU funding data were collected, they were subsequently supplemented with data on economic activity in the observed municipalities. To assess the impact of funding at such a granular level, the study leverages the potential of remote sensing data. The research is guided by the hypothesis that increased economic growth is accompanied by observable changes in spatial-structural parameters that satellite images can pick up on (as shown in Figure 1 above).
Satellite data record not only the nature of the Earth’s surface and changes over time, but also night-time light emissions. It is common in economic research to use nocturnal light emissions as a proxy for economic growth. Especially in less developed countries, this method is often used due to a lack of consistent economic data. Recognised studies have shown that night light development and (economic) growth are positively related. The purpose of this study was to determine the extent to which night-time light changes are related to EU investment at the spatial level of communities (see Figure 2 for Myszków, Poland).
Figure 2 — Increase in night-time light emissions over Myszków, Poland
Economic analysis shows a positive link between EU funding and local growth
The analysis revealed that EU funding amounts are positively related to the change in nocturnal light emissions in the municipalities under investigation. This effect takes on a value of up to 0.0116, depending on the period considered and the model specification.
This means that a 1% increase in EU funding, all other factors being equal, was associated on average with increased growth in night-time light emissions of up to 0.0116% in the municipalities considered. If one assumes the same correlation between night light growth and GDP growth at the municipal level as at the NUTS-3 level, this corresponds to higher growth of around 0.001%. EU funding policy thus shows a measurable, positive — albeit very low — correlation with economic growth at the municipal level in our pilot region.
The study also finds that there is a positive relationship between economic development in a municipality and EU funding in neighbouring municipalities. This means that EU funding not only benefits the directly-funded commune, but also those in the immediate surrounding area, and can thus trigger spillover effects.
Outlook: Earth observation and remote sensing data for policy evaluation
Satellite imagery is famous for providing a bird’s-eye view of processes on the Earth’s surface. A rapidly growing body of literature draws on satellite imagery to analyse economic questions. However, most of these studies tend to focus on comparisons of larger administrative units like countries or, in Europe, large NUTS-1 regions.
The study shows that remote sensing data can also be used to capture the local impact of place-based policies on economic development, even in a pan-European context. The Europe seen from the stars project provides an analysis of the relationship between EU project funding and local economic development, and is thus an example of how novel data at high spatial granularity can help in targeting important questions of regional policy. Covering a wider geographical area would make it possible to pinpoint the causal effect of EU cohesion policy on economic activity, for example by combining municipal data with eligibility thresholds in funding activity, or enabling a matching analysis that compares small-scale policy effects in similar regions.
Beyond this approach based on economic activities, remote sensing data and techniques can document and analyse land cover, land use, and atmospheric components and their changes over time with a consistent data base of high spatial granularity. Recent studies using remote sensing data, for example, have mapped and spatially analysed contiguous settlement areas and corridors of development in both national and transnational European settings. In another study, the driving emitters of air pollutants such as NO2 have been mapped, enabling the documentation of areas exposed to air pollution. Other studies combining remote sensing and other geospatial data reveal opportunities to bring heterogeneous data together in order better to understand the interdependencies of socio-economic or demographic indicators, and the shaping of space. An example of this is the interdependent analysis of the built and natural environment with quality of life. A multi-dimensional and systematic combination of these and other geo data for the documentation and assessment of EU cohesion policy, however, is still pending.
Data-driven analysis is a useful tool for underpinning debates on EU spending policies with greater empirical evidence, and can help the EU to develop a more tailored cohesion and structural policy. The work with different forms of data — specifically the innovative satellite-imaging technique — is a good starting point for further research into the granular effects of EU public spending.
Public auditors can use such data to corroborate other findings, or to highlight outliers in relation to other available data. The ever-increasing volume of new data available in a digital format may allow auditors to cover a larger population, or to zoom in on specific local situations. This is particularly helpful with regard to the effects of locally oriented programmes and projects that are financed through cohesion funding. The use of satellite data can be one of the tools when auditing cohesion policies and their impact, making “the view from the stars” more relevant for both local and EU policymakers.
(1) The nomenclature of territorial units for statistics (Nomenclature des Unités Territoriales Statistiques — NUTS) is a geographical system used by Eurostat, according to which EU territory is divided into three hierarchical levels, known as NUTS-1 (major social-economic regions), NUTS-2 (basic regions for the application of regional policies) and NUTS-3 (small regions for specific diagnoses).
This article was first published on the 1/2022 issue of the ECA Journal. The contents of the interviews and the articles are the sole responsibility of the interviewees and authors and do not necessarily reflect the opinion of the European Court of Auditors.