The capital flight quadrilemma: Democratic trade-offs and international investment

ABSTRACT This article argues that capital flight of real investment presents governments with a quadrilemma. First, governments can tailor their policies to attract investors – but this is incompatible with a whole range of alternative policy choices. Second, they can simply accept capital flight – but this is incompatible with a robust capital stock and tax base. Third, they can harmonize its taxes and regulations with other states – but this is incompatible with international independence. Fourth, they can impose capital controls – but this is incompatible with international capital mobility. These incompatibilities make up four different goals, the value of which are described. Strategies may be mixed, but the pursuit of any three goals must always come at the expense of the fourth.

choose some mixture of four ideal-typical strategies, each of which satisfies three relevant goals but runs directly counter to a fourth goal (see Table 1 below).
First, a state can retain capital mobility, international independence, and a robust capital base by tailoring its policies to attract investors -but this precludes a whole range of alternative policy choices. Second, a state can retain policies investors dislike, alongside capital mobility and international independence, by simply accepting the erosion of its capital base -but then they cannot cultivate the capital stock and tax base. Third, a state can retain mobility, capital base and investor-limiting policies by harmonizing its taxes and regulations with other states -but then they cannot be international independent. Finally, a state can retain independence, capital base and investor-limiting policies by putting up barriers to capital flows -but this means restricting international capital mobility. Pursuing any three of these goals must come at the cost of the fourth goal.
The quadrilemma implies that there can be no simple solution to the problem of real capital flight. For individual countries, the normative question is how to weigh the different goals at stake. This article is devoted to elucidating the structure of this normative choice, and although it is informed by research on how states and investors have behaved in the recent past, it does not present any new empirical findings.
After some brief background, the four main sections of the article consider each strategy in its ideal-typical form, alongside a discussion of the goal with which the strategy is incompatible. These goals are not supposed be fundamental moral values in themselves. Rather, the goals are merely supposed to be valuable, and I discuss which deeper values they serve. These sections have a similar structure: first they describe the strategy, then the contradictory goal and why states should be able to pursue that goal, and then consider other objections, either to the strategy itself or to its alleged tension with the relevant goal. The final section then moves from ideal-typical strategies to consider more mixed approaches, summing up the quadrilemma in a graphical representation.

Background
Before describing the quadrilemma in detail, I will briefly situate my thesis in the existing literature and define the scope of my inquiry more precisely.
The questions I tackle have been raised by two bodies of literature. The general phenomenon of private investment influencing democratic decision-making has attracted attention under a range of names. Michał Kalecki (1943) wrote about the  Lindblom (1977Lindblom ( , 1982 the 'privileged position of business, ' Fred Block (1977) (and following him Joshua Cohen (1989)) 'structural constraints, ' Przeworski and Wallerstein (1988) 'structural dependence,' Brian Barry (2002) 'capitalists having power over the government,' Thomas Christiano (2010a) the 'uneasy relationship between democracy and capital,' Stuart White (2011) 'capital strikes', andWolfgang Streeck (2017) 'investment strikes' -terms that also occur in journalistic discussions ('A Capital Strike', 2009;Burke 2012;Frase 2011). In addition, there is growing body of work on 'tax competition': strategic tax-setting by states to attract internationally mobile capital. Many empirical studies in economics and political science have assessed the prevalence of the phenomena. More recently, tax competition has also been assessed in normative terms, most notably by Peter Dietsch and Thomas Rixen (Van Apeldoorn 2018; See also Avi-Yonah 2000;Dietsch 2015;Dietsch andRixen 2014, 2016;Ronzoni 2009Ronzoni , 2014. Like the tax competition literature, but unlike the literature on capital strikes, I focus specifically on the international context of capital flight. Almost nowhere today has anything approaching a closed economy, and, as I will argue below, closed economies can largely avoid capital strikes in any case. Unlike the tax competition literature (and to a lesser extent the capital strikes literature), I do not focus solely on taxes. Taxes are merely the most obvious policies states can change to attract investors.
The author whose works comes closest to my argument is Dani Rodrik (2012). In Rodrik's 'political trilemma of the world economy', the incompatible goals are 'hyperglobalisation', 'democratic politics' and 'the nation state'. Rodrik's thesis is broader than mine, referring not only to competition for real investment but to the whole spectrum of international economics, including trade, exchange rates and multilateral organizations. This greater scope understandably comes at the cost of precision regarding particular issues. Applied to real capital flight, Rodrik's three-option model neglects a fourth option: simply accepting capital base erosion.
