The economic potential of the Third Strand of the Investment Plan for Europe: a general equilibrium assessment

Il Mulino - Rivisteweb Martin Christensen, Andrea Conte, Filippo Di Pietro, Patrizio Lecca, Giovanni Mandras The Economic Potential of the Third Strand of the Investment Plan for Europe: A General Equilibrium Assessment (doi: 10.14650/103217) Scienze Regionali (ISSN 1720-3929) Fascicolo Speciale, gennaio-aprile 2022 Ente di afferenza: () Copyright c by Societ`a editrice il Mulino, Bologna. Tutti i diritti sono riservati. Per altre informazioni si veda https://www.rivisteweb.it Licenza d’uso L’articolo `e messo a disposizione dell’utente in licenza per uso esclusivamente privato e personale, senza scopo di lucro e senza fini direttamente o indirettamente commerciali. Salvo quanto espressamente previsto dalla licenza d’uso Rivisteweb, `e fatto divieto di riprodurre, trasmettere, distribuire o altrimenti utilizzare l’articolo, per qualsiasi scopo o fine. Tutti i diritti sono riservati. The Economic Potential of the Third Strand of the Investment Plan for Europe: A General Equilibrium Assessment Martin Christensen, Andrea Conte, Filippo Di Pietro, Patrizio Lecca, Giovanni Mandras Scienze Regionali, vol. 21, numero speciale 2022, pp. 127-158 ISSN 1720-3929 © Società editrice il Mulino Received: 1/10/2020; Accepted: 20/09/2021 Abstract: This paper uses a general equilibrium framework to evaluate the macroeconomic impacts of the legislative measures contained in the third strand of the Investment Plan for Europe. Differently from most of the existing European policies evaluated with economic impact assessments, these measures do not involve any funds or money injections in the economy. Rather, they aim at either lowering or removing the remaining obstacles to a fully functioning European single market. We find that the removal of those barriers may lead to substantial economic gains for the regions in the European Union. Our results indicate considerable heterogeneity in economic impact across regions. Keywords: Investment Plan for Europe, impact assessment, general equilibrium model. JEL classification: C54, C68, O52. 1. Introduction In the European Union (EU), goods and services can travel freely without any internal borders or other regulatory barriers. This is commonly thought of having the potential to promote economic growth thanks to economies of Martin Christensen: European Commission, Joint Research Centre, Calle Inca Garcilaso 3, 41092 Sevilla, Spain. E-mail: [email protected], https://orcid.org/0000-0003-4870-7171, corresponding author Andrea Conte: European Commission, Joint Research Centre, Calle Inca Garcilaso 3, 41092 Sevilla, Spain. E-mail: [email protected] Filippo Di Pietro: Universidad de Sevilla, Departamento de economía financiera y dirección de operaciones, Avenida Ramón y Cajal, s/n, 41018 Sevilla, Spain. E-mail: [email protected], https://orcid. org/0000-0003-1573-8553 Patrizio Lecca: PBL Netherlands Environmental Assessment Agency, Bezuidenhoutseweg 30, 2594 AV The Hague, Netherlands. E-mail: [email protected], https://orcid.org/0000-0002-1053-3869 Giovanni Mandras: European Commission, Joint Research Centre, Calle Inca Garcilaso 3, 41092 Sevilla, Spain. E-mail: [email protected], https://orcid.org/0000-0003-1306-0224 We are grateful to the guest editor, Simone Salotti, and two anonymous reviewers for useful comments and suggestions. The views expressed are purely those of the author and may not in any circumstances be regarded as stating an official position of the European Commission. 128 | Martin Christensen, Andrea Conte, Filippo Di Pietro, Patrizio Lecca, Giovanni Mandras scale, increased competition, and innovation, via increased trade and better allocations of production factors and capital flows (Melitz, 2003; Resmini, 2020). Despite the high level of integration among the EU Member States, some cross-border barriers do exist (Mariniello et al., 2015). Non-tariff barriers such as different regulatory regimes pose obstacles to the free movement of goods, services and investments (Springford, 2012). At the end of 2014, the European Commission adopted the Investment Plan for Europe which, among other things, aimed at tackling some of these barriers as a response to the 2008 economic and financial crisis. The Plan consisted of a combination of public and private investments with the initial objective to reach a total direct monetary injection of €315 billion (a target that was met and exceeded four years after the adoption of the Plan, European Commission, 2018a). The initiative was structured along the following three strands or pillars (European Commission, 2014): i)the European Fund for Strategic Investments (EFSI), providing an EU guarantee to mobilise private investment; ii) technical assistance and visibility for investment opportunities; iii)the removal of regulatory barriers to investment both nationally and at the EU level. This paper contains an ex-ante macroeconomic assessment of some of the measures of the third strand related to the development of the European common market. We concentrate here on measures related to the Capital Markets Union (CMU), the Single Market Strategy (SMS), and the Energy Union. We separately evaluate the system-wide economic impact of each initiative identifying for each case an appropriate transmission channel, that is the process through which the initiatives are expected to affect the behaviour of economic agents. To the best of our knowledge, there have been no other attempts at evaluating the macroeconomic effects of the measures analysed here in a general equilibrium framework. These