5

How does the future look for the EU?

How does the future look for the EU?

How does the future look for the EU?

Previous chapters have shown that the integration process in Europe was prompted by an initial commitment by France and Germany to jointly support collaboration rather than confrontation. The creation of the EU and its initial progress were supported by this agreement. Assessments of relative power show that Germany, which still remains the dominant nation in this region, was most supportive of early structural changes. A key German commitment was to provide the largest transfers to the EU for economic adjustments due to integration. As European integration deepened and the EU gained more members, the preponderance of Germany in the regional hierarchy gradually declined. France, Italy, and the UK closed the gap between themselves and Germany. As Germany’s capacity to provide necessary resources for integration reached its limit, integration plateaued because of the lowering of guarantees for ameliorating negative implications of integration. This shift is reflected in our results that show at higher levels of integration, further deepening no longer depends on the dominant states within the hierarchy. Instead, we show that deepening of integration is driven by convergence of values and continued trust in EU institutions. In sum, satisfaction with the status quo becomes the driver in the later stages of integration.

The ongoing Brexit process discloses that members that have lagged in their commitment to “ever closer union” may choose to leave the EU and shift regulatory decisions back to their national governments. This is the case for the UK, which was often uncomfortable with integration steps that went beyond the common market. Recall that Britain chose not to join the Eurozone and rejected Schengen. Furthermore, we show that Britain’s values are not converging with those of the EU core. Certainly, Britain’s value convergence is more with those of the US than its European neighbors. Moreover, we also show that since joining the EU, Britain maintained a lower level of trust in EU institutions than any other member state. Because of this reluctance to integrate, Britain has consistently opposed political proposals that would move the EU further towards a political union. By removing Britain as a member of the EU, we speculate and test here if further deepening of integration may follow among the rest of the member states.

National decisions addressing the forced migration from war-torn nations – particularly Syria, Iraq, Afghanistan, and Central and North Africa – have placed serious strains on the freedom of citizens to move within the EU. Many suspect that fear of Middle Eastern migrants disproportionately relocating to Britain was a major factor in the vote for Brexit. Refugees have been welcomed in Germany, but many countries in the EU, most prominently Hungary, Poland, and Bulgaria, have refused to accept immigrants and sealed their borders counter to EU rules. Greece and Italy have taken the brunt of new entries but now seek to disperse the new immigrants. While migrations after World War II dwarf the size and scope of the refugee flows currently underway, the political impact on the EU is palpable. We explore the migration divide within the EU to assess the impact of this externally induced division in European unity and the willingness to further integrate.

An additional concern revolves around European security. It is conceivable that the Trump administration might substantially reduce US commitment to NATO. This, in turn, raises a serious challenge for Europe’s CFSP. Strengthening the CFSP against external threats may require substantial financial commitment by member states. Such financial requirements cannot proceed without revision of current EMU fiscal limits. However, the current political opposition to furthering the centralization of financial transitions within the EU has shown that the undertaking of such drastic measures is anything but easy. Yet, if the Trump administration casts serious uncertainties upon the future of NATO and the security interests of the EU, the defense needs of the EU could force member states to revise the fiscal restrictions on national budgets and push them to adopt a supranational-level solution by adopting a fiscal union.

In this chapter, we address these challenges through macro and micro analyses. First, we present a macro-level analysis of the EU as a global actor in competition with other giants given the current state of integration among its members. In other words, holding all variables constant what would be the EU’s position in the global hierarchy? Once this picture is clear, we next examine the member states’ policy preferences over key challenges that are central to further deepening of integration. Finally, we provide a forecast analysis of the challenges presented in Chapter 3, using the theoretical model presented in Chapter 2 and tested in Chapter 4.

Competition among the giants

In this section, we look at the EU as a global actor. In the past, competition among the more powerful states of the global hierarchy has led to confrontations under conditions of parity when competing sides rejected the status quo. This observation is fundamental to power transition theory. Allison Hamlin and Jacek Kugler summarized the long-term global record of observations from 1700 through the present.1 Limitations of data restrict somewhat the length of this analysis, but Figure 5.1 shows that starting with the French Revolution, the most severe global wars were waged at parity among dissatisfied contenders.

Figure 5.1: Power Parity and Severity of Conflict.

In Figure 5.1 the X-axis represents power where parity, measured by GDP, is achieved at 1.0. The Y-axis indicates the severity of conflict in terms of total deaths. The figure shows us that the overwhelming number of severe conflicts occur in or around parity – challenging the notion that a balance of power preserves peace. The one glaring exception is the conflict Japan initiated against the United States during World War II.2

Parity and dissatisfaction can lead to severe conflict and power asymmetry and dissatisfaction can lead to limited conflict. However, parity and satisfaction, as we have shown, can lead to cooperation while a combination of power asymmetry and satisfaction improves the likelihood for early integration. Let us then present the basic data before we provide a very tentative assessment of the likelihood of conflict and cooperation among the major powers.

The productivity of top members of the global hierarchy can be used to anticipate the likely trajectory that sets the necessary preconditions for major conflict or cooperation. Figure 5.2 summarizes the relative GDP and productivity of major global powers. The EU’s power is an aggregate of members controlled for when they joined the Union. Anticipating Brexit, we reduce the capability of the EU by removing the UK starting in 2020. The thickness of each time assessment indicates the productivity of individuals within each society. The United States dominates the global system from 1950 to the present, but this preponderance is now challenged by China. The European Union holds a steady third place even without Britain. Russia, despite its current visibility, will not be a major player in the future. Instead, India, because of its population size, is expected to emerge as a major power sometime beyond 2035.

Figure 5.2: Power and Productivity of Global Contenders, 1950–2030 (in PPP 1990 US trillions).

Another key consideration is the EU’s regional hegemony. As the dominant power in Europe, the EU, as a single actor, can possibly provide stability. If integration deepens between EU member states, the resulting Union would be capable of meeting any economic or security threat posed by Russia. On the other hand, neither Germany, France, nor the UK has individual capabilities to confront Russia’s growing challenge in Europe. Given the power projections presented in Figure 5.2, we examine the likelihood of cooperation or conflict between the EU and Russia, and the EU/US and Russia/China.