My topic is defined more precisely in several ways. I focus on real investments (FDI), rather than 'hot money' flows of liquid assets, or 'virtual capital flight' of accounting profits and portfolio capital. Hot money flows are important primarily for exchange rates. They give rise to a well-established 'monetary policy trilemma' ('impossible trinity') that is distinct from the quadrilemma described here (though with some points of analogy). Virtual capital flight concerns where profits and assets are located for accounting purposes, as distinct from where the underlying real economic activity is located. Virtual capital flight raises different normative issues (focusing on how rights to tax economic activities should be allocated), and different practical issues (focusing on how international co-operation to tackle the problem can be structured) (Dietsch and Rixen 2014;Pogge and Mehta 2016).
Real capital flight presents a quadrilemmatic trade-off: four potentially desirable goals, only three of which can be fully reached at any point in time. This is a quadrilemma in the same sense as Rodrik's trilemma and the monetary policy trilemma. Quad/trilemmas of this sort allow for compromise strategies: a state could mix a little of all four strategies to achieve something of all four goals. However, the possibility of such compromises does not invalidate the quadrilemmatic structure of the choice situation. What makes this structure distinct from trade-offs in general is the particular structure in which three goals can be maximized, but never all four. This implies that attempting to improve on one of the goals must come at the expense of at least one of the other goals. Graphical representations are introduced below to make this clearer.
The quadrilemma thesis is premised on an assumption that investors seek to maximize their returns. There are two ways we can imagine this assumption being violated. On the one hand, we can imagine that investors might behave more aggressively towards governments. In the slump of 1937, Franklin Roosevelt suspected a capital strike in a strong sense when he thought investors were conspiring to deliberately wreck the economy in order to hurt him politically (Kennedy 1999, 352). The moral issues at stake in a deliberate, coordinated capital strike of this sort would be relatively clear. Capital flight driven purely by profit-seeking, on the other hand, is both a morally more complex as well as empirically much more common phenomena. I thus refer to capital flight solely in the sense of normal, uncoordinated profit-seeking, any political effects of which are by-products.
On the other hand, we can also imagine that investors could be more deferential towards government, refraining from shopping around for favourable regulatory regimes. At the limit, if investors completely ignored international differences in rates of return due to different taxes and regulations, the quadrilemma would cease to hold. Although investor restraint can theoretically mitigate the capital flight quadrilemma, this is not treated as a significant variable in the empirical literature on topic.
Normatively, the investor self-interest assumption demonstrates the need to distinguish carefully between the question of what investors should do in response to government policies, and the question of what governments should do. 1 The latter question is part of nonideal theory in the sense that instead of asking what all actors should do, it asks what some actors (states) should do, given empirically plausible assumptions about how other actors (investors) are likely to respond. 2 The question for states is whether the benefits of policies outweigh the costs, given that the distribution of those costs will be affected by investors attempting to pass costs on to others. If the state should avoid policies because of adverse reactions by private investors, we have a basis for saying that private investment impedes a democracy's ability to choose its policy goals. Assessing this claim is a necessary prerequisite for evaluating socialist arguments that we need to socialize investment decisions to protect democratic decision-making (Cohen 1989;White 2011). Now that we have a clearer idea of our subject, I now proceed to develop my own framework, describing each of the ideal-typical strategies and goals which make up the quadrilemma.

Investor-friendly policy or investor-limiting policy
I refer to the first ideal-typical strategy as investor-friendly policy: taxes and regulations that are conducive to the interests of investors and so attract/retain investment in the country. As I define them, investor-friendly policies prioritize gains to relatively mobile factors of production (mobile capital) over relatively immobile factors (labour, land, immobile capital). By pursuing such policies, states can satisfy three of the four goals of the quadrilemma: they can cultivate their capital base while retaining capital mobility, and without loss of international independence.
It is important to distinguish two different reasons countries might choose investorfriendly policies. In the case of deliberate regulatory competition, investor-friendly policies are adopted with the specific goal of attracting investment from other countries. This is exemplified by Ireland, which cut its corporate tax rate to 10% with the explicit goal of attracting foreign direct investment (Dietsch and Rixen 2014, 161).
However, we cannot assume that all investor-friendly policies are instances of regulatory competition. Some countries might choose a smaller state just because they judge that kind of society to be better, independently of any competitive advantages visà-vis other countries. We can distinguish these cases using a counterfactual test (Dietsch and Rixen 2014, 164): would a country have chosen the policy in question even if there was no prospect of capital flight? In this case, any capital inflow is just a fortuitous side-effect of the policy. These are cases of genuine regulatory disagreement rather than competition.
Regulatory competition by multiple countries has a distinctive dynamic often described as a 'race to the bottom'. This is illustrated by the case of taxation in the simple 'baseline model': two identical countries sharing a common international tax base, each deciding what tax rate to choose (Genschel and Schwarz 2013; The model originated in Zodrow and Mieszkowski 1986). The first country has an incentive to cut tax rates in order to increase tax revenues by attracting some of the tax base of the second country. But the second country can easily win back that tax base by cutting its tax rates too. The two countries are left with the same share of the tax base they began with, only now with a lower level of taxation. The countries are in a prisoners' dilemma situation that pushes tax rates and revenues in both countries below the level they would otherwise choose.