initiatives are regulatory proposals not directly involving monetary injections into the economy. Rather, they are designed to lower barriers to investment and to improve the management of resources and the system-wide efficiency in the economic and financial systems by modifying the regulatory frameworks of each area of intervention accordingly. The analysis is carried out with the dynamic spatial computable general equilibrium (CGE) model called RHOMOLO (Lecca et al., 2018; 2020) and with its Input-Output (IO) version RHOMOLO-IO (Mandras et al., 2019). Note that a version of RHOMOLO is routinely used for the yearly impact assessment of EFSI, which is the first strand of the Investment Plan (Christensen et al., 2019; European Investment Bank, 2018). The RHOMOLO model includes 267 EU and UK regions interlinked through trade subject to transportation costs. This rich regional specification introduces spatial heterogeneity into the model at the level of consumption and production. This allow us to consider the impact from regional differences in income, factor endowments, and trade exposures. Inclusion of agent heteroge- The Economic Potential of the Third Strand of the Investment Plan for Europe | 129 neity has been put forward as one of the main recommendation for enhancing macroeconomic models (see e.g. Vines, Wills, 2018)1. Using a regional model also allow us to consider the region specific impacts of the policies we assess. Hence, our analysis relates to the literature on regional impact of respectively spatially targeted government policies and spatially blind policies. The latter being considered to be those that are universal in their coverage such as those associated with trade, regulation of land use or labour (Hewings, 2014). The policies we are examining in this paper may be considered spatially blind in design but are not spatially blind in effect. The policies we consider change the regulatory burden and affect factor prices and production costs of firms unevenly across regions and result in spatial heterogeneity of economic impact. We find that the regulatory changes that we consider leads to a rise in aggregate EU GDP and Employment. However, because of the spatial heterogeneity in impact some regions gain largely while other experience only modest gains or even loses as a result of the regulatory changes. Clearly, the analysis hinges upon the ability to identify the appropriate transmission channels for the policy change and in quantifying their effects on productivity and production costs. This becomes further difficult due to the fragmentation of the value chain across regions (Hewings, 2014; Temursho, Miller, 2020). In our analysis we therefore consider a range of possible impacts represented by three alternative scenarios; low, medium and high. This report is structured as follows. In Section 2 we present an overview of the building blocks and theoretical foundations of the RHOMOLO model and of its IO counterpart. Sections 3 describes the policy initiatives and identify their appropriate transmission channels. Results are presented in Section 4. Finally, Section 5 concludes. 2. An introduction to RHOMOLO and RHOMOLO-IO 2.1. The RHOMOLO model in a nutshell In this section, we present an overview of the RHOMOLO-V3 model whose full mathematical exposition can be found in Lecca et al. (2018). The theoretical structure of the model is common to other numerical general equilibrium models, with its key distinguishing feature being its geographical granularity. The economy consists of a set of 267 EU and UK NUTS2 regions plus one single exogenous region representing the rest of the World (ROW). Spatial interactions between regional economies are captured through trade of goods and services, income flows, and factor mobility. 1 Although allowing for spatial heterogeneity the version of RHOMOLO used here adopt the assumption of regional representative households and firms. 130 | Martin Christensen, Andrea Conte, Filippo Di Pietro, Patrizio Lecca, Giovanni Mandras The model has a set of ten different economic sectors (industries), with a subset operating under monopolistic competition à la Dixit and Stiglitz (1977). In each region, the monopolistic sectors contain a number of identical firms each producing a differentiated variety that is considered as an imperfect substitute for the other varieties produced within the same region and elsewhere. The number of varieties in the sectors is endogenous and determined from the zero-profit equilibrium condition (according to which profits must be equal to fixed costs). In turn, this means that, in equilibrium, prices equal average costs. In the remaining sectors, firms operate under perfect competition. The version of the model used in this paper employs the following ten aggregations of NACE 2 economic sectors: A, B-E, C, F, G-I, J, K-L, M-N, O-Q, and R-U (see Table 1). All are treated as imperfectly competitive sectors except A, O-Q, and R-U which are modelled as perfectly competitive. Final goods are consumed by households and government (in the form of private and public capital goods), whilst firms consume intermediate inputs. Regional goods are produced by combining value added (labour and capital) with domestic and imported intermediates, creating vertical linkages between firms. Trade both between and within regions is costly, implying that the shipping of goods entails transport costs assumed to be of the iceberg type as in Krugman (1991)2. Transport costs are identical across varieties but spe- 2 The Iceberg type trade costs is a simple way to generate costly trade in the model. However, Table 1: List of