Constitutions set out the basic rules of the political game. Accordingly, for those who care about democracy, a key question is which constitutional rules make it most likely that democracy endures? Indeed, in our present age of global democratic backsliding, this is perhaps the most important topic for constitutional law and theory.

One familiar challenge of constitutional design, made newly relevant in the era of backsliding, concerns the problem of government power: in short, how easy or hard it should be for government to act. The perils of unchecked government power are obvious. There are countless ways for an unchecked government to undermine democracy, including by seeking to undermine free and fair elections, trample individual rights, or repress political opponents and minority groups. (These concerns may be especially acute for chief executives, who can act with what Hamilton called “energy” and are not beset by the collective action problems that make legislating difficult.) Anxiety about malicious uses of state power resonates with what Judith Shklar famously described as a “liberalism of fear,” which “begins with the assumption that the power to govern is the power to inflict fear and cruelty.” This line of thinking generates a clear lesson for constitutional law: Limit the power of government.

And yet. If the state’s power is excessively limited, government may prove unable to meet pressing public problems. Public law rules might restrain executive action by imposing excessive procedural hurdles, public participation requirements, or analytic requirements. They might restrain legislatures by creating supermajority voting rules or other hurdles that make it difficult to enact new statutes or easier for narrow interest groups to capture the legislative process. Judicial interpretations of constitutional and statutory law alike might further hem in the powers of the executive and the legislature. And a culture of adversarial legalism, coupled with high polarization, can mean that any major executive or legislative action is virtually guaranteed to end up in court. Government may become a “vetocracy,” too shackled to act—even when the public demands it.

Vetocracy not only impedes good public policy; it also poses a lethal risk to democracy itself. What happens if existing democratic institutions cannot effectively respond to pressing challenges? What if voters concerned about society’s problems, or their own economic well-being, see government as unable or unwilling to help? These failures can make voters question democracy and ask whether a strongman could muscle his way through the country’s paralyzed institutions and deliver what democracy cannot.

Constitutional design in a democracy must find a middle path between these two extremes. An unrestrained government can undermine democracy today, but a vetocracy can undermine democracy tomorrow. How, though, can the right balance be struck?

One possible path forward is to maintain checks on government power while dramatically expanding state capacity. This could involve expanding and strengthening the government workforce: hiring more legislative staff to improve the lawmaking process, and staffing the executive bureaucracy with high-quality civil servants so that the government can act effectively. While far from the most exciting democracy reform, increasing capacity can allow government to continue to perform effectively even in the face of checks on its power.

Further, public law should recognize that not all forms of government power hold the same potential for abuse, and that the stakes of government inaction vary by context. This means that different rules should apply to different domains. Some types of government power, like the prosecutorial power, hold significant potential for abuse and little risk to the public interest if government has to clear more hurdles before proceeding; those types of power should be strongly checked, and often are. But other powers, like the power to issue a permit for a solar farm, hold much less potential for abuse, and the costs of inaction are higher. For these sorts of powers, the state should be able to act with a relatively free hand.

Public law rules already treat the two types of government actions differently, as they should. But the difference is not as pronounced as one might think: just think of the countless stories of relatively simple government projects that become mired in procedural delays, with little or nothing to show after years (or even decades). The types of checks that make sense for the powers of government most liable to abuse—such as the power to prosecute political opponents or the power to regulate the media—often do not make sense for ordinary policymaking.

Government should also steer clear of reforms that only seem to enhance democracy while in fact undermining it. Public participation requirements are one example. Enacted to enhance democracy by bringing the public into government decision-making processes, such requirements can become an empty ritual: they can slow down the policymaking process without meaningfully incorporating input from ordinary citizens, and they are often vulnerable to capture by special interests. Democracy’s defenders, especially the lawyers among us, often focus on process, but we should be wary of processes, even well-intentioned ones, that turn out in practice to obstruct the substantive outcomes that citizens rightly expect from their government.

More broadly, lawmakers and citizens alike should avoid the temptation to frame the question of constitutional design in binary terms: a “big” or “small” government, a “weak” or “strong” state, and so on. This way of thinking oversimplifies the issues at stake and may lead us to forget that for any power of government, the checks on that power should be tailored, in part, to the potential for government abuse (which point toward weaker checks) and the costs of inaction (which point toward stronger checks).

These suggestions are necessarily tentative. But my hope is that framing the questions facing democracy in terms of the dilemma of government power can help us to generate new ideas and better understand existing problems. Democracy requires limits on public power—but it also requires a government that is powerful and capable enough to deliver the substantive outcomes that citizens reasonably expect from their government.