One important assumption in this simple model is that the two countries have economies of similar size. Countries sufficiently smaller than their neighbours can actually increase their net revenues by becoming tax havens. This is because for smaller countries, the increase in the tax base has a much greater effect than the decrease in the tax rate. However, these benefits for small tax havens only occur at the cost of greater losses for their larger neighbours. Many empirical studies have attempted to test these predictions and those based on more sophisticated models. Put crudely, the consensus is that tax competition has occurred, but not to the extent of the full-on 'race to the bottom' predicted by the baseline model. 3 Some possible reasons for this will be discussed below.
The strategy of investor-friendly policy by definition runs against policies that discourage investment. I use the term investor-limiting policy as a shorthand for these contradictory policies. The most obvious element of investor-limiting policy is the taxation of investors, and most discussions (both empirical and normative) reduce regulatory competition to tax competition. 4 However, other policies such as labour rights or environmental standards can also limit the attractiveness of a country for foreign investors.
As a goal, investor-limiting policy is frustrated by investor-friendly policy but can be achieved by the three other ideal-typical strategies of the quadrilemma: by capital barriers, by international harmonization, and by accepting capital flight.
Including investor-limiting policy as one of the four goals at stake means that something of value is sacrificed when a country chooses to abandon investor-limiting policy in order to better attain the other three goals of the quadrilemma. This claim can be justified in two ways: by appeal to justice, or by appeal to democracy.
I will take the latter approach, although the former is worth briefly discussing. The appeal to substantive conceptions of justice leads to the familiar left/right divide on economic policy. Investor-limiting policy is advocated by those with more egalitarian conceptions of justice and/or more optimistic views about state capacity, and criticized by those with more libertarian conceptions of justice and/or more pessimistic views about state capacity. Consequently, depending on one's starting point capital flight can be either lamented (Brock 2008;Cappelen 2001) or celebrated (Hardin 2009) for its tendency to promote investor-friendly policy. The quadrilemma is a difficult choice for those on the left who see value in investor-limiting policy. However, insofar as those on the right deny that there are costs of investor-friendly policy, this strategy dominates the others and the quadrilemma disappears.
Rather than relying on a substantive conception of justice to ground the claim that investor-limiting policy is valuable, I will instead focus on a different normative level, that of democracy. 5 My normative premise is that it is valuable for a government's policy to reflect the preferences of its people. 6 On this level, there is nothing regrettable in itself about a government pursuing investor-friendly policy out of what I called genuine regulatory disagreement. But there is something regrettable about a government forced to abandon investor-limiting policy because that is the only way it can protect its tax base without compromising capital mobility or international independence. What makes this case regrettable is that citizens' first preference would have been to combine investor-limiting policy with the other three goals of the trilemma, and they must instead accept investor-friendly policy as their second preference. Investor-friendly policy is a valuable goal because it is goal that democracies may and indeed often have chosen to pursue.
It might be objected that the fact that investor-friendly policies lead to capital flight is not democratically regrettable, because it merely reveals the genuine costs of those policies. For example, suppose a government requires stricter environmental standards on producing fridges. Consequently, fewer fridges will be produced and investment in the industry will fall. But there is nothing sinister here: just a market efficiently signalling the costs of a different production process.
In reply, we can concede that investor responses to investor-limiting policies like environmental standards can often be characterized as efficient signals of the real costs of policies, in a closed economy. However, this objection does not work in an open, international economy. In the context of international capital flight, investor responses are often inefficient and do not signal the genuine costs of policies. When choosing between two investment opportunities, so long as both are under the same tax and regulatory system, investors will choose the opportunity with the lowest costs and greatest benefits. However, when regulatory systems differ, investors will choose what would otherwise be less productive investments merely because taxes in that country are lower (Avi-Yonah 2000, pp. 1604-1616. Moreover, capital flight can put pressure not only on redistributive policies, but also on policies (like environmental standards) designed to remedy market failures (Sinn 2003).
Shifting from a substantive conception of justice to a theory of democracy does not shield my normative premise from all controversy. Defenders of regulatory competition may say there is nothing regrettable about democracies being deterred from what they say are unjust or unwise policies. I will not enter into this debate here. The democratic value of being able to pursue investor-limiting policy is one of my premises, not one of my conclusions. Leaving aside more general questions about the relationship between justice and democracy, 7 I assume it is valuable for democracy to hold sway on topics such as tax rates, labour rights, environmental protections and infrastructure investments -areas that are constitutive of investor-liming policy.
