Democratization Policy Council

"An initiative for accountable democratization policy worldwide"

Lessons from the Balkan Playbook: Counting on the Census, and Deciding Who Counts

24 June 2019 — Valery Perry and Kurt Bassuener

Over the past two-plus years it has become increasingly clear that illiberal politicians and policymakers in the west have either been directly inspired by, or have unconsciously imitated the behavior of, their more experienced peers in the Balkans who, since the wars of the early 1990s, have perfected a political modus operandi devoid of accountability and grounded in self-dealing. As one of the authors was quoted in George Packer’s recent Atlantic piece, Donald Trump is “America’s first Balkan president”.

As norms are shredded and democratic checks and balances in the US are stretched to their limits – domestically and internationally – it is useful to explore the parallels, both to warn US audiences of the potential consequences of democratic backsliding in even the most established democracies, but also to remind citizens in the western Balkans that awareness of and resistance to such tactics is a universal, and never-ending, struggle, with some times more challenging than others. For the US and some other established democracies now exhibit the telltale symptoms of a malady long endemic in the Balkans.

One recent example of Balkan politics being replicated is related to the US census, which will be held in 2020. The organization of a decennial count is one of the few specific, technical requirements for governance that is explicitly required in the US constitution (Article I, Section 2). While there are many other methods of statistical collection that occur at the federal and state level on an ongoing basis, the official census, organized every ten years, is critical for two main reasons: it determines the apportionment and delineation of Congressional seats (power), and it serves as the basis for a substantial number of funding decisions (money).

Some similarities (in intent and method) are already evident in controversy about the US census, and the similarly contentious census that was held in BiH in 2013 – the first post-war census in the country. Most notable is the politicization of both census processes through engineering aimed at reducing response rates based on highly targeted efforts grounded in identity politics that are viewed as potentially benefitting one political party, though at the expense of accurate and reliable numbers.

In BiH, one of the main factors that both delayed – and ultimately shaped – the census was the preoccupation of the ruling parties on three identity questions. It was well known at the time that these questions were not required by Eurostat (meaning that BiH could access every cent of potential EU funds etc. without asking such questions). Further, good practices when asking such questions were known, but were minimally followed. The result was a process in which the ethnic identity question was paramount and framed as such. It received far more attention than questions related to income, education, method of commuting, etc. This was flagged by the authors, BiH civic actors, and others to the EU and the bilateral donor, Sweden, which bankrolled the census. The donors could have made their funding conditional on the question NOT being included, but determined that would be too prescriptive. The fact that the parties could agree was seen as a win in itself – despite the fact that the foundation of their agreement was inherently retrograde and disintegrative.

Further, this focus led to inevitable debates and disagreements on the results, with the politicization of these questions undercutting faith in the data collection and aggregation process. Some egregious procedural problems were identified during the census process. And far from helping prevent, deter, or at least offset the politicization of the process, the donors who largely funded the census helped compound the damage, doing precious little to ensure the results were appropriately aggregated, and allowed political actors to play these games with these external census funds.

The US is seeing similar controversy unfold, which will soon be determined in the Supreme Court. In this case, the issue is about whether the census should include a question on whether the respondent is an American citizen. While at various times and in various ways in US history some questions related to citizenship have been asked, “Never before has the federal government used the census to directly ask for the citizenship status of every person living in every household in the United States.”

While in theory all individuals present in the US at the time of the census could participate in the count, there are credible concerns that individuals or families that have one or more undocumented resident will, due to data privacy concerns and reasonable fears of future possible legal consequences, opt to avoid participation, leaving that family’s numbers off of the final count. This would reduce the number of respondents in urban regions – where many new immigrants (both documented and undocumented) often live. (A useful summary of the issue is available here.)

The Trump administration has claimed that the citizenship question is needed to ensure the implementation of the Voting Rights Act, However, the lack of interest in expanding voting rights, and the plethora of other ways that this could be achieved that are unrelated to the census, has been contested since the addition of this question was proposed. In fact, the administration and Republican Party have made a concerted effort to roll-back the Voting Rights Act and voter access, so their credibility from that standpoint is less than zero.

Even more troubling is the recent discovery of argumentation and data that seems to demonstrate that key consultants were in fact completely aware that adding the question would reduce the number of non-white American respondents, and in turn reduce the numbers of people in urban (nearly always Democratic) areas, and, inter alia, reduce their representation and budget allocations while increasing representation in rural districts. This follows years of data-driven electoral engineering (the state of Wisconsin offers a useful case study of the impact such activity can have on a state). This, together with past and ongoing gerrymandering to craft electoral districts that guarantee “safe seats” would help to consolidate Republican minority control.

Identity politics is being manipulated to establish and entrench political dominion, as it was in BiH.

Further, the results of gerrymandering – more precise than ever based on sophisticated, computer-aided electoral district delineation - leads more and more to a system based not on functional, representative, accountable, rules-based decentralization, but a partisan and identity-based fragmentation. This has echoes of governance in BiH, where ethnonational de facto federalism based on identity is prioritized over functional federalism based on the geographic reality of efficiently providing quality services to citizens.

The American Civil Liberties Union (ACLU – a large NGO that engages in significant legal aid and strategic litigation in the interest of protecting civil liberties; V. Perry is a contributing member) has filed a notice to the Supreme Court arguing that the initiative to add this citizenship question is politically motivated, and has been proposed and defended in bad faith and through factual misrepresentation. There is speculation that the court could allow the census to go forward with this census question in spite of the newly available facts, perhaps on procedural/administrative grounds often favored by conservative yet legacy-conscious Justice Roberts.

Finally, this issue is also troubling for the role that it is playing in detracting from policy grounded in independent facts and data, both by undermining the data collection methods, but also by reducing the general confidence that the process is itself normal, apolitical and independent. This would fit with Steve Bannon’s stated goal to “destroy the administrative state.” There has been some speculation that if this question is added, a civic strategy of “resistance” in which citizens might opt to give untrue responses, or themselves refuse to participate, could further undercut the validity of the data.

Such a response to deliberate political engineering would represent yet another intentional step to erode public confidence in governance and the existence of factual information grounded in scientific processes upon which policy may be made. This in turn would drive further alienation of individuals from their polity, taking away their sense of agency and atomizing individuals to the point where effective collective action on any policy issues is made more difficult, if not impossible. This is the desired end goal of political elites and their donors/supporters disinterested in accountable and truly representative politics and policy.

This was similarly the case with the BiH census, where the process was transparently politically instrumentalized from the beginning – and the results are disputed by the parties which agreed to include the identity questions for political ends. Public confidence in officialdom in that country is abysmal; a 2015 poll in the wake of large 2014 street protests is telling, and frustration has only grown since. Such an approach succeeds in making destruction of public trust a two-way street, resulting in a downward spiral that works for enemies of democracy and good governance worldwide.

So Americans and those who have looked to the US for governance practice should take note of the forthcoming Supreme Court’s ruling, argumentation and impact. If the citizenship question is permitted, it will undoubtedly act as an accelerant in the unwinding of American social trust and civic practice.