Cora Chan is a professor of law at the University of Hong Kong. The funding of the Research Grants Council of Hong Kong (RGC Ref. no. 17600524) is acknowledged.

Assessments of democracy are usually made at the national level, but in fact the quality of democracy can vary greatly within nations. There can be authoritarian regions within national polities deemed democratic, as the example of America’s Deep South from the 1890s to the 1970s shows. This piece focuses on the reverse phenomenon of liberal enclaves, i.e., subnational regions that enjoy substantially more freedoms, fairer elections, and/or a stronger rule of law than the national entity. Such enclaves are rare, but by no means non-existent. Hong Kong in China is a notable contemporary example. Parallels can also be found in empires (e.g., Finland in Tsarist Russia), occupied territories (e.g., Denmark under Nazi occupation), and polities undergoing democratization (e.g., regions in Northern Mexico prior to national democratization). Furthermore, liberal enclaves may arise in democratically backsliding nations (think California in the Trump-led USA). Despite the enormous scholarly and practical value of the phenomenon of liberal enclaves, it has largely been overlooked. My key message here is that if we care about sustaining democracy in an age of democratic backsliding, we should pay more attention to it.

Liberal enclaves matter to democracy for four reasons. First, they matter to the populations living in those enclaves, who are able to enjoy more democracy than their counterparts in the rest of the nation. Second, liberal enclaves help to preserve what democratic traditions remain in authoritarian or backsliding polities, thereby facilitating re-liberalization at the national level should the opportunity arise. Third, they are sites for experimenting with, reimagining, and reforming democracy, thereby helping us rectify the weaknesses in democratic institutions that have led to their erosion. Finally, the civil society, groups, and leaders cultivated in liberal enclaves, as well as democratic institutions therein, such as free and fair elections and independent courts, can set in motion dynamics that decelerate autocratization or contribute to liberalization at the national level. In short, liberal enclaves act as shelters, reservoirs, laboratories, and engines for democracy in an authoritarian environment.

How then can such enclaves endure under authoritarian pressures? The answer depends on numerous factors, including the type of authoritarian and territorial regime involved and the nation’s stage of autocratization. The pathways to withstanding a backsliding yet still by-and-large democratic federal regime may differ from those in a consolidated, unitary, one-party state. Nonetheless, I offer some observations from my study of liberal enclaves in authoritarian regimes. These observations are also relevant to backsliding yet still-democratic regimes because they forewarn us of what might happen if the backsliding continues.

First, the perceived level of threat to the authoritarian regime’s rule posed by maintenance of the liberal enclave is a key factor affecting such maintenance: the lower the authoritarian regime’s perceived threat, the more likely the regime is to tolerate the enclave. The tricky thing is that a strong civil society is needed to maintain freedoms, but if it acquires such a character that the authoritarian regime perceives its rule to be endangered, then the regime may respond with repression. Second, the relative economic power of the enclave and the center matters: the more the center’s economy depends on the liberalness of the enclave, the more likely it is to tolerate the enclave. Third, the presence of regional institutions (such as a governor’s office) with the will and tact to protect the region’s liberal character is vital, and it is important to recognize that compromises with the center may be needed. Fourth, if the center is much more powerful than the enclave, the latter’s liberal status can deteriorate rapidly once the former moves toward repression. Institutional behavior can change overnight; civil society can be forced into silence in months. But amidst it all, and this is my final point, there are remnants from the previous liberal era—traditions, legal doctrines, and habits of heart and mind—that serve to keep spaces that are not politically sensitive relatively free. The courts may be highly deferential toward the regime where its core interests are at stake, but they may otherwise continue to rigorously protect constitutionalist democratic norms. This heterogeneity in constitutionalism within a subnational region, as with that at the national level, offers us both hope and possibilities for sustaining and even strengthening democracy at a time when it is facing a global crisis.