NACE Rev. 2 sectors in RHOMOLO Code NACE Rev. 2 Sectors’ description AAgriculture, Forestry and Fishing B-D-E Mining and Quarrying + Electricity, Gas, Steam and Air Conditioning Supply + Water Supply; Sewerage, Waste Management and Remediation Activities CManufacturing FConstruction G-I Wholesale and Retail Trade; Repair of Motor Vehicles and Motorcycles + Transportation and Storage + Accommodation and Food Service Activities JInformation and Communication K-L Financial and Insurance Activities + Real Estate Activities M-N Professional, Scientific and Technical Activities + Administrative and Support Service Activities O-Q Public Administration and Defence; Compulsory Social Security + Education + Human Health and Social Work Activities R-U Arts, Entertainment and Recreation + Other Service Activities + Activities of Households As Employers; Undifferentiated Goods- and Services-Producing Activities of Households for Own Use + Activities of Extraterritorial Organisations and Bodies Source: Lecca et al. (2018). The Economic Potential of the Third Strand of the Investment Plan for Europe | 131 cific to sectors and trading partners (region pairs). The spatial configuration of the system has a direct impact on regions’ competitiveness because firms located in regions that are more accessible can source their intermediate inputs at lower prices and thus gain larger market shares in local markets. Regarding the labour market setup, the model distinguishes three different labour categories corresponding to the level of skill or education: low, medium, and high. For each labour type, the default wage setting relationship is represented by a wage curve (Blanchflower, Oswald, 1994), whose implication is that lower levels of unemployment increase the workers’ bargaining power, thereby increasing real wages. Government expenditure includes current consumption on goods and services, capital expenditures dedicated to public infrastructure, and net transfers to households. Revenues are generated by labour and capital income taxes on household income and indirect taxes on production. In the simulations reported here, government spending is considered an exogenous policy variable. The model is recursively dynamic with myopic expectations, and it is solved sequentially with stocks being upgraded at the beginning of each year. The model is calibrated to a steady-state baseline with no changes in policy. The RHOMOLO model briefly described here is used in the next sections to evaluate the policy initiatives related to the CMU and to the SMS. For the case of Energy Union, we adopt a modified version of RHOMOLO where all supply constraints are removed and direct input substitution effects are neglected. The model is thus transformed to a conventional Inter-Regional IO model (RHOMOLO-IO) where the price elasticity of supply is infinite and the price elasticity of demand is zero. 2.2. The RHOMOLO-IO framework The RHOMOLO-IO framework uses data organised in IO tables with sectoral disaggregated regional economic accounts (see Mandras et al., 2019, for more details). The IO tables represent a snapshot of the flows of products and services produced and consumed within the economy in a single year. The basic principle of the IO tables is to identify and disaggregate all the monetary flows between industries (inter-industry expenditure flows), consumers and supplies of factors in the economy. Under a number of assumptions, IO tables can be used as the basis for an economic model where exogenous final demands drives total output (Leontief, 1986; Miller, Blair, 2009). Supply is infinitely elastic and the determination of inputs is based on fixed technical coefficients. The transmission the iceberg assumption has the limitation that it is difficult to assessment regulatory and innovation improvements in the transportation sector. 132 | Martin Christensen, Andrea Conte, Filippo Di Pietro, Patrizio Lecca, Giovanni Mandras mechanisms linking changes in exogenous demands to changes in aggregate and sectoral activities are called multipliers. These represent the knock-on effects throughout the economy, generated by the change in final demand. In other words, IO multipliers allow us to measure to what extent an increase/ decrease in final demand of one sector, entail expansionary/contractionary effects in the output of all sectors, the perturbed sector included. The activity generated by each sector resulting from the initial demand disturbance is known as the indirect effect. In computational terms the multiplier effect is thus given by the direct plus indirect effects divided by direct effect. IO multipliers can be used to quantify the economic impact derived from a demand-side shock assuming that the average relationships in the IO table apply at the margin. Finally, the two key assumptions in IO are: a)the supply-side of the economy is entirely passive and, b) the production technology for all sectors is represented by fixed coefficients (i.e. an increase in the production of any one sector’s output means a proportional increase in that sector’s input requirements). This means that inputs substitutability is neglected. 