The EU and Russia

After 1989, Russia endured a severe and protracted economic downturn. Hyperinflation plunged millions of people into poverty. A sharp fall in GDP, plummeting foreign trade, a high unemployment rate, an extraordinary increase in income differentials, a rise in corruption, and a massive budget deficit dramatically affected income. Coupled with low oil prices, the negative economic performance created serious economic problems in the reckless early 1990s. Real GDP fell by 8.7 percent in 1993 and 12.7 percent in 1994, indicating a severe recession.3 The EU and US made significant contributions to Russian economic recovery through various assistance programs.4 Many Western analysts and policy-makers anticipated that privatization, price and trade liberalization reforms, and currency convertibility would ultimately lead to macroeconomic stability in Russia. Such changes would then put this country on its way to becoming a more democratic and faster growing market economy. Early agreements seemed to lead in this direction. In 1998, the United States and the European Union invited Russia to join the Group of Seven (G7). This major step recognized the prominent role that Russia could play among the top industrialized and highly developed democracies despite its temporary economic decline. Moreover, relations with the West improved further in 2012, when after 19 years of tortuous stop–start negotiations, the World Trade Organization (WTO) officially welcomed the Russian Federation to join the organization as its 156th member. Russia was finally granted the right to have a say in global trade forums. Admitting Russia would bring the country international prestige, enhance its economic and political interactions, and adhere it more closely to the West.

However, the opposite occurred. Failure of the 1995 Oslo negotiations between Presidents Clinton and Yeltsin to bring Russia into NATO, coupled with the admission of the former Soviet Baltic states, enraged the Russians and paved the way for a more hostile relationship. Vladimir Putin increased state control of the economy, started a massive military build-up, intervened and occupied parts of Georgia and Ukraine, annexed Crimea, and caused dangerous regional instability. Russia’s offensive postures made Western leaders very concerned about the future of regional stability on the eastern border of the EU. The early post-Soviet optimism about a full reconciliation between Russia and the West dissipated and initial rapprochement between the EU and Russia that started after the collapse of the USSR ended.

A standoff with Russia was prompted by EU attempts to incorporate Ukraine as an EU associate member. A prolonged crisis in Ukraine began in November 2013 and deposed President Viktor Yanukovych who had suspended implementation of an association agreement with the EU. Mass protests known as the “Euromaidan” toppled the government. This in turn led to a reaction by Russia that culminated in the annexation of Crimea and continued support for east Ukrainian dissidents who seek to partition the country or destabilize the government in Kiev.

In view of this deteriorating status quo, Figure 5.3 shows the power distribution between the EU and Russia controlling for the UK dropping out of the Union in 2019. The distribution of power shows the EU as the dominant power in the European region for the foreseeable future. With or without US support, a conventional confrontation in Europe favors the EU.5 On the other hand, were the EU to collapse (and we do not forecast this eventuality), direct confrontations between Germany and Russia or France and Russia would favor Russia as these societies are all very similar in power to Britain.

Figure 5.3: Power and Productivity of the EU, Russia, and the UK, 1950–2030 (in PPP 1990 US trillions).

Currently, tensions in EU–Russia relations continue despite uncertainty regarding the US position. It is not likely therefore that the status quo will improve in the next few years sufficiently to anticipate cooperation between the EU and Russia. Thus, we examine the probability of conflict between the EU and Russia using a specification of the power transition model provided by Efird, Genna, and Kugler.6 For estimation purposes, we use a conflict–cooperation continuum that ranges from 1 to 10, where 1 represents federated integration and 10 represents severe conflict. Using the Efird, Genna, and Kugler model with value convergence and power asymmetry to reflect hierarchy and the status quo, we can preliminarily forecast the likely interactions between the EU and Russia. Figure 5.4 indicates a stable confrontational relationship through 2035. Unless major political changes take place, the confrontational posture currently in play will persist. Although we do not forecast war, the results indicate a set of preconditions for a continuing low level of hostilities.

Figure 5.4: Estimated Relations between the EU and Russia, 2016–2035.

These results are concerning but concurrently reassuring. Despite the visibility of the Ukraine crisis and the sharp focus on Crimea, the likelihood of an escalation from crisis to war is very limited. Even without US participation, the EU towers over Russia. The combined forces of France and Britain provide a significant deterrence to a potential Russian threat. In a conventional conflict, Russia could inflict serious short-term costs on the EU. Yet, in a prolonged confrontation, Russia would not be able to match the combined capabilities of the EU. Only in the unlikely event of EU disintegration (which we do not anticipate) would Russia prevail in a confrontation with the EU. At the same time, no individual member state can match Russian capabilities – so dissent about external security among EU members and distancing from the US can both contribute to a rise in the likelihood of conflict along the eastern border of the EU.

NATO and the China–Russia relationship

In this section we examine relations between NATO and a Sino-Russian alliance. Figures 5.5 and 5.6, using the same method used in the previous section, outline the structural conditions and likelihood of cooperation and conflict, respectively. Initial results show that – as common sense indicates - the EU–US remains the preponderant power in the global arena. Even with a competitive China–US relationship, where China is expected to overtake the latter as the dominant global economic power by 2025–2030, NATO allies maintain preponderance over a Russia–China alliance for the foreseeable future.

Figure 5.5: Power and Productivity of the EU–US and China–Russia, 1950–2030 (in PPP 1990 US trillions).

Figure 5.6: Estimated Relations between the EU–US and China–Russia, 2016–2035.

The reason is simple. Russia is a visible, active international actor that is no longer a very powerful country. As its population continues to decline, and assuming that productivity remains static, Russia will remain an important power – like Germany, Japan, or France – but can no longer contest global dominance. Indeed, the asymmetric relationship between China and Russia is such that if these countries disagree, Russia will become a vassal of China. This is increasingly evident despite prominence in the Middle East. As Russian trade shifts east away from the EU, China, if it so desires, could increasingly control the economic future of its much smaller partner. With a retreat towards Europe blocked by the Ukraine dispute (see above), the options for Russia are limited.

We replicate the procedures described for Figure 5.4 to assess the relationship between East and West. Figure 5.6 show a slightly higher confrontation potential between NATO and China/Russia than between that the EU and Russia. The results are consistent with derivations from power transition that as parity approaches, confrontation between dissatisfied countries rises. The asymmetry of power between East and West constrains the possibility of a high cost war (10). In sum, a conflict between East and West is unlikely for this time period. However, prospects for cooperation along the eastern border of the EU are equally unlikely. The analysis shows that if current interactions are maintained, the stability of the international system will be increasingly challenged.