I now turn to a different kind of concern, which can also be reconstructed as an objection to the alleged value of investor-limiting policy. I have been writing as though 'investor-friendly policy' could be glossed as something like 'neo-liberalism'. But it may turn out that what actually attracts real investment is more like the welfare state: perhaps investors are attracted to countries with good infrastructure and healthy, welleducated workers. Regulatory competition would then lead to a race to the top rather than a race to the bottom. This has been one of the main lines of criticism against the baseline model. 8 If the policies states might want to adopt for the benefit of noninvestors are actually consistent with a robust capital base, investor-limiting policy is not a valuable goal at all. Investor-friendly policy would dominate the other three idealtypical strategies. Instead of an agonizing trade-off, we would have a simple solution.
One way of putting this objection to the test is by re-purposing the counterfactual test employed above: to what extent do policies that attract international investment overlap with the policies that would be chosen in a closed economy? Put this way, there is clearly a lot of room for overlap, but the overlap is never likely to be complete. When it comes to public goods, provision that optimizes for a single goal (workforce productivity) will not perfectly coincide with provision that optimizes for a blend of additional goals (humanistic education or biodiversity, for example). When it comes to redistribution, the divergence is even starker. Perhaps there are some degrees of redistribution from investors to non-investors where investors actually benefit, because workers become more productive and consumers more spendthrift (Przeworski and Wallerstein 1988, 27). But this is surely the exceptional case: normally, redistribution from investors to others does not benefit investors.

Accepting base erosion or a robust capital base
The second ideal-typical strategy available is to accept capital flight and continue as before. By doing so a country can allow capital mobility, pursue investor-limiting policies, and remain internationally independent. However, by its nature this strategy goes against a concern for a country's capital base, whether this means attracting new investment or retaining existing capital in the country. Retaining and cultivating the capital base, on the other hand, can be achieved through the other three ideal-typical strategies in the quadrilemma: investor-friendly policy, capital barriers, and international harmonization.
A country's capital base is worth cultivating for two reasons. First, there is a benefit to economic welfare from using the capital itself. Capital investments raise productivity, increasing income or making it possible to expand leisure time without loss of material prosperity. Second, capital can be taxed. A broad tax base is instrumentally valuable for public spending. Public spending itself can be evaluated in the same ways as investorlimiting policies: either on the basis of a substantive conception of justice, or on the basis of democratic choice.
Accepting base erosion is an easy option to overlook. As the default option, it is rarely considered as a conscious strategy. Omitting it as an option suggests an unwarranted credulity towards claims about the capital-base costs of investorlimiting policies. In many cases, the impact of investor-limiting policies on the capital base will be relatively minor, and so it will make more sense to accept this minor impact rather than revising the whole policy. When investors or their political allies claim that we must endorse their preferred vision of society in order to keep up with the global race, the correct response in many cases is surely to echo Mandy Rice-Davies : 'well, he would say that, wouldn't he?' For example, in their analysis of the Australian mining case, Stephen Bell and Andrew Hindmoor (2014) conclude that the impact of potential capital flight was fairly insignificant in purely economic terms. The reason Labour politicians abandoned the tax was not because they really believed it would damage the economy, Rather, they believed that voters believed it would, and the politicians did not want voters to regard them as incompetent or irresponsible. Moreover, both the public's perception of the risk posed and politicians' perception of the public's perception were strongly influenced by the spirited public relations campaign of the mining industry.
The possibility of simply accepting base erosion is also integral to economic models of tax competition. Countries in the baseline model maximize their tax revenues by setting a tax rate that optimally trades off between two effects of (higher) tax rates: revenue gained because of the higher rate itself (the investor-friendly/limiting policy dimension) and revenue lost because of a smaller tax base after capital flight (the accepting base erosion/robust capital base dimension). Empirically, the very fact that regulatory competition has not yet led to a general race to the bottom suggests (in the absence of serious attempts at harmonization or capital controls) that acceptance has been the most common response to the quadrilemma in the world so far. Because states can choose to accept base erosion, we cannot directly infer the influence of capital flight from looking only at general trends in tax rates (or other policies). The influence of capital flight might instead show up as a (relative) decline in capital base.
Perhaps the most important line of objection to the baseline model has been that states cannot simply be modelled as tax-revenue-maximizers. Political scientists have amended the model to incorporate more complex decision-making processes within the state (Plümper, Troeger, and Winner 2009, pp. 764-765). From my perspective, these amendments strengthen rather than weaken the case for a quadrilemma. That investorfriendly policy has not been the universal response to the quadrilemma is precisely because governments care about goals others than maximizing revenue. Different countries have responded to capital flight in different ways (Plümper, Troeger, and Winner 2009, 781), and this is what we would expect if we understand their responses as preferences for different trade-offs among the goals of the quadrilemma.