3. The simulation strategy 3.1. The cross-border investments initiatives We evaluate the macroeconomic impact of removing national regulatory barriers to facilitate cross-border activity and investment within the CMU. In particular, we concentrate on the evaluation of the following proposals: i) the proposal for a Regulation on crowdfunding3; ii) the proposal to facilitate cross-border distribution of investment funds; iii) the proposal on the assignment of financial claims and a guidance on third party effects of securities to improve legal security for cross-border investment. The broad idea behind such a proposed regulatory framework is to create the necessary incentives to improve the market-based financial system as a complement to bank-based finance of investments (Veron, Wolff, 2015; Quaglia et al., 2016). This should improve the efficiency of the whole EU financial market and reduce the price of capital, as postulated by most of the theoretical constructions adopted in the current economic literature on the deepening and integration of the EU capital market. Essentially, standard neoclassical theory predicts positive growth effects from more integrated financial markets based on the reduction of capital cost driven by capital 3 With a corresponding amendment to Directive 2014/65/EU on the Markets in Financial Instruments MiFID II. The Economic Potential of the Third Strand of the Investment Plan for Europe | 133 flows moving from capital-abundant regions towards capital-scarce ones. However, the existing empirical evidence does not unambiguously confirm this theoretical prediction and tends to conclude that there are only limited effects at best. For instance, Bekaert et al. (2005) and Quinn and Toyoda (2008) find positive growth effects, but Edison et al. (2002) only find mixed effects. On the contrary, Neumann et al. (2009), Stiglitz (2004), and Levchenko et al. (2009) find negative effects of financial integration on growth through the reduction of the cost of capital. According to Gehringer (2013), financial openness generated a strong positive impact on economic growth and factor productivity in Europe. Other studies affirm that financial openness directly affects factor productivity by stimulating financial development (see, among others, Rajan, Zingales, 2003). According to Gourinchas and Jeanne (2006), analysing the effects of international financial integration on productivity is more important than examining its investment growth effects. Typically, financial openness is expected to have a positive impact on productivity via a better and more efficient allocation of resources (Kose et al., 2009), as well as due to easier access to investment opportunities (Giannetti et al., 2002). Furthermore, financial openness may result in less capital constraints, permitting the economy to engage in more productive investments (Acemoglu et al., 2006). In addition, capital account openness may spur financial development (Baltagi et al., 2009). A well-developed financial system contributes to a better risk diversification and encourages foreign investors to shift at least part of their investments from safe and low-yield to risky but more profitable locations (Obstfeld, 1994; Sandri, 2010). In line with the economic literature summarised above, productivity shocks appear to be the most appropriate transmission channel to evaluate the potential impact of more integrated capital markets in the EU. Thus, we frame our simulation strategy in accordance with this approach within RHOMOLO. In particular, we aim at capturing the system-wide impact of greater financial openness within the EU through permanent changes in total factor productivity (TFP). The TFP shock we implement is based on Gehringer (2013) who estimates the impact of financial openness on TFP growth, which is the TFP elasticity to financial openness. To the best of our knowledge, this study is the most up-to-date analysis measuring such elasticity in relationship with the EU capital market. The estimated elasticity is equal to 0.043 with a standard error of 0.012 (Gehringer, 2013, Table 2). Applying such elasticity to the time series index of financial openness by Chinn and Ito (2008) and data on TFP levels for the 28 EU countries from 1996 to 2017 (AMECO database), we are able to reconstruct the evolution of TFP up to 2017 for each country solely driven by financial integration. We simulate three alternative scenarios, Low, Central, and High, based on the uncertainty attached to the estimated TFP elasticity to financial open- 134 | Martin Christensen, Andrea Conte, Filippo Di Pietro, Patrizio Lecca, Giovanni Mandras ness quantified via the standard deviation associated to the point estimate by Gehringer (2013). Namely, for the construction of the Central scenario we use the point estimate of 0.043, while the Low and High scenarios use the point estimate -/+ two standard deviations, respectively. The three newly created TFP series (one for each scenario) are then taken as benchmark and used to construct the TFP shocks for the model. The idea behind our simulation is the following: We assume that the development of the capital markets union associated to the regulatory proposals on cross-border activities may exert effects comparable to those observed during the 3 years either before the adoption of the euro (for the old EU Member States) or before the entrance into the EU (for newer members). Thus, for the Member States that adopted the euro in 1999, the TFP shock is defined as the period by period percentage changes computed during the period 1997-1999 with respect to the 1996 TFP level (bear in mind that such changes are solely due to financial openness by construction, as explained above). The third period’s TFP change is then kept as a permanent shock in the model. We use the same approach for the countries Denmark, Sweden and the UK that have negotiated the right to opt-out from participation to the Eurozone. For the countries that joined the EU at a later stage, and are therefore still in the adjustment phase, the TFP shock is created using a different reference period which is the three years preceding the entrance in the EU4. For instance, Romania joined the EU in 2007, it is currently outside the Eurozone, but committed to the adoption of the euro once it fulfils the necessary conditions. In this case, the reference period is 2005-2007. 