Estimating further integration

The pervious section assumed stable or deepening integration. The question that remains unanswered is: what are the prospects for the member states to deepen the integration? By utilizing the results obtained in the multinomial regression model in Chapter 4, we can estimate the likelihood of further integration. Since it is difficult to forecast integration through the use of the multinomial model, we construct an Integration Factors Index (IFI), which provides a benchmark for the furthering of integration (see Appendix B for details). The IFI reflects the average levels of value convergence, trust, and hierarchy at every stage of integration.7

Table 5.1 presents the IFI estimates and mean compositions for the time period between 2019 and 2021. As in our prior analysis, we assume that the UK is going to leave the EU in 2019. IFI values above 0.10 indicate movement towards further integration; 0.09–0.06 suggest stability at current levels; 0.05 and below suggest lower levels of integration. The estimates provide important but rather complex results. We begin from the first section where we hold the integration achievement score (IAS) to the level of a customs union. Even at this low level of integration, the UK will seek further separation from the EU as indicated by the low IFI values. This is reinforced at all three trust scenarios – low, normal, and high. Therefore, the UK is expected to seek a hard exit and to continue to distance itself from the EU.

Table 5.1: Impact of Brexit with Values, Trust Scenarios, and Hierarchy, 2019–2021.
Integration Achievement Score (IAS) Independent Variables Na Mean Integration Factors Index
Country: UK
UK’s Integration Level after Brexit (Customs Union) Value Convergence 3 1.3086252
Hierarchy 3 0.2947608
Low Trust 3 0.1461217 0.03
Normal Trust 3 0.1826510 0.04
High Trust 3 0.2191803 0.05
Countries: Poland, Sweden, Romania, the Czech Rep., Hungary, Denmark, Bulgaria, and Croatia
3.50 (Economic Union) Value Convergence 24 1.4995956
Hierarchy 24 0.2947608
Low Trust 24 0.3335608 0.07
Normal Trust 24 0.4165723 0.08
High Trust 24 0.4995837 0.10
Countries: France, Spain, Italy, Netherlands, Belgium, Austria, Greece, Portugal, Finland, Ireland, Slovakia, Slovenia, Latvia, Lithuania, Luxembourg, Estonia, Cyprus, and Malta
3.83 (Eurozone + partial financial integration) Value Convergence 54 1.0267418
Hierarchy 54 0.2947608
Low Trust 54 0.3135296 0.09
Normal Trust 54 0.3916334 0.11
High Trust 54 0.4697372 0.14
Note:a

Country-year.

We then set the IAS value to the level of the economic union (3.50) to see how the members listed are likely to press for further integration at varying levels of trust. At low or normal trust levels, members will likely stay at their current levels of integration. If trust increases, they may take steps towards further integration. The incentive to do so, however, is not high since the IFI value of 0.10 is a weak indicator of further integration. The last section of Table 5.1 increases the IAS to the EMU level (3.87). States that have low trust are likely to resist major increases in integration (i.e. fiscal union). However, if trust increases to normal or high levels, integration is likely to increase swiftly.

These results indicate that further divergence of a two-speed Europe is likely (see Chapter 3, Figure 3.1). The EU will be stable but will only inch forward to further integration if trust in the EU increases. Unlike current analysis that predicts other countries will follow the example of Brexit, we predict that no country currently in the Economic Union or the Eurozone will leave the EU anytime soon. In sum, the EU is fairly stable.

Finally, if we consider prior expectations derived from a British lack of trust in the EU and its movement away from value convergence with the core EU nations, a comparison of the results in Table 5.1 with similar assessments of the EU’s probability of deepening of integration without the UK, we find that integration will increase after Brexit. Furthermore, Brexit would increase Germany’s power asymmetry in the European hierarchy and improve values convergence among the remaining leading EU countries. This leads to improved prospects for further integration to rise concurrently.

Interaction between power transition and agent-based models

Power transition is an aggregate model that accounts for the structural preconditions for integration or conflict. Such aggregate models provide a long-term assessment of the likelihood of future stability but cannot provide short-term predictions or assess impending salient issues. Also, they cannot indicate the tactical short-term moves that can determine the direction agents (in this case nations) might take along the cooperation–conflict continuum. To close this gap we shift our focus from macro- to micro-assessments.

Power transition’s macro-analysis and the agent-based model’s (ABM) micro-results (decision theory, expected utility, and game theory) are connected by common assumptions. It is fundamental to note that micro-solutions are consistent with macro-expectations. Bueno de Mesquita and Lalman show that crisis escalates to conflict following the preditions of power transition.8 They show that conditions derived by the agent-based model and power transition are identical to the short-term tactical evaluations derived from an agent-based approach. Jacek Kugler independently shows that when additional variables used in the agent-based models are standardized, similar results to that of power transition emerge, suggesting cooperation among competitors when trust is high and confrontation when trust is absent.9 Simply stated, comparisons confirm that the main macro-insights of power transition can be derived at the micro-level that adds complexity to the interactions.10

Agent-based models

There are a number of ABMs used in the literature. The majority of them use stakeholder analysis to anticipate net gains agents expect to achieve from interactions with other agents. We apply two distinct computer programs to provide a picture of the current policy debates affecting European integration. giCompute and Senturion are used to assess the current and likely future development of integration in the EU (see Appendix C for descriptions of these programs).

We use the same data set, slightly transformed, for both ABM programs (see Appendix C). giCompute is an advanced decision analytics platform that assesses ranked preferences of states on selected issues to determine where they stand and what policy they will now support. This perspective incorporates the principles of game theory and decision analysis to calculate the positions leaders adopt and the impact of their actions on the overall outcome on any given issue. Senturion is similar to giCompute, but in addition has the capability to anticipate the outcome of issues for a period when stakeholders are consistent. Generally forecasts are reliable for a period between one and three years.11 The outcomes are predictions about policy outcomes based on current inputs. The model allows us to explore strategic opportunities the stakeholders could utilize.

In this analysis, we take into account the preferences of the major EU members and major external stakeholders on three salient issues identified in Chapter 3 that can affect the future of integration in the EU: fiscal union (political union), CFSP, and migration/refugees. The fourth issue, Brexit, is considered in each of the three assessments. Based on careful reading of the current literature, we define the specifics of the key issues, assess the influence of key stakeholders, and the importance they attach to each position on the issue, and then order preferences of member states and influential external actors across the three issues. The results are presented in Appendix C. The models then estimate potential outcomes across each issue to assess how integration is currently impacted – as we detail below – by migration and security.

