A Thai taxi motorbike rider holds a flag calling on Thais to vote during a campaign to promote the referendum on a new constitution in Bangkok. (Rungroj Yongrit/European Pressphoto Agency)

A man walks into a bookstore and asks for a copy of the constitution. “We don’t sell periodical literature,” replies the manager. This joke dates to 1958, when France passed the Fifth Republic constitution.

So spare a thought for Thai voters, who on Aug. 7 approved the country’s 20th constitution since the end of absolute monarchy in 1932. In fact, Thailand has changed constitutions on average every 4.2 years, about as frequently as other countries change governments.

If citizens don’t agree with the constitution, government has a problem

But while Thailand’s experience may be exceptional, there’s a huge global principle here, one that affects established and emerging democracies alike: How much power should elected authorities wield?

A constitution sets the rules that govern how the game of politics is played. Above all, it defines how countries choose political leaders, how these leaders govern, and the scope of their authority. A stable democracy combines civic consensus over these rules with political competition within them.

The legitimacy of a constitution ensures that, win or lose, everyone accepts the outcome of an election and the decisions of a new government. If a significant group ceases to accept the rules, civil conflict and breakdown of authority are likely to follow.

But Thailand, more than any other country, has persistently failed to secure a consensus on its rules of the game. Since the 1970s rapid modernization has exacerbated this failure by throwing up new social forces to challenge the established order. A rising urban middle class — and more recently an increasingly assertive rural majority — has demanded participation in politics and access to resources. These groups have met with resistance from entrenched interests and institutions.

Thailand continues to fine-tune its rules of democracy

Since 2001 this color-coded confrontation of “yellow shirts” (royalist establishment) and “red shirts” (the populist opposition) has deepened. The rural majority mobilized by Thaksin Shinawatra, prime minister from 2001-2006, has repeatedly returned governments only to be removed by coups or other devices (including disqualification for hosting a cooking show). Thailand has become caught in a “Groundhog Day” of election-coup-constitution-election.

The military government in power since the May 2014 coup drafted this latest constitution, which tries to restore democratic elections while constraining future democratic governments. It does so by hemming in elected authority with unelected institutions and other devices. For example, the constitution adds an appointed upper chamber and the possibility of an unelected prime minister. A range of bodies, filled with “good people,” as they are known in Thailand, is also empowered to step in and override government decisions.

The deep polarization of Thai politics means the new constitution is unlikely to enjoy the consensus it requires in order to last. The country does not seem likely to escape its version of “Groundhog Day.”

What are the options for fine-tuning the practice of democracy?

Thailand’s experience offers clues to the prospects for democracy elsewhere in East Asia, where elites face growing popular demands for participation. As the risks and costs of resisting change rise, governments have three options:

1) A phased-in democracy —  The approach used in 19th century Western democracies was to expand participation gradually through controlled widening of the franchise. But this option is not available in the 21st. With a handful of exceptions, mostly in the Middle East, the franchise is now all-or-nothing: voting rights can no longer be granted selectively. Most non-democracies already grant nominal universal suffrage in any case, though the elections themselves may be meaningless.

2) Free elections, with falsified results — A more recent elite strategy is electoral autocracy: the appearance, but not reality, of democratic elections. This is achieved through the systematic use of “administrative resources” to skew and, if necessary, falsify, the election outcome. These methods work best where states are strong and civil society under-resourced, as in much of the former Soviet Union.

But this approach has risks, especially when digitally empowered and connected citizens can detect and publicize fraud more effectively than ever. In recent years, mass protests against manipulated elections have led to government overthrow in Ukraine, Georgia, Serbia and elsewhere.

3) Free elections, but reel in government authority —  A third strategy is to allow largely free and fair elections but constrain the scope of democratic authority. Elites thereby hope to satisfy mass demands for electoral participation while protecting their own interests. This is what Thailand will look like under the new 2016 constitution.

And this was the approach taken in neighboring Myanmar where, following last year’s historic election, Aung San Suu Kyi’s National League for Democracy now governs within a framework designed by the previous regime to entrench military influence. This could also prove a model for elite responses to growing pressures elsewhere, including Vietnam, Laos and perhaps one day even China.

Democracy is challenged in established democracies as well

The scope of democratic governance has become a live issue in stable Western democracies, too. Here, this scope has been narrowing for decades as governments have ceded power to the institutions and practices of globalization.

These constraints on democratic decision-making come not from “good people” representing elites, but from “experts” — technocrats, unelected institutions and markets —  that purport to know better than citizens do how to achieve their interests. A reappraisal from below is now underway. The theme of “taking back control,” which framed the Brexit campaign, resonates across the European Union —  and in the United States.

A common question underlies all these concerns about democracy in East and West alike. Where should the boundary between democratic and non-democratic authority be drawn? In non-democracies, this will be negotiated between self-protecting elites and populations demanding greater inclusion. In democracies, it will be negotiated between elite beneficiaries of globalization and populations that value national autonomy more highly.

Posed in different ways in different countries, this issue of the proper scope of elected authority is now the most important question facing democracy around the world.

Nigel Gould-Davies teaches at Mahidol University International College in Thailand, and is an associate fellow of Chatham House.