3.2. The Start-up and Scale-up initiative The Start-up and Scale-up EU initiative aims to create a favourable environment to innovation, entrepreneurship, and competitiveness (European Commission, 2016a; 2016b). With this initiative, the European Commission is determined to eliminate the administrative and regulatory barriers currently existing in the fragmented EU Single Market that may particularly constraint start-up and scale-up companies in their growth potential. We focus here on the part of the initiative built to improve access to finance for start-ups in the EU. The European Commission and the EIB Group launched a Pan-European Venture Capital Fund of Funds to attract private funding from institutional 4 We thus approximate the potential gain from the cross-border investments initiatives for newer Member States by changes in their financial openness index three years prior to entrance in the EU. An alternative approach could be to assume an increase in TFP similar to that observed (on average) in the Member of the EU before the introduction of the euro. This would result in a higher TFP growth in seven of the newer Member States and result in lower TFP growth in six of the newer Member States. The Economic Potential of the Third Strand of the Investment Plan for Europe | 141 years after the implementation of the policy9, in all three scenarios. Then, the economy expands further year by year through capital accumulation. In 2030, the GDP is likely to increase by between 0.21% and 0.75%, with the central estimates being equal to +0.48% relative to the initial steady-state. As for employment, our modelling simulations suggest a minimum impact of +0.07% and a maximum impact of +0.26% from the initial steady-state (corresponding to 157 and 547 thousand FTEs respectively), with the central estimate being equal to +0.17% (equivalent to 353 thousand FTEs). Interestingly, in percentage term the rise in employment is less than the rise of GDP, meaning that capital is increasing proportionally more than GDP. This reflects substitution effects in favour of capital generated by a relatively greater reduction of the price of capital compared to the price of labour, both measured in efficiency units. To examine the regional impact of the TFP shock, Figure 1 shows boxplots of the percentage changes relative to the initial steady-state for selected 9 Recall that we assume the reform to yield three consecutive years of TFP improvement after which the TFP gain becomes permanent. Table 4: Macroeconomic impact of the cross-border investments initiatives (percentage changes from baseline) – Low, Central, and High scenarios 2020 2025 2030 Low GDP 0.15 0.19 0.21 Employment 0.03 0.06 0.07 Household consumption 0.09 0.13 0.15 Export to the ROW 0.29 0.34 0.37 CPI -0.09 -0.10 -0.10 Central GDP 0.35 0.44 0.48 Employment 0.08 0.14 0.17 Household consumption 0.20 0.29 0.33 Export to the ROW 0.66 0.76 0.83 CPI -0.21 -0.22 -0.23 High GDP 0.54 0.68 0.75 Employment 0.12 0.21 0.26 Household consumption 0.32 0.46 0.52 Export to the ROW 1.03 1.19 1.30 CPI -0.33 -0.34 -0.36 Source: RHOMOLO calculations. 142 | Martin Christensen, Andrea Conte, Filippo Di Pietro, Patrizio Lecca, Giovanni Mandras variables in the central scenario for the years 2020 and 2030. Lower factor prices causes a general decline in the consumer price index (CPI) for all regions with the median regions experience a decline in CPI of 0.22% in 2030. Improved competitiveness results in higher export, GDP, consumption and employment for most regions. We notice considerable regional variation in impacts from the TFP shock. While most regions benefit from the change in regulation, a small number of regions suffer losses. For example, the median region experience a rise in GDP of 0.49% in 2030 with the regional change in GDP varying from -0.11% to 1.19%. The rise in employment for the median regions is 0.15% in 2030 with the change in employment ranging from -0.13% to 0.74%. The largest gains in GDP and employment are found in the Baltic countries, and in regions in Romania and Ireland (as shown for GDP in Figure 2). Swedish regions also experience some of the largest gains in GDP. Malta experiences a loss in GDP while Cyprus and regions in Poland experience only modest gains in GDP. We also find modest gains in GDP in Greece and Southern Italy. Losses in employment are found in regions in South, Insular and North-West Italy, and in Malta. Typically, regions placed in the lower quartile of GDP gains are also placed in the lower quartile of employment gains. With the aim of identifying the sources of the increased competitiveness at the regional level, in Figure 3 we plot changes in the regional GDP (horizontal axes) against changes in exports (vertical axes) obtained under the Central scenario. We show result for the short run i.e. after one year (panel a) and the long run i.e. the new steady-state (panel b). Both in the short run and in the long run we observe a positive correlation between changes in Figure 1: Regional impact of the cross-border investments initiatives – Central scenario. Source: Authors’ calculation on RHOMOLO. The Economic Potential of the Third Strand of the Investment Plan for Europe | 143 the GDP and exports, with the correlation being weaker in the short run as supply constraints are in place to mimic adjustment delays of firms. As explained