Data

The inputs for the two agent-based models are the same. Outputs depend on accurate information obtained from face–to-face interviews with subject matter experts. Their inputs need to be supported and confirmed by information obtained from published sources. Four inputs are required:

Central issues of integration

The integration process is affected by salient policy differences over key issues. The process of integration takes place under competitive or cooperative conditions and involves stages that vary from a customs union, a common market,an economic union, monetary union (Eurozone), and fiscal and political union (including military matters). States satisfied with EU institutions wish to deepen integration, while those that are dissatisfied wish to move back or even abandon the process. Many are committed to the current status quo and do not support changes. To provide a forensic investigation of the socio-economic factors and a road map for the future of European integration, we concentrate on deepening of integration as affected by migration and security concerns.

For sure, EU members must face challenges created by entering a monetary union without a matching fiscal, military, and political integration. The introduction of the euro generated differences over the future direction of the EU. The main supporters of deepening integration, France and Germany, still have some differences over the direction of a more integrated EU. France supports debt mutualism and fiscal flexibility before committing to deeper integration and the creation of a unified security arrangement.12 In contrast, Germany prefers fiscal discipline and integration before considering mutualism. For Germany, a security arrangement cannot be effectively implemented without prior consensus on fiscal discipline.

An additional item that divides members is migration, which was prominently mentioned in the Brexit debate. Finally, the movement towards an integrated European military force, in response largely to threats from Russia and concerns about US policy shifts that affect the stability of NATO, impacts the willingness to integrate. Table 5.2 summarizes the options open to policy-makers on the level of integration, migration, and security.

Table 5.2: Policy Options across Integration, Migration, and Security.
Issue Policy Options
Integration Disintegration Customs Union Common Market Economic Union Eurozone Political Fiscal Union
Migration Fortress Europe Turn back refugees Help from afar only Safe zones outside Europe Resettle not in Europe Resettle in Europe with controls Accept all refugees with open arms
Security Neutral NATO NATO–EU partnership CFSP

Using giCompute, we assess the relative capabilities of each member state by the votes they hold in the EU. We then rank them to create the landscape of preferences for analysis. Russia, the US, and Turkey are added when they can directly affect integration. Each actor is assigned a value between 100 and −100 based on their preferences: 100 equals maximum disagreement on a particular issue and −100 equals maximum support on the same issue. Zero is the point at which the actor holds a neutral position on the issue. Figure 5.7 shows the outcome of our giCompute analysis. Overall, there is a lack of consensus among the parties about the future of the European Union. The most likely current path for the EU is the preservation of the Eurozone, but a path to deepening political integration is not feasible. Specifically, there is a 47.5 percent chance that the Eurozone will not lose members. Concurrently, there is a 40 percent chance that the EU will revert back to solely an economic union without the common currency. Further scaling down of integration is unlikely. The likelihood of reverting back solely to a customs union is 5 percent and to a common market is 7.5 percent.

Figure 5.7: Level of EU Integration Controlling for Current Migration and Security Policy.

Next we examined preferences of states controlling for anticipated results on the migration policy. To test the robustness of these results, we set the migration policy outcome to resettle migrants outside Europe, which reflects the results of the dynamic analysis to be presented later in this chapter. Consistent with prevailing assessments, the migration issue is an important factor in weakening European integration. Figure 5.8 displays the results. Our results show – counterintuitively – that once the migration issue is resolved, commitment to further integration actually diminishes. Note that there is some possibility of political/fiscal union, but it is very small at 5 percent. The survival of the Eurozone drops from 47.5 percent to 27.5 percent. Economic union also declines from 40 percent to 32.5 percent, while the chances of reverting to a common market increase 7.5 percent to 32.5 percent. Reversion to a customs union or complete disintegration is minimal at 2.5 percent. This counterintuitive result suggests that resolving the migration/refugee issue will not improve the chances for the deepening of integration.

Figure 5.8: Level of EU Integration Controlling for Anticipated Migration and Security Policy.

Note an interesting non-obvious implication if these results hold. The final result coincides with the original posture of Britain regarding migration. Such results challenge the notion that migration is the driver of Brexit. If EU members are converging on the policy the UK desires, a reversal of the exit policy should follow. We see no indication of such a reversal by the UK, suggesting that migration is not the driver of Brexit.

Finally we explore the likelihoods of deepening integration among members of both Schengen and the Eurozone. Such EU members should be the most committed to deepening integration in the EU. The results of this simulation are presented in Figure 5.9. As anticipated, this restricted set of states is far more committed to the EU. The chances of Schengen and Eurozone countries leaving the Eurozone is lower (moving down from a combined 65 percent to 30 percent). The likelihood of lowering the level of integration below the common market is nonexistent. Moreover, deepening of integration has a greater chance – 20 percent – of taking place among this restricted group than in the larger context. The Schengen and Eurozone group of states is not likely to dissolve.

Figure 5.9: Level of EU Integration Controlling for Anticipated Migration and Security Policy, Schengen and Eurozone Countries.

Dynamic analysis

To more directly address the challenges posed by the integration, migration, and security issues, we apply a dynamic ABM analysis using Senturion to anticipate the likely resolution of these issues in the near future.

Integration

The financial and fiscal crises of the past few years revealed the need for further fiscal union for monetary stability. Many economists indicate that the Eurozone requires further fiscal deepening in order to avoid a repetition of the economic crises, ranging from Greece to Ireland, that have weakened the EU in recent years.13 Consistent with analysis earlier in this chapter, the following scale is used to capture the integration continuum previously analyzed incorporating distance between policy options.

Figure 5.10a is the base case or the initial positions of the agents as provided by expert assessments. The height of each agent represents the weighted power that the stakeholder can exercise on the issue. The positions’ labels indicate the policy options (degree of integration) across the issue. The shading of each column represents distinct groups of states and each state is labeled. Figure 5.10b represents the anticipated outcome of the issue (degree of integration) resulting from the political interactions among the agents.

Figure 5.10: (a) Base Case Outcome of Integration Policy. (b) Outcome of Integration Policy (Round 5).