Accepting base erosion is the option which seems to be missing from Rodrik's trilemma. However, the comparison between his trilemma and the quadrilemma is not straightforward. As noted above, his subject matter is much broader, and he does not specifically claim that a trilemma applies to real capital flight, although this is a natural reading. 9 Moreover, Rodrik's trilemma can be preserved insofar as the normative importance of both the capital base and of investor-limiting policy can both be evaluated in terms of the value of democracy. Nonetheless, it is important to distinguish between these two different ways in which a country's democratic choices can be reduced, and Rodrik's trilemma does not allow this. A country might well choose to preserve demanding environmental or labour laws and accept that the cost of doing so is less investment and tax revenue. This is a very different approach than that of a country which de-regulates in order to attract investment. Putting the former option on the table mitigates Rodrik's push for capital barriers or international harmonization as the only democratic options. 10 Indeed, some might push this point still further, and argue that the costs of accepting base erosion are so minor that this option dominates the others in the quadrilemma. However, while economic fundamentals such as labour costs and infrastructure matter more for investment location decisions, even small differences extended over a long enough time period can accumulate to a significant impact. Moreover, an important possibility suggested by empirical scholars such as Kimberly Clausing (2016) and George Zodrow (2010) is that the relative moderation of real tax competition thus far may itself be partly due to the prevalence of virtual tax competition. If states can succeed in making it harder for companies to avoid paying tax where their economic 9 The strategies in Rodrik's (2012) trilemma are 'Global Governance' (corresponding to what I refer to as international harmonization), the 'Golden Straitjacket' (corresponding to investor-friendly policy) and the 'Bretton Woods compromise' (corresponding to capital barriers). Looking at Rodrik's earlier work one can see how his political trilemma hypothesis grew out of an extension of the monetary policy trilemma rather than from reflecting on the case of real capital flight (Rodrik 2000). In his brief specific discussion of corporate tax competition, Rodrik is very focused on the change (the extent to which rates have been reduced), and does not discuss the continuity (the extent to which corporate profits continue to be taxed) (Rodrik 2012, pp. 193-194). While I think Rodrik's analysis of real capital flight is somewhat simplistic, I do not really regard this as a criticism, given the scope of his work. I see the capital flight quadrilemma rather as a deeper and more precise analysis of this particular aspect of the broader theme. 10 Rodrik's own normative preference for democracy pushes him towards either capital barriers or international harmonization. Of the two, he opts for capital barriers, leaning on what I refer to below as the 'non-ideal democratic' argument to criticize the international harmonization strategy. Interestingly, when Rodrik first introduced the trilemma in his earlier work (2000), he made the prediction that 'In the next 100 years or so, I see a world in which the reach of markets, jurisdictions, and politics are each truly and commensurately global as the most likely outcome. I may also be biased, since that is the option that I personally like best.' While this is technically compatible with his later (2012) stance, when he returned to the subject a decade later (a period which included the financial crisis) his emphasis shifted strongly towards retrenching the nation-state.
activities are really located, companies will have stronger incentives to respond to regulatory differences when locating their real investments. Solving the problem of virtual tax competition will thus only render the capital flight quadrilemma more prominent. Ultimately, the empirical question of how far states have suffered from pursuing more investor-limiting policies has not been settled and it is not my purpose to do so. Instead, this article attempts to clarify the extensive debate around capital flight by theorizing of the underlying structure of the trade-offs involved. We should also ask how far the apparent viability of accepting base erosion actually instead reflects the persistence of one of the other strategies: capital barriers. This brings us to the third element in the trade-off.

Capital barriers or capital mobility
The first two strategies accept (enthusiastically or resignedly) the tendency of investors to seek investor-friendly jurisdictions. The remaining two strategies attempt to eliminate capital flight by removing one of the conditions of its existence -capital mobility in the first case, international policy variation in the second. These strategies are not being pursued seriously by any countries at present. For this reason they tend to be absent in empirical discussions of capital flight and regulatory competition.
This section considers barriers against capital flight: By putting up barriers, a country can retain its capital base while pursue investor-limiting policies, without loss of international independence. Of course, this comes at the cost of capital mobility, which could instead be advanced through the strategies of investor-friendly policy, international harmonization or accepting base erosion.
Capital barriers can be natural or artificial. Artificial barriers are known as capital outflow controls. The simplest form is a proportional tax on outgoing capital, up to a full 100% (a complete ban). As a lesser step, some authors have suggested that a small general financial transactions tax (Tobin tax) would help ameliorate regulatory competition (Ronzoni 2014, 52). Any artificial barriers to capital mobility would supplement natural barriers that already exist.