above, the fall in the price of primary factors generates a fall in commodity prices boosting regional competitiveness particularly with the ROW. It is worth noticing that the regions above the fitting lines are those benefiting the most from such improved competitiveness effects, while the rest of the regions experience smaller changes in export. In Table 5 we divide the regions into quartiles based on GDP per capital in the baseline, the regions with the lowest GDP per capital in quartile one and the regions with the highest GDP per capital in quartile four. For each quartile group we calculate the average change in GDP per capita relative to the baseline. We find that all quartile groups on average benefit from the reform proposal, with the riches EU regions befitting the most in terms of GDP (on average a rise of 0.51% relative to the baseline). This suggests that the cross-border investments initiatives would contribute to economic growth in the EU but would not lead to economic convergence across regions. Figure 2: Regional GDP impact in 2030 of the cross-border investments initiatives – Central scenario. Source: Authors’ calculation on RHOMOLO. 144 | Martin Christensen, Andrea Conte, Filippo Di Pietro, Patrizio Lecca, Giovanni Mandras 4.2. The Start-up and Scale-up initiative Figure 4 plots the percentage changes of EU GDP from the initial steadystate obtained under the three scenarios. We observe a permanent economic Figure 3: Impact on the relative competitiveness of regions: short-run and long-run – Central scenario. Source: Authors’ calculation on RHOMOLO. Table 5: Relative change in GDP per capita by baseline quartile groups – Central scenarios 2020 2025 2030 Quartile 1 0.240 0.329 0.379 Quartile 2 0.322 0.403 0.443 Quartile 3 0.361 0.448 0.489 Quartile 4 0.367 0.464 0.510 Source: RHOMOLO calculations. The Economic Potential of the Third Strand of the Investment Plan for Europe | 145 expansion in all three scenarios, with the GDP increase between 0.022% and 0.036% above baseline values in year 2030 (corresponding to €2.8 and €4.7 billion added to EU GDP, respectively). For the first three simulation years the assumed survival rate of start-up firms is identical for all scenarios. After three years the assumed survival rate of start-ups declines in the low and central scenario resulting in lower GDP impact. After six years, the assumed survival rate of start-up firms in the low scenario declines further and this results in the lower GDP increase. Table 6 sheds light on the adjustment mechanism by reporting the percentage deviation from the initial steady-state of selected variables for the years: 2020, 2025, 2030. We can see that GDP, employment and household consumption all increase year by year eventually settling to a new steadystate where all variables grow at a constant rate. The reduction in the risk premium spurs a rise in investments and accumulation of capital. At first, the one-year time to build assumption on productive capital impose a short-run capacity constraint. This means that in the first simulation year the economy responds to the shock as if it was a conventional demand-side shock with no direct supply-side response. The higher investment demand puts initial upward pressure on factor prices and domestic commodity prices resulting in a decline in exports to ROW. The EU is running a current account deficit to help finance the rise in investments. However, as capital accumulates in the medium to long term direct supply-side effects through increased capital stock accompany the demand-side effect of the shock. Capital accumulation puts downward pressure on commodity prices and these falls below the initial steady-state and the gain in competitiveness leads to a rise in exports to Figure 4: Macroeconomic impact on GDP (Percentage changes from baseline) – Low, Central, and High scenarios. Source: Authors’ calculation on RHOMOLO. 146 | Martin Christensen, Andrea Conte, Filippo Di Pietro, Patrizio Lecca, Giovanni Mandras ROW. We note that the rise in GDP is higher than the rise in employment as capital accumulation leads to a fall in the price of capital and a substitution away from labour. Employment in 2030 rises between 0.014% and 0.024% (corresponding to an increase by between 30 and 50 thousands FTEs with the central estimates being equal to 39 thousands FTEs). Figure 5 depicts the distribution of regional changes in selected variables in the central scenario for the years 2020 and 2030. Capital accumulation gradually raises regional GDP, exports and employment. In 2030, all regions experience a rise in GDP and employment relative to the initial steady-state. The increase in GDP and employment for the median region is respectively 0.03% and 0.02%. Household consumption rises in most regions. However, regional divergence means that a few regions experience relative large increases in household consumption (North-Eastern Scotland, Bucharest and Helsinki) while other regions suffers small declines in household consumption (Cheshire, Merseyside, the Finnish region of Åland and regions in North- East Italy). We find the lowest gains in GDP and Employment in Cyprus, Estonia and Spanish regions (see Figure 6). The highest gains in GDP and employment can be found in a geographically dispersed group of regions Table 6: Macroeconomic impact of the large procurement projects package (percentage changes from baseline) – Low, Central, and High scenarios 2020 2025 2030 Low GDP 0.0129 0.0217 0.0216 Employment 0.0090 0.0139 0.0141 Household consumption 0.0132 0.0181 0.0186 Export to the ROW -0.0410 0.0107 0.0073 CPI 0.0108 -0.0031 -0.0016 Central GDP 0.0129 0.0237 0.0282 Employment 0.0090 0.0154 0.0184 Household consumption 0.0132 0.0206 0.0244 Export to the ROW -0.0410 -0.0028 0.0042 CPI 0.0108 0.0006 -0.0009 High GDP 0.0129 0.0288 0.0362 Employment 0.0090 0.0188 0.0237 Household consumption 0.0132 0.0253 0.0315 Export to the ROW -0.0410 -0.0086 0.0035 CPI 0.0108 0.0021 -0.0007 Source: RHOMOLO calculations. The Economic Potential of the Third Strand of the Investment Plan for Europe | 147 Figure 5: Regional impact of the Start-up and Scale-up initiative – Central scenario. Source: Authors’ calculation on RHOMOLO. Figure 6: Regional GDP impact in 2030 of the Start-up and Scale-up initiative – Central scenario. Source: Authors’ calculation on RHOMOLO. 