Comparing the two figures indicates no change in positions between the base case and the anticipated outcome. Commitments to current positions are fixed. Consistent with previous results, in the short term, given current policy commitments, further integration is unlikely. Germany, France, and Belgium will seek to move further towards a fiscal union – but are unlikely to succeed. All members of the Union will coalesce at the current Eurozone level. Poland, Croatia, Romania, and even Hungary, are expected to join the Eurozone despite their concerns over migration. On the other hand, Britain will likely emerge following Brexit with an agreement that falls between a customs union and a common market. This outcome will be closer to but not exactly a hard exit, thus making it a “soft exit” of a sort. Our static and dynamic ABM analyses indicate that the EU is robust despite a complex structure. States that are members of the Eurozone are the most likely to deepen integration. This observation is consistent with the previous findings reported in the macro- and micro-analyses. Movement towards a political union does not have much support in the foreseeable short term.

Brexit is the result of long-term policy dissatisfaction in the UK with the rules of the EU. Assessments in Figure 5.10b do not forecast other member states willing to follow the British example. Changes in government can shift policy preferences sufficiently and thereby raise up the issue of an EU exit. Our results here suggest, however, that they are not likely to diminish integration below the economic union level and this will most likely remain in place.

Migration

The alternative policies regarding migration were prompted by increased movement of people displaced mainly by the Syrian, Iraq, and Afghanistan crises and instability in Central and North Africa. This is not the largest or most consequential migration crisis the European continent has experienced.14 A substantive difference between past and current migrations is the large differences in values held by the Europeans and the new migrants.15 These concerns have prompted many to suggest that the migration crisis will weaken efforts at further integration. Recall that the migration crisis has caused some states to abandon their Schengen commitments and promoted arguments on how to best resolve the crisis. We explore the likely future path of the resolution of this issue in order to determine the validity of such claims.

Following previous methods, the policy options on migration available to European leaders are as follows and are given the following values:

Figures 5.11a and 5.11b plot out the outcomes of migration policy following the same procedures found in Figures 5.10a and 5.10b. The base case analysis in Figure 5.11a shows wider dispersion of policy preferences across the EU members. These are the disagreements that have recently caused such serious strife among European countries. The policy outcomes of the migration debate are detailed in Figure 5.11b. Key decision-makers are likely to coalesce as the debate is reduced to levels of support for new migrants outside Europe. EU member states are expected to adopt a common policy that supports safe zones outside Europe. This is a marked departure from the current status quo that allows migrants to resettle in Europe with controls.

Figure 5.11: (a) Base Case Outcome of Migration Policy. (b) Outcome of Migration Policy (Round 5).

While the migrant problem will not disappear, this analysis indicates that EU members will find a common migration policy position that will dissipate the crisis and that migration will become a far less salient issue. All EU nations are expected to converge on a policy that slows the influx of migrants into the EU. All will eventually favor different safe zones outside Europe. Hungary and Poland will lead dissent and continue to reject any additional immigration aid for migrants already in place or outside the Schengen region. Once more, the anticipated final outcome coincides with the original position held by Britain, again suggesting that Brexit was not solely or mainly motivated by differences between the EU and the UK on migration policy.

Security

Security concerns of European states are serious. For the first time since World War II, states covered by the NATO umbrella are not totally convinced that the US will come to their aid if needed. Warmer Russia–US relations combined with US support for Brexit, and loosening ties between Germany and the US, pose questions about the future security of Europe. Before these events took place, Genna and Justwan carried out an elaborate systematic analysis of European attitudes towards CFSP with data ending in 1990.16 This analysis indicated that positive support for a broader CFSP driven by Germany and France was sufficient to shift reliance on national security to collective security.17 The question now is whether conditions have changed sufficiently to alter the EU’s reluctance to generate its own security.

To address this challenge, we assess the current positions of the EU member states on security. The following are the possible positions of each state along the following scale:

The base case in Figure 5.12a shows substantial divisions within the EU membership over the future of European security. Austria, Ireland, and to a lesser degree Sweden, want to maintain neutrality. Britain, Denmark, Norway, and Iceland all advocate a strict reliance on NATO. The majority of members support a NATO and EU partnership. France, Germany, and Greece want to create a robust CFSP to supplement and even partly replace NATO.

Figure 5.12: (a) Base Case Outcome of Security Policy. (b) Outcome of Security Policy (Round 5).

Figure 5.13: Simulation: Impact of Russia Advocating EU Neutrality.

Figure 5.12b displays the outcome of the dynamic analysis. Major shifts are anticipated suggesting that despite the initial differences, all European states will likely respond to veiled US pressures by supporting a NATO framework while increasing the role of the EU in the Atlantic alliance. With the exception of Ireland and Austria – which remain neutral – Europeans wish to remain in NATO with increased European contribution to security.

Given the fluidity of this issue, the final analysis simulates the possible impact of Russia on EU security positions. Russia has suggested that NATO is obsolete and that the EU should choose neutrality to diminish the rising tensions. To test this proposition, we add Russia in a neutral position. The outcome of this simulation is reported in Figure 5.13.

Compared with the previous result, Russia can persuade Sweden to remain neutral. Britain, Iceland, Norway, and Denmark will now support NATO rather than join the rest of Europe in support of a NATO–EU partnership. Russia is an important competitor but will not drive Europe’s security decisions. We also attempted to simulate the outcome if NATO were to disintegrate. However, the data following such an event is so speculative that multiple scales can be constructed that provide little light on reality. However, given the coherence of the EU in the NATO analysis and under the Russian pressure, were NATO to falter we expect that most EU member states would coalesce into a CFSP rather than disperse as individual nations.

Conclusion

The European Union is stable with limited room for further integration. The migration challenges are serious but will not derail the Union. The most committed Eurozone–Schengen members of the EU are most likely to deepen integration further, reinforcing the two–track EU already in place. The greatest threat to European integration comes from declining trust in EU institutions. Reversing this trend is essential for integration to prosper and move towards a fiscal union.

Migration is a passing concern as EU member states will coalesce on a common policy that excludes migrants instead of supporting them in place. Migration will not soon disappear because migrants already in Europe cannot be simply removed. Moreover, demand for a larger workforce can help accommodate the new Europeans already in place.

Security is a new and rising concern. NATO dominates all opposition – limiting the likelihood of major war for the foreseeable future. The EU alone can stand up to pressure from Russia. To do so a buildup of weapon systems is needed to compensate for a potential US withdrawal from Europe. Such a move will require amending the EMU spending rules. Analysis suggests that while EU members favor a NATO partnership, they will also willingly join together for their common defense, implying that if the need arises, under German and French leadership, the creation of a viable CFSP would follow. Such a move implies a willingness to accept further fiscal and military integration – culminating in a federated state.

