Unbundling Corruption: Why it matters and how to do it

By Yuen Yuen Ang, Political Scientist at the University of Michigan, and the author of How China Escaped the Poverty Trap and China’s Gilded Age: The Paradox of Economic Growth and Vast Corruption

Corruption-whistleblower-shutterstock_1581042757Even amid a global pandemic, corruption persists and manifests itself in multiple forms, ranging from corrupt police extorting truck drivers delivering essential goods, rigged procurement contracts, to politically connected corporations receiving huge bailouts from the government while small businesses are starved of loans they desperately need to stay afloat. Although all of these actions are corrupt, they involve very different actors and stakes; some are transactional while others are extractive; and each brings about vastly different consequences.

Yet the conventional way of measuring corruption across countries does not capture qualitative distinctions across types of corruption. Instead, standard indices—most notably, the Corruption Perception Index (CPI)—measure corruption as a one-dimensional problem, ranging from 0 to 100. Consistently, rich countries rank at the top while poor countries are stuck at the bottom.

Because these indices are the primary—indeed the only—source for tracking corruption around the world, they have profoundly shaped the way everyone thinks about corruption. They result in a fixation with overall quantities of corruption and global rankings, at the expense of measuring and understanding the effects of different qualities of corruption.

Instead of asking “Which country is most (or least) corrupt?”, both analysts and practitioners should consider a different set of questions: “Which country is dominated by what type of corruption? Why? With what consequences? How can we fight different types of corruption?”

In order to address these questions, we must begin by creating an unbundled index of corruption.

How standard indices mislead

Global indices of corruption include the CPI, released annually by Transparency International (TI), and the World Bank’s “Control of Corruption” Index, part of the Worldwide Governance Indicators (WGI). These perception-based indices are constructed by gathering responses from surveys conducted by third parties (for example, the Economist Intelligence Unit and Political Risks Services Guide) and then combining them to construct a single score for each country. In 2019, the CPI ranked Denmark and New Zealand the two least corrupt countries in the world, tying at 87 out of 100 points, while Somalia, scoring only 9 points, was the most corrupt.

Measuring corruption around the world, year after year, is expensive and difficult. Thus, existing indices have provided an immense public service by allowing us to compare levels of corruption across countries. They also help raise awareness and mobilise public support to fight corruption.

Yet existing indices have also misled in two key ways. First, they do not distinguish among different types of corruption. Instead, they mush all corruption into a single score. Second, they under-measure a particular type of corruption found in rich capitalist economies, what I term “access money”—elite exchanges of power and wealth that can be legalised. As a result, rich countries always appear clean. This reinforces the impression that “corruption is [only] something that happens to less fortunate people in poor nations,” to quote Glaeser and Goldin.

Unbundling corruption

A necessary step toward creating a better measure of corruption is to unbundle corruption. In China’s Gilded Age (see also my review article, “Unbundling Corruption”), I advance a typology of four distinct types of corruption: petty theft, grand theft, speed money, and access money.

Figure 1

Fig 1 - typology

In particular, I draw a clear distinction between two types of transactional corruption. “Speed money” refers to petty bribes paid to low-level bureaucrats to circumvent hurdles or delays. “Access money,” on the other hand, are high-stakes rewards offered by elite capitalists to powerful officials in exchange for exclusive, lucrative privileges. Whereas petty theft, grand theft, and speed money are almost always illegal, access money can encompass both illegal (e.g., massive graft) and legal actions (e.g., political finance and lobbying). When people think of “bribery,” they usually only consider speed money, whereas access money is neglected.

Unbundled Corruption Index (UCI)TM

The idea that corruption comes in distinct varieties is not new—but it has generally not been translated into global indices. My research provides, to my best knowledge, the first indicator of distinct types of corruption across countries. I call it the Unbundled Corruption Index (UCI).TM The UCI is based on an original survey of experts in 15 countries that measures the perceived prevalence of the four categories of corruption identified in my framework (see Figure 1).

My survey makes a targeted attempt to measure the elusive category of access money—the purchase of lucrative privileges, both illegal and legal. Rather than ask country experts to rate overall levels of corruption in a country, which can be very misleading, I employ a “stylised vignette” approach to more accurately capture perceptions of various types of corruption. Table 1 summarises how the UCI addresses the three major limitations of standard indices like the CPI.

Table 1

Table 1

Source: Ang, China’s Gilded Age, pp. 31

For ease of comparison, I visualise the results in the format shown below. Each figure displays the total UCI score (listed below country name), and the distribution of this aggregate score across four categories. The category that takes up the highest proportion of score is interpreted as the dominant mode of corruption, highlighted in orange.

Figure 2

Fig 2 - UCI all

The most immediate advantage of the UCI, compared to standard indices, is that it allows us to compare not only overall levels of corruption, but also structures of corruption. For example, compare China and India. Although the two countries have similar total UCI and CPI scores, their patterns of corruption diverge. In China, the most dominant type of corruption is access money (elite exchanges of power and wealth), whereas in India, it is speed money.

Figure 3

Fig 3 - China vs India

The UCI also reveals that some high-income countries with low scores on petty theft, grand theft, and speed money can simultaneously, have moderately high levels of access money. A case in point is the United States, whose score on access money exceeds that of Ghana

Figure 4

Fig 4 - USA vs Ghana.

Three practical implications

Unbundling corruption is no mere academic exercise—it has three important practical implications. First, it affects financial decisions that could amount to billions of dollars. As Christiane Arndt and Charles Oman at the OECD pointed out, international investors and aid agencies use existing corruption and governance indicators to assess country and sovereign risk, which inform major decisions on investment, loans, and aid. Yet as my study finds, bundled ratings under-count forms of corruption commonly found in wealthy economies and over-count those in poor countries. One implication of this bias is that low-income countries could be financially penalised in the form of higher interest rates and lower aid or investment flows.

Second, because standard indices routinely under-measure legalised modes of access money, such corruption easily escapes the glare of public scrutiny and censure. Yet compared to bribery and embezzlement, its socio-economic damage is potentially no less severe. Consider, for example, the role of state capture in blocking climate action or precipitating the 2008 U.S. financial crisis. Such corruption also drives the rise of extreme inequality and populism.

Encouragingly, the OECD has begun to bring attention to “access money” in the forms of influence in policy-making, lobbying, and political finance through case studies. The UCI advances this agenda by providing the prototype for a global indicator of such corruption.

Third, an unbundled measure of corruption will help policymakers tailor anticorruption strategies to diverse contexts. The OECD has advocated for “micro,” “sector-specific,” and “project-specific” strategies that target corruption in particular contexts (see also Mark Pyman’s “Curbing Corruption”), rather than one-size-fits-all solutions. The UCI offers a balance between binary divisions, which are too coarse, and specific solutions, which are too diverse. This balance is essential for context-sensitive approaches to be widely applied in theory and in practice.

In principle, we should measure what we value, yet the reality is often the opposite—we value what we can measure. Because the easiest way to measure and rank corruption is to reduce it to a one-dimensional scale, this has become the default method. It has narrowed and even distorted our understanding of corruption. Unbundling corruption is a necessary step toward recognising the fact that corruption comes in distinct varieties, including legal modes of access money.