148 | Martin Christensen, Andrea Conte, Filippo Di Pietro, Patrizio Lecca, Giovanni Mandras (including Latvia, Northern Scotland, Bucharest and Bratislava region) characterised by having relatively low calibrated risk premia. We formulate the simulated policy shock as a uniform percentage point reduction in the risk premia across all regions. Hence, the lower the regional initial pre-shock risk premium the higher the resulting relative change in regional user cost of capital. Regions with lower initial risk premia will therefore experience the largest relative rise in private investments. In Table 7, we consider the average change in GDP per capita by quartile groups. The table reveal that all quartile groups experience a rise in GDP per capita relative to the baseline. We observe the largest rise in GDP per capita in year 2030 for quartile one (the regions with the lowest GDP per capital in the baseline). However, the variation in average GDP per capita deviations relative to the baseline are modest across quartile groups. This suggest a modest economic convergence across EU regions from the Startup and Scale-up initiative. 4.3. The large procurement projects package Table 8 contains the results of the three scenarios expressed in percentage changes from the initial steady-state for selected variables for the years 2020, 2025, and 2030. The aggregate economic gains from the large procurement projects package are relatively small. Our simulations suggest that improvements in public procurement procedures could result in GDP rise by 2030 of between 0.0041% and 0.0074% relative to the initial steady-state. The central estimate being a rise in GDP by 2030 of 0.0057%. Employment by 2030 rises by between 0.0014% and 0.0025% (correspond to increase by between 3 and 5 thousands FTEs). The increase in public capital efficiency generates system-wide efficiency gains. Services from public capital enters as an unpaid factor in production and improved public capital efficiency therefore reduces production costs. This means that the economy becomes more competitive through positive terms of trade effects leading to a rise in exports to ROW and an improvement in the current account. Table 7: Relative change in GDP per capita by baseline quartile groups – Central scenarios 2020 2025 2030 Quartile 1 0.013 0.027 0.034 Quartile 2 0.013 0.022 0.026 Quartile 3 0.014 0.024 0.029 Quartile 4 0.013 0.024 0.029 Source: Authors’ calculation on RHOMOLO. The Economic Potential of the Third Strand of the Investment Plan for Europe | 149 We find that the increase in public capital efficiency benefit relatively more the primary (A and B-E) and the manufacturing sectors than the service sectors as suggested by Figure 7 plotting the percentage changes from the steady-state for the central scenario for the years 2020, 2025, and 2030. Figure 8 shows the regional changes in selected variables in the central scenario. All regions benefit from lower CPI and in turn a rise in total export and GDP. Our results suggest that the rise in GDP for the median region is 0.0055% in 2030. Most regions also experience a rise in employment and household consumption. However, regional variation exists with the highest gains in GDP and employment found in regions in Belgium, Romania and Spain (as illustrated for GDP in Figure 9). We find modest gains in Cyprus, and regions in Greece and Finland. Regions in North- West, South and Insular Italy experience modest gains in GDP paired with a modest decline in employment and household consumption. Table 9 compares the relative changes in GDP per capita across quartile groups. We find modest gains in GDP per capital across all regions. Table 8: Macroeconomic impact on key economic variables (percentage changes from baseline) – Low, Central, and High scenarios 2020 2025 2030 Low GDP 0.0010 0.0026 0.0041 Employment 0.0002 0.0008 0.0014 Household consumption 0.0005 0.0016 0.0026 Export to the ROW 0.0019 0.0048 0.0073 CPI -0.0006 -0.0015 -0.0022 Central GDP 0.0013 0.0037 0.0057 Employment 0.0003 0.0011 0.0020 Household consumption 0.0008 0.0023 0.0037 Export to the ROW 0.0027 0.0067 0.0103 CPI -0.0009 -0.0021 -0.0031 High GDP 0.0017 0.0047 0.0074 Employment 0.0004 0.0015 0.0025 Household consumption 0.0010 0.0030 0.0048 Export to the ROW 0.0034 0.0087 0.0132 CPI -0.0011 -0.0027 -0.0040 Source: Authors’ calculation on RHOMOLO. 