Appendix B

Overview of giCompute

giCompute is a hybrid conjoint analysis-reverse induction game theory model conceived for the purpose of quantitatively estimating outcomes of multi-factor and multi-agent issues in a transparent way. The platform helps transfer the qualitative judgment of analysts and practitioners involved in transnational disputes, armed conflict, legal disputes, and mergers to measure gains and losses relative to agent preferences across any set of defined options and scenarios. The objective is to compute and visualize the likely versus optimal outcomes for agents based on their relative preference attributes and influence pertaining to any particular dispute or negotiation. The platform computes all the possible relative gains and losses pertaining to any set of scenarios dependent on the number of preference attributes for agents across a defined scale involving a dispute or negotiation. The platform then transfers the calculated agent payoff for each scenario and visualizes multi-agent landscape preferences.

giCompute in application to negotiations or disputes applies a hybrid approach that adapts elements of a method used in market research known as conjoint analysis and integrates it with the game-theory principles of status quo positioning. Conjoint analysis is a statistical technique used to determine how consumers value different attributes of a product or service. The process estimates the overall utility associated with each component based on the rank-ordered preference on components of a product by the consumer or market segment. The utility values identify the relative value consumers place on a product characteristic. This information then allows one to develop a net utility maximized product, isolating the more desirable product based on consumer references. giCompute calculates full factorial design using pure ordinal ranking of component preference. The giCompute approach calculates relative aggregate agent payoffs based on the number of components defining the issue and then calculates the payoffs based on agent preferences on the various policy issues.

Subject matter experts provide filtered information on issues deemed central to EU integration. The scales for each issue – Integration, Migration, and Security - are constructed. Then the following data is collected.

  1. Who are the national agents that have potential to influence the outcome?
  2. What is the relative potential influence of each group, and then within each group, what is the relative influence of each agent belonging to a group?
  3. What is the stated position that each agent holds?
  4. How important is the issue to each agent given all other issues in the nation’s agenda?

giCompute uses the same inputs to define agents, and measures of relative potential influence of groups and agents. The position of agents on the scale is used to create the rank variable. Each actor ranks highest the position closest to the initial stated agent position. Ranked second is the closest defined policy. Ranked third is the next closest policy and so on. Distance from the initial policy left or right position determines rank.

Unlike Senturion, which interacts influence with importance, giCompute uses influence in this analysis, measured here by the voting strength of EU members.18

Appendix C

Overview of Senturion

Senturion has been used for a variety of cases ranging from complex diplomatic and business negotiations to anticipation and evolution of conflicts. Given all the components in the data, Senturion simulates political dynamics and future interactions to predict the outcome of negotiations. The model uses Black’s median voter theory, which shows the winning position in a pairwise comparison of all alternatives is the median. Distance from the median defines an agent’s risk profile – those close to the median are risk averse and those farther from the median are risk acceptant.19 An agent attempts to achieve its goal subject to risk tradeoffs estimated in a game theoretical context.20 Using such information the model maps potential decisions by calculating pairwise interactions for each dyad colored by risk perceptions. Given the median outcome, how will agents attempt to change other actors’ positions and to what degree will they succeed? The model reports individual actor moves and their influence on the overall median. Iterations stop when agents see no potential to influence others. Stability emerges when a winning coalition coalesces around a median outcome – the smaller the dispersion the more stable the outcome. Dissent will take place when coalitions form but do not coalesce. Confrontations take place when the distance between such coalitions is large. Predictions are reliable for the duration of policy control by current governments or if the replacement does not shift current positions. This analysis applies to all political disputes where agents have a say in the outcome, but does not apply to market-driven events determined by the pricing of goods.

Senturion assumes that agents hold rational preferences that are (1) complete – all choices are disclosed; (2) ordered – all choices can be ranked; and (3) transitive – consistent ranking of choices from best to worst holds. Applications are only valid to single picked monotonic and symmetrically declining utility preferences away from the preferred positon of an actor. The decision-making landscape is built from data on policy positions by competing relevant actors whose commitment to their position is affected by influence weighted by the importance they attach to the outcome of the contested issue. Senturion anticipates future behavior, repeatedly analyzing changing structures resulting from accepted proposals across numerous actors, all of which seek to maximize their net gains. Perceptions are created by considering the gains and losses of not acting at all compared with the anticipated outcomes of intervening conditioned on expectations that third parties will join, oppose, or remain neutral. Each agent perceives potential payoffs differently based on their risk propensity and the importance attached to the issue. Change is driven in part by real opportunities, and also by unseen opportunities generated by differences in risk and importance across agents. The Senturion agent-based model performed with the highest level of accuracy and granularity, achieving an accuracy of more than 90 percent for baseline projections in policy analysis.21

Data

Issue 1

European Security 0 — Neutral

         50 — NATO

         75 — NATO

         100 — CFSP

Agent Group Proxy Veto Position Influence Group Influence Importance
Austria* Eurozone No No 0 10 275 90
Belgium Eurozone No No 80 12 275 90
Cyprus* Eurozone No No 75 4 275 80
Estonia Eurozone No No 75 4 275 80
Finland Eurozone No No 75 7 275 85
France Eurozone No No 100 29 275 90
Germany Eurozone No No 95 29 275 90
Greece Eurozone No No 90 12 275 80
Ireland* Eurozone No No 0 7 275 90
Italy Eurozone No No 75 29 275 85
Latvia Eurozone No No 75 4 275 80
Lithuania Eurozone No No 75 7 275 80
Luxembourg Eurozone No No 85 4 275 85
Malta Eurozone No No 100 3 275 85
Netherlands Eurozone No No 75 13 275 90
Portugal Eurozone No No 75 12 275 85
Slovenia Eurozone No No 75 4 275 85
Slovakia Eurozone No No 75 7 275 85
Spain Eurozone No No 80 27 275 85
Bulgaria Eurozone No No 75 10 275 80
Czech Eurozone No No 75 12 275 90
UK EU No No 50 29 275 95
Denmark EU No No 50 7 77 90
Hungary EU No No 75 12 77 85
Poland EU No No 80 27 77 85
Romania EU No No 75 14 77 85
Sweden* EU No No 25 10 77 90
Croatia* EU No No 75 7 77 85
Norway EEA No No 50 100 40 90
Iceland EEA No No 50 10 40 90
Liechtenstein* EEA No No 75 2 40 80
Switzerland* EEA No No 0 75 40 95
Turkey CU No No 75 100 20 70
US US No No 75 100 400 90
*

Non-NATO countries.