Natural barriers have been another important way of explaining why the complete race to the bottom predicted by the baseline model has not occurred. These barriers take the form of things like transport costs, exchange rate uncertainties, fear of expropriation, and limited knowledge of foreign investment opportunities (Sinn 2003;Zodrow 2010). Insofar as there are natural barriers, this simply means that a government's response to the quadrilemma is to that extent pre-determined rather than up for choice. The corollary of this has not been recognized: when governments seek to reduce natural barriers (for example by improving transport infrastructure or concluding investment treaties with foreign states), any increased capital mobility that results must come at the expense of one of the other goals of the quadrilemma. Thus, even if we suppose that accepting base erosion is relatively painless at present, the costs of this approach may well mount as natural barriers to capital mobility are continuing to decline (Zodrow 2010).
Complete barriers to capital mobility mean a country has a closed economy when it comes to capital transactions. In such a country, capital could not leave in search of a more investor-friendly environment elsewhere, and so the regulatory policies of other countries would cease to have an impact on a country's capital base. Capital barriers at lesser levels reduce the loss of capital base that ensues when a country pursues investorlimiting policy.
By definition, capital barriers come at the cost of capital mobility. Capital mobility is treated as a goal in the quadrilemma because it allows greater gains from trade. If capital is not mobile, investments will not be made where they are most efficient. For example, with capital outflow controls, domestic investors looking abroad must bear in mind the added cost of capital controls, and foreign investors will be deterred from entering the country by the knowledge that it will be expensive if they later choose to leave. This creates an absolute 'deadweight' loss of economic welfare. Some of this loss may accrue to foreigners, but much will accrue to residents of the state in question. Countries that need to attract capital from abroad face a particularly acute problem. Capital outflow controls are only helpful if a country actually has capital to begin with; it is something of a solution for a first-world problem (Przeworski and Wallerstein 1988, 20). This is another way of saying that for some countries, for reasons beyond their control, capital mobility may be an especially important goal.
I now briefly consider four objections to the use of capital controls, aside from their fundamental cost in terms of capital mobility.
The first and most basic objection to capital controls is the claim that capital strikes can still occur in a closed economy. If this were the case, there would still be pressure towards investor-friendly policies, and so private investment decisions would still influence democratic decision-making. However, this claim was put to rest by Adam Przeworski and Michael Wallerstein's (1988) analysis of the structural dependency of the state on capital. Przeworski and Wallerstein's analysis shows that a state with a closed economy can move directly against the interests of investors without fear of negative economic consequences. They show that taxes in a closed economy can be designed to eliminate negative incentives on investment that might otherwise deter governments from the policy. Taking away investors' money is the type of policy that most obviously limits investors' interests. Thus, if states with closed economies are not constrained to avoid even this direct assault on investors' interests, we can take it as strong evidence that they are not economically constrained to avoid investor-limiting policy more generally.
Although the details of their model are complex, the essence of Przeworski and Wallerstein's argument is relatively simple. The longstanding worry that taxing investors reduces investment is valid when it comes to simple taxes on income. However, a government concerned with maintaining private investment while raising revenue can avoid this trade-off simply by exempting investment from taxation. Exempting investment from taxation would remove any disincentives to invest and place the burden of taxes entirely on consumption (expenditure). This is consistent with a very high tax rates and highly redistributive policies. The conclusion is subject to a caveat, which is that there will be transition costs during the anticipation period between the announcement and implementation of a tax (Przeworski and Wallerstein 1988, 23). Investors will tend to reduce investment during the anticipation period, aiming to consume now while consumption is cheap rather than later when it will be taxed. Once this transition cost is paid, however, governments in closed economies can pursue more or less investor-friendly policies without impacting on investment decisions. This in turn means investment decisions need not impact on democratic decisions about whether to pursue a more or less investor-friendly policy.
A second objection to capital outflow controls is to doubt their effectiveness (Dietsch 2015;Eichengreen 1999). Investors whose capital is trapped in a country have a strong incentive to circumvent the controls, for example by disguising capital transactions as over-priced commercial transactions. How far capital outflow controls can be designed to have their intended effect is an open empirical question, and difficult to answer until we can observe a serious attempt at implementation under contemporary conditions. With an eye on such worries, Dietsch (2015, 69) claims that capital outflow controls are in any case unnecessary to defend the tax base so long as a state can effectively tax foreign-source income. While it is true that taxing foreign-source income is preferable where possible, this only works where the investor remains a taxable citizen of the home country. If the investor moves along with their capital or resides in a third country, only capital controls can prevent capital flight.
A third objection to capital controls is that they are difficult to publicly deliberate about in advance, something potentially regrettable from a democratic perspective. For maximum effectiveness, capital controls should be introduced without warning in the dead of night. This is because anticipation effects are likely to be particularly strong, with investors trying to get out quickly before the controls are implemented.
A final problem (from the perspective of an individual state) is that permanent capital outflow controls have historically been opposed by the World Trade Organization and the International Monetary Fund, and may even prompt sanctions. Of course, this is a policy choice by these bodies, the justice or legitimacy of which can be questioned.