150 | Martin Christensen, Andrea Conte, Filippo Di Pietro, Patrizio Lecca, Giovanni Mandras Figure 7: Sectoral output effects (percentage changes from baseline) – Central scenario. Source: Authors’ calculation on RHOMOLO. Figure 8: Regional impact of the large procurement projects package – Central scenario. Source: Authors’ calculation on RHOMOLO. Regions in quartile one (the regions with the lowest GDP per capita in the baseline) experience marginally higher long run gains in GDP per capita than other quartile groups. The Economic Potential of the Third Strand of the Investment Plan for Europe | 157 European Investment Bank (2018), Assessing the Macroeconomic Impact of the EIB Group. Economic – Impact Studies. Luxembourg: European Investment Bank. DOI: 10.2867/140089. Freire-González J. (2017), Evidence of Direct and Indirect Rebound Effect in Households in EU-27 Countries. Energy Policy, 102: 270-276. DOI:10.1016/j. enpol.2016.12.002. Gehringer A. (2013), Growth, Productivity and Capital Accumulation: The Effects of Financial Liberalization in the Case of European Integration. International Review of Economics & Finance, 25: 291-309. DOI:10.1016/j.iref.2012.07.015. Giannetti M., Guiso L., Jappelli T., Padula M., Pagano M. (2002), Financial Market Integration, Corporate Financing and Economic Growth. Brussels: Directorate General Economic and Financial Affairs (DG ECFIN), European Commission. European Economy – Economic Papers 2008-2015 n. 179. Gourinchas P., Jeanne O. (2006), The Elusive Gains from International Financial Integration. The Review of Economic Studies, 73, 3: 715-741. DOI:10.1111/j.1467- 937X.2006.00393.x. Hewings G. J. D. (2014), Spatially Blind Trade and Fiscal Impact Policies and Their Impact on Regional Economies. The Quarterly Review of Economics and Finance, 54, 4: 590-602. DOI:10.1016/j.qref.2014.04.007. IMF (2015), Making Public Investment More Efficient. Washington, DC: International Monetary Fund. IMF Policy Papers, June. Kose M. A., Prasad E. S., Terrones M. E. (2009), Does Openness to International Financial Flows Raise Productivity Growth? Journal of International Money and Finance, 28, 4: 554-558. DOI:10.1016/j.jimonfin.2009.01.005. Krugman P. (1991), Increasing Returns and Economic Geography. Journal of Political Economy, 99, 3: 483-499. DOI:10.1086/261763. Lecca P., Barbero J., Christensen M., Conte A., Di Comite F., Diaz-Lanchas J., Diukanova O., Mandras G., Persyn D., Sakkas S. (2018), RHOMOLO V3: A Spatial Modelling Framework. Luxembourg: Publications Office of the European Union. DOI: 10.2760/671622, JRC111861. Lecca P., Christensen M., Conte A., Mandras G., Salotti S. (2020), Upward Pressure on Wages and the Interregional Trade Spillover Effects under Demand-Side Shocks. Papers in Regional Science, 99, 1: 165-182. DOI:10.1111/pirs.12472. Lecca P., McGregor P. G., Swales J. K., Turner K. (2014), The Added Value from a General Equilibrium Analysis of Increased Efficiency in Household Energy Use. Ecological Economics, 100(C): 51-62. DOI:10.1016/j.ecolecon.2014.01.008. Leontief W. (1986), Input-Output Economics. Oxford: Oxford University Press. Levchenko A. A., Rancière R., Thoenig M. (2009), Growth and Risk at the Industry Level: The Real Effects of Financial Liberalization. Journal of Development Economics, 89, 2: 210-222. DOI:10.1016/j.jdeveco.2008.06.003. Mandras G., Conte A., Salotti S. (2019), The RHOMOLO-IO Modelling Framework: A Flexible Input-Output Tool for Policy Analysis. Seville: European Commission. JRC Working Papers on Territorial Modelling and Analysis n. 06. Mariniello M., Sapir A., Terzi A. (2015), The Long Road Towards the European Single Market. Brussels: Bruegel. Bruegel Working Paper n. 01. 158 | Martin Christensen, Andrea Conte, Filippo Di Pietro, Patrizio Lecca, Giovanni Mandras Melitz M. (2003), The Impact of Trade on Intra-Industry Reallocations and Aggregate Industry Productivity. Econometrica, 71, 6: 1695-1725. DOI:10.1111/1468- 0262.00467. Miller R. E., Blair P. D. (2009), Input-Output Analysis. Foundations and Extensions. Cambridge, UK: Cambridge University Press. Neumann R. M., Penl R., Tanku A. (2009), Volatility of Capital Flows and Financial Liberalization: Do Specific Flows Respond Differently? International Review of Economics and Finance, 18, 3: 488-501. DOI:10.1016/j.iref.2008.04.005. Obstfeld M. (1994), Risk-Taking, Global Diversification, and Growth. American Economic Review, 84, 5: 1310-1329. Available at: https://www.jstor.org/stable/2117774. Quaglia L., Howarth D., Liebe M. (2016), The Political Economy of European Capital Markets Union. Journal of Common Market Studies, 54, S1: 185-203. DOI:10.1111/jcms.12429. Quinn D. P., Toyoda A. M. (2008), Does Capital Account Liberalization Lead to Economic Growth? An Empirical Investigation. Review of Financial Studies 21, 3: 1403-1449. DOI:10.1093/rfs/hhn034. Rajan R. G., Zingales L. (2003), The Great Reversals: The Politics of Financial Development in the Twentieth Century. Journal of Financial Economics, 69, 1: 5-50. DOI:10.1016/S0304-405X(03)00125-9. Resmini L. (2020), Openness and Regional Growth. Scienze Regionali – Italian Journal of Regional Science, 19, 3: 395-430. DOI:10.14650/98285. Sandri D. (2010), Growth and Capital Flows with Risky Entrepreneurship. Washington, DC: IMF. IMF Working Paper n. 10/37. Springford J. (2012), How to Build European Services Markets? London: Centre for European Reform, September. Stiglitz J. E. (2004), Capital-Market Liberalization, Globalization, and the IMF. Oxford Review of Economic Policy, 20, 1: 57-71. DOI:10.1093/oxrep/grh004. Temursho U., Miller R. E. (2020), Distance-Based Shared Responsibility. Journal of Cleaner Production, 257, 120481. DOI:10.1016/j.jclepro.2020.120481. Veron N., Wolff G. B. (2015), Capital Markets Union: A Vision for the Long Term. Journal of Financial Regulation, 2, 1: 130-153. DOI:10.1093/jfr/fjw006. Vines D., Wills S. (2018), The Rebuilding Macroeconomic Theory Project: An Analytical Assessment. Oxford Review of Economic Policy, 34, 1-2: 1-42. DOI:10.1093/oxrep/grx062. Vogel L. (2009), Macroeconomic Effects of Cost Savings in Public Procurement. Brussels: European Commission. European Economy Economic Papers n. 389. DOI:10.2765/29810.