Issue 2

Integration-Disintegration of EU 0 — Disintegration

               15 — Free Trade Area

               25 — Custom Union

               50 — Common Market

               65 — Economic Union

               75 — Economic Union & Eurozone

               100 — Political Union: Economic and Fiscal Union

Agent Group Client/Proxy Veto Position Influence Group Influence Importance
Austria Eurozone No No 75 10 275 80
Belgium Eurozone No No 100 12 275 90
Cyprus Eurozone No No 85 4 275 60
Estonia Eurozone No No 75 4 275 85
Finland Eurozone No No 85 7 275 90
France Eurozone No No 85 29 275 90
Germany Eurozone No No 85 29 275 90
Greece Eurozone No No 65 12 275 80
Ireland Eurozone No No 75 7 275 90
Italy Eurozone No No 75 29 275 85
Latvia Eurozone No No 75 4 275 85
Lithuania Eurozone No No 75 7 275 85
Luxembourg Eurozone No No 100 4 275 90
Malta Eurozone No No 75 3 275 90
Netherlands Eurozone No No 75 13 275 90
Portugal Eurozone No No 75 12 275 85
Slovenia Eurozone No No 75 4 275 75
Slovakia Eurozone No No 75 7 275 90
Spain Eurozone No No 75 27 275 85
UK Eurozone No No 35 29 275 85
Bulgaria Eurozone No No 75 10 275 75
Czech Eurozone No No 75 12 275 90
Denmark EU No No 65 7 77 90
Hungary EU No No 75 12 77 75
Poland EU No No 75 27 77 80
Romania EU No No 75 14 77 75
Sweden EU No No 65 10 77 90
Croatia EU No No 75 7 77 75
Norway EEA No No 50 100 40 90
Iceland EEA No No 50 10 40 90
Liechtenstein EEA No No 50 2 40 75
Switzerland EEA No No 50 75 40 90
Turkey CU No No 65 100 20 85
US US No No 75 100 400 1
Issue 3

Immigration to EU 0 — Fortress Europe

         20 — Turn Back Refugees

         40 — Help From Afar Only

         50 — Safe Zones Outside Europe

         70 — Resettle Not in Europe

         80 — Current SQ: Resettle in Europe with Controls

         100 — Accept all Refugees with Open Arms

Agent Group Client/Proxy Veto Position Influence Group Influence Importance
Austria Eurozone No No 70 10 275 50
Belgium Eurozone No No 60 12 275 70
Cyprus Eurozone No No 20 4 275 80
Estonia Eurozone No No 55 4 275 40
Finland Eurozone No No 55 7 275 40
France Eurozone No No 80 29 275 85
Germany Eurozone No No 80 29 275 70
Greece Eurozone No No 80 12 275 60
Ireland Eurozone No No 80 7 275 40
Italy Eurozone No No 60 29 275 80
Latvia Eurozone No No 55 4 275 40
Lithuania Eurozone No No 55 7 275 40
Luxembourg Eurozone No No 60 4 275 70
Malta Eurozone No No 20 3 275 80
Netherlands Eurozone No No 65 13 275 60
Portugal Eurozone No No 80 12 275 50
Slovenia Eurozone No No 20 4 275 80
Slovakia Eurozone No No 20 7 275 90
Spain Eurozone No No 70 27 275 60
Bulgaria Eurozone No No 25 10 275 75
Czech Eurozone No No 20 12 275 80
UK EU No No 50 29 275 85
Denmark EU No No 40 7 77 80
Hungary EU No No 5 12 77 90
Poland EU No No 10 27 77 65
Romania EU No No 75 14 77 40
Sweden EU No No 35 10 77 40
Croatia EU No No 20 7 77 80
Norway EEA No No 60 100 40 40
Iceland EEA No No 40 10 40 30
Liechtenstein EEA No No 40 2 40 50
Switzerland EEA No No 40 75 40 80
Turkey CU No No 70 100 20 80
US US No No 40 100 400 20

Notes

1Allison Hamlin and Jacek Kugler, in Regional Politics, edited by Jacek Kugler and Ronald Tammen (forthcoming). In this work satisfaction with the status quo is measured by the transfer of military resources to coalition members. Large transfers indicate support while counter transfers indicate disagreement. Tests reported in this work show strong results indicating that nations that confront each other and are dissatisfied with the status quo may wage war at parity. This empirical argument rejects the entrenched notion that a balance of power preserves peace.

2Power transition shows that conflicts over dominance of the global system are overwhelmingly waged under conditions of parity and dissatisfaction. Contenders escalate to global conflicts over very valuable population and output. Previous global encounters were waged over control of Eurasia because more than 50 percent of global population and GDP were and still are at stake. The winner emerges as the dominant global nation. Peripheral disputes that lead to severe but asymmetric limited conflicts – Korea, Vietnam, Afghanistan, and Iran among the most recent – are resolved with far less loss of life and treasure by the great powers involved, even though they may be devastating for the smaller main participants. A. F. K. Organski and Jacek Kugler, The War Ledger (Chicago: Chicago University Press, 1980); Jacek Kugler and Douglas Lemke, Parity and War (Ann Arbor: University of Michigan Press, 1996); Ronald Tammen et al., Power Transitions (New York: Chatham House, 2000).

3Yelena Tuzova and Faryal Qayum, “Global Oil Glut and its Impact on Putin’s Russia,” Energy Policy 90 (2015): 140–151.

4For example, see the TACIS and PHARE programs.

5A nuclear confrontation would undoubtedly involve France, Britain, and the US. Given such a contest China would also likely join. See KyungKook Kang and Jacek Kugler, “Nuclear Weapons: Stability of Terror,” in Debating a Post-American World: What Lies Ahead?, ed. by S. Clark and S. Hoque (New York: Routledge, 2011).