International harmonization or international independence
The final strategy states may pursue is international harmonization. International harmonization reduces the national differences that are one of the necessary conditions for capital flight. It effectively turns all the participating countries into one big closed economy. As we saw in the previous section, in a closed economy a state can carry out investor-limiting policies without loss of investment. Participating countries maintain capital mobility without compromising their capital bases. However, international harmonization comes at the cost of international independence, a goal which can instead be advanced through strategies of capital barriers, investor-friendly policy or accepting base erosion.
Unlike the other ideal-typical strategies, harmonization will almost certainly have to be supplemented with some element of an alternative strategy. As long as some other countries do not wish to harmonize or cannot be forced to do so, the harmonizing bloc will have to adopt one of the other three strategies with respect to those outsiders. The bloc must either (a) accept capital flight to outsiders, (b) adopt a relatively investorliming policy in order to prevent capital flight (which would tend to remove the point of harmonization), or (c) enforce capital controls on transactions to outsiders. The fact that harmonization needs to be supplemented with another strategy is often overlooked by its advocates. Dietsch (2015), for example, endorses harmonization on a minimal investor-limiting policy. Yet, he does not envisage any problem of capital flight to outsider countries, and he rejects the use of capital controls. Technically, there is no inconsistency in Dietsch's stance, because his ideal theory is about what all countries should do; but this leaves us no guidance about what should be done if some countries do not follow his prescription.
International harmonization contradicts a goal I refer to as international independence. This is because harmonization requires individual member-states to subordinate their own policy preferences to the policies agreed for the bloc as a whole. Of course, the extent to which independence is compromised reflects the depth of harmonization. There might also be a loss of independence for states outside the harmonizing bloc which feel pressured to adopt the bloc's policies in order to avoid being subject to capital controls. Their loss of independence would be even more severe since they would have no say at all in shaping the bloc's policies.
There are at least four different reasons why one might value international independence of the kind that is reduced by international policy harmonization.
First (and most often invoked in discussions of capital flight (Dietsch and Rixen 2014;Ronzoni 2014)) is a 'liberal' argument for international independence: policymaking on a national as opposed to regional or global level allows for a more finegrained responsiveness to local policy preferences.
Second is an 'experimental' case for international independence, advanced by J.M. Keynes (1933) in the context of capital flight. 11 Keynes stressed the epistemic value realized by different countries experimenting with different economic systems. For Keynes, facilitating such experiments was a reason to adopt capital controls to limit the homogenizing effects of capital mobility and regulatory competition.
Third is a 'non-ideal democratic' argument for international independence. If serious policy decisions are made at the level of a harmonized international bloc, these decisions should made democratically. However, democracy at this kind of geographical scale and level of diversity may not be feasible. Streeck (2017, 188) puts this argument in its starkest terms: 'in the world as it is, democracy cannot be had without state sovereignty.' 12 Finally, there is a 'communitarian' case for international independence. Citizens often seem to view international independence as intrinsically valuable. Whether or not we agree with this view, their preferences may hold some weight.
International independence is used here as a term of art referring strictly to independence vis-à-vis other states. It should not be confused with a state's ability to set policy more generally. 13 International independence normally contributes to a state's ability to set policy more generally. However, a country that prioritizes cultivating foreign investment above all else would have no remaining space for policy decisions, even if it retained full international independence. The liberal, experimental and democratic arguments would apply equally strongly against such a course of action. These three arguments thus function mainly to recommend capital barriers or accepting base erosion rather than international harmonization or investor-friendly policy. 11 For discussion see Crotty 1983. 12 For a contrary view see Weinstock 2009. 13 My distinction between 'international independence' and 'ability to set policy more generally' maps Ronzoni's (2012) distinction between 'negative' and 'positive sovereignty'. Dietsch 2015, Chapter 4 uses a similar distinction.
However, the communitarian case for independence applies only against international harmonization, and does not also provide an argument against investor-friendly policy.
In some situations countries adopt relatively similar policies as a result of genuine regulatory agreement rather than deliberate harmonization. According to the experimental argument for international independence, this kind of accidental harmonization is still something to be regretted. However, this is not the case for the other three arguments. On those accounts, to the extent that harmonization occurs accidentally through regulatory agreement, the quadrilemma is genuinely mitigated and progress is possible on all four goals at once. Nonetheless, a wide range of genuine regulatory disagreement between countries is unlikely to disappear any time soon.

Representing the quadrilemma
Putting together these various goals and strategies, we have the capital flight quadrilemma, represented above in Table 1 in terms of ideal-typical strategies. I will now describe a better way of representing the four-way trade-off, not as a choice between four discrete options but as the choice of a point in a continuous space of options.