6We apply the formulation provided by Brian Efird, Jacek Kugler, and Gaspare M. Genna, “From War to Integration: Generalizing Power Transition Theory,” International Interactions 29 (2003): 293–313, which produces very stable results across all types of conflict. In the latest Conflict and Cooperation estimation, we calculated the Hierarchy of the Challenger and the Hierarchy of the Dominant consistent with the work here that differs somewhat from Efird et al. The model used is as follows:

  • CC = Conflict − Cooperation. Continuum ranges from 0 to 10
  • $$ when (GDPi > GDPj)
  • Vc = Value Convergence
  • HC = Hierarchy of the Challenger
  • HD = Hierarchy of the Dominant

7Integration Factors Index (IFI) is derived as an interaction between the means as follows:

The lower the values distance between countries, the higher the level of integration. In the analysis, a decreasing mean of value convergence increases the IFI; an increasing mean of trust raises the level of integration; and an increasing mean of hierarchy increases integration. The resulting IFI measures the impact of all drivers of integration. For more details, see Appendix B.

8Bruce Bueno de Mesquita and David Lalman, War and Reason: Domestic and International Imperatives (New Haven: Yale University Press, 1992).

9Jacek Kugler, “Terror without Deterrence: Reassessing the Role of Nuclear Weapons,” Journal of Conflict Resolution 28, no. 3 (1984): 470–506. This argument is extended in Jacek Kugler and Suzanne Werner, “Conditional Anarchy: The Constraining Power of the Status Quo,” Paper presented at the annual meeting of the Midwest Political Science Association in Chicago, April 1993.

10Macro-assessments cannot simply replicate micro-evaluations. The reason is that risk and intensity that affect micro-assessments are associated with specific leaders and cannot be forecast into the future. Likewise, accurate assessment of tactical short-term decisions are muted on the long-term implications that macro-models address directly.

11Senturion provides a description of current conditions and policy assessments based on the nature of interactions among stakeholders and the types of coalitions that will form. The mechanics of this agent-based modeling are based in part on Black’s median voter theorem that the median voter controls outcomes in majority win situations and assumes monotonicity of scales used to assess expectations and the escalation of political disputes (Duncan Black, “On the Rationale of Group Decision-making,” Journal of Political Economy 56, (1948): 23–34; Duncan Black, The Theory of Committees and Elections (Cambridge, UK: Cambridge University Press, 1958); Jeffrey S. Banks, “Equilibrium Behavior in Crisis Bargaining Games,” American Journal of Political Science 34 (1990): 599–614. The models’ predictions are based on the assumption that national leaders make rational decisions that optimize the political desired net gains of the parties. Decisions are not extensions of past behavior but calculations based on strategic interactions. The decision analytics platform uses minimal information and ranks the policy preferences to obtain the probability that agents in the EU will choose a position along the integration scale. Assessments are verified by altering marginal positions to determine if integration is deepened or reversed motivated by variations in either the migration or the security issues. These ABM models facilitate a mapping of the relationships based on stakeholder’s individual perceptions of reality vis-à-vis every other stakeholder who observing the same payoffs may assess outcomes differently because of their risk and salience levels. The dynamic ABM model employed incorporates importance to the calculations and evaluates risk propensity. The dynamic ABM model simulates the anticipated shifts in position of individual stakeholders in response to pressure from competitors during implicit or explicit bargaining. The results report changing rounds in detail that produce a likely final output. All actors can simultaneously make proposals and exert influence on one another during each round but only viable proposals are accepted. Stakeholders evaluate options and build coalitions by shifting positions on the issue that is deemed to be critical in each situation. The policy proposals and the subsequent responses begin to give insights into the short-term negotiation process, anticipating dynamics and outcomes that in our case deal with the future of EU integration. The agent-based modeling forecasts the likely settlement of specified policy issues as a function of negotiations in and environment where competition, confrontation, and cooperation are dictated by the distribution of power of stakeholders, the importance they attribute to the issue, and the risk they are willing to accept to achieve goals.

12Markus Konrad Brunnermeier, Harold James, and Jean-Pierre Landau, The Euro and the Battle of Ideas (Princeton: Princeton University Press, 2016).

13Brunnermeier et al., The Euro and the Battle of Ideas.

14The migrations following World War II were massively larger. See Tadeusz Kugler, Kyungkook Kang, Jacek Kugler, M. Arbetman, and John Thomas, “Demographic and Economic Consequences of Conflict,” International Studies Quarterly 57. (2013): 1–12; Tadeusz Kugler, “Demography and International Relations: Politics, Economics, Sociology and Public Health,” in Interdisciplinary Approaches to International Studies, ed. by Patrick James, Steve Yetiv, and Stefan Fritsch (London: Palgrave Macmillan, 2016).

15For some works in this area, see Malina Voicu, Ingvill C. Mochmann, and Hermann Dulmer (eds.), Values, Economic Crisis and Democracy (London: Routledge, 2016); Richard C. Jones (ed.), Immigrants Outside Megalopolis (Lanham, M.D.: Lexington Books, 2008); Nils Holtug, Kasper Lippert-Rasmussen, and Sune Lægaard, (eds.), Nationalism and Multiculturalism in a World of Immigration (New York: Macmillan, 2009).

16Gaspare Genna and Florian Justwan, “The European Power Hierarchy, Member State Trust, and Public Support for the Common Security and Defense Policy,” Pan-European Conference on the European Union, European Consortium for Political Research (Trento, Italy, 2016).

17Ibid.

18Replication shows that changing these inputs does not substantively affect results for the EU.

19Duncan Black, “On the Rationale of Group Decision-making,” Journal of Political Economy 56 (1948): 23–34; John W. Pratt. “Risk Aversion in the Small and in the Large,” Econometrica 32, no. 1/2 (1964): 122–136; Arrow, Kenneth, “The Theory of Risk Aversion,” in Helsinki Aspects of the Theory of Risk Bearing, ed. by Yrjo Jahnssonin Saatio (Helsinki: Yrjo Jahnssonin Saatio, 1965).

20David Lalman, “Conflict Resolution and Peace,” American Journal of Political Science 32, no. 3 (1987): 590–615.

21For a technical assessment see various contributions in Jacek Kugler and Yi Feng (eds.), “The Expected Utility Approach to Policy Decision Making,” International Interactions 23 (1997); Mark Abdollahian and Carole Alsharabati, “Modeling the Strategic Effects of Risk and Perceptions in Linkage Politics,” Rationality and Society 15, no. 1 (2003); Mark Abdollahian, Michael Baranick, Brian Efird, and Jacek Kugler, Senturion: Predictive Political Simulation Model, Defense and Technology Paper 32 (Washington, D.C.: Center for Technology and National Security Policy, 2006).