Only very rarely do states adopt one of the pure ideal-typical strategies. Normally, states will adopt a mixture with elements of each. For example, a state might be part of some broad harmonizing bloc setting some minimal standards of investor-limiting policy. To protect the bloc, this will require capital controls on outsider countries. At the same time, within the minimally harmonizing bloc, the country might make some attempt to offer taxes and regulations that are attractive to investors, while also accepting some degree of base erosion to more competitive jurisdictions. A state's choice is thus about the proportions in which to mix the four ideal-types.
To represent this choice, I start with the simpler case of a three-way trade-off. Figure 1 represents the trade-off between investor-limiting policies, capital mobility and international independence (see below). Different policy choices are represented by different points within the triangle. The distance between a point and one of the triangle's sides represents how well that policy realizes a goal. Thus, the closer a point is to the bottom of the triangle, the better it realizes the goal of investor-limiting policy. This is why the three ideal-typical strategies sit in the corners of the triangle. The pure strategy of capital barriers, for example, sits as close as possible to the lines representing international independence and investor-limiting policies, but is as far as possible from the line representing capital mobility. International harmonization, by contrast, is as close as possible to investor-limiting policies and capital mobility but as far as possible from international independence. A point at the very centre of the triangle would represent the mixed policy of a state which combined some capital barriers, some international harmonization, and some effort to attract international investors.
Of course, Figure 1 only shows a trilemma, not a quadrilemma. What Figure 1 is missing is the possibility of simply accepting some base erosion. Put another way, Figure 1 assumes that the goal of a robust capital base is non-negotiable. In order to represent this fourth goal, we need an extra dimension. Imagine there is a fourth corner to Figure 1 situated above the centre of diagram, representing the ideal-typical strategy of accepting base erosion. Instead of a triangle, we now have a three-sided pyramid. Policy positions are now represented by a position within the volume of the pyramid, and each goal is represented by a different face. The pyramid is shown in Figure 2 below, with the four ideal-typical strategies labelled at the four vertices.
The strategy a country should choose depends on how it evaluates the trade-off between the four goals. This has two elements. First, there are empirical questions about how any given strategy will perform with respect to each goal. For example, what are the economic costs of different levels of capital controls, if they can be made effective? How much capital flight will there actually be under different combinations of policy, capital mobility and harmonization? The answers to questions like these also depend on the choices of other countries. For example, to the extent that other countries adopt strongly investor-friendly policies, choosing to accept base erosion comes at a greater cost to the capital base. Second, there are normative or evaluative questions about how trade-offs between the four goals should be weighed. Democracy requires that the strategy a country chooses should reflect the judgements of its citizens about the relative values of the four goals. One important implication of this is that international bodies such as the IMF and WTO should cease their unqualified opposition to long-term capital outflow controls. The opposition to capital controls implies a belief that capital mobility is more important than the other three goals of the trilemma. But this cannot be defended in all situations, and in any case it is a choice that democracies should make for themselves.
The value of democracy does not, however, preclude theorists from offering arguments about what relative valuation of the goals they believe is appropriate. As already mentioned, the trade-off should at least be sensitive to the situation of the country in question. In particular, poorer countries for which economic growth is more morally urgent should generally value capital mobility and the capital base more highly than richer countries for which it is not. 14

Conclusion
The central claim of this article has been that countries face a four-way trade-off between capital mobility, investor-limiting policies, international independence, and the capital base. This is the primary way private investment decisions can impede democratic decision-making. The free movement of investments must come at the expense of at least one of the other goals in the quadrilemma. And these other goals all reflect the value of democracy in different ways. It is valuable that states can pursue investor-limiting policy and a robust capital base because democracies should be free to pursue economic and social policies of their choosing. International independence too has important democratic elements. Independence allows experimentation and responsiveness to local preferences and reflects the value people place on national independence. Moreover, national states thus far have a better democratic track record than supranational institutions. While the quadrilemma shows how international capitalism can conflict with democracy, I submit that we understand the issue more clearly when we disaggregate the different democratic elements at stake into the different goals of the quadrilemma. This is the best way of conceptualizing the issue of regulatory (including tax) competition in normative terms. Because of the trade-off, there can be no such thing as a 'solution' to real capital flight: all options come with costs.
I analysed the quadrilemma as a decision each country should make based on its interests in each of the four goals. In the first instance, the balancing between the goods of capitalism and democracy must take place at this level. However, as we have seen, the choice one country makes will also affect how far other countries are able to meet the four goals. An important goal for further research is thus a framework for fairly balancing different countries' preferences for responding to the quadrilemma.
In the medium term, political actors concerned by tax competition can make substantial and important progress in tackling 'virtual' competition over facilitating tax avoidance. A fair system for accounting for profits across borders need not require compromises between private investment and democracy. However, the point of the capital flight quadrilemma is that such a compromise is ultimately unavoidable when it comes to real investment. If globalization continues, how to respond to the quadrilemma is a question that will only become more urgent in the coming century.

Disclosure statement
No potential conflict of interest was reported by the author(s).