China’s anti-corruption drive. Small steps made in lots of different directions

How effective is Xi Jinping’s anti-corruption drive? If you believe the latest data published by anti-corruption NGO Transparency International then progress is indeed being made.  But, it’s slow going.

Trying to measure how much corruption exists is often a fool’s errand.  Corruption comes in many shapes and sizes and takes on a variety of different forms.  This doesn’t stop Transparency International, one of the world’s leading anti-corruption NGOs, from publishing an annual Corruption Perceptions Index (CPI).  Data is taken from a range of surveys on corruption and related issues and put in to what is effectively a poll of polls.  States are then ranked on a scale of 1 (the worst score possible) to 100 (the best).

There are many reasons to be just a little sceptical about the voracity of the data in the CPI, but that doesn’t stop it catching many policy-makers’ eyes.  No one wants to see their country going the wrong way in tables like these.  Chinese policy-makers are unlikely to be any different.

The 2016 data was released on Wednesday 25th January.  It makes interesting reading. Perennial front-runners Denmark and New Zealand again head the pack, coming in joint first with scores of 90.  Finland, Sweden and Switzerland are just behind them and they make up the top five.  Somalia remains bottom of the pile (10 points), with South Sudan (11), North Korea (12) and Syria (13) also doing particularly poorly.  Suriname showed the most improvement, scoring 45 in 2016 as opposed to 36 the year before.  The impact of high profile scandals surrounding the awarding of football world cups no doubt contributed to Qatar being the state that performed worst compared to 2015; down ten points from 71 to 61.

China has traditionally not done particularly well in the CPI.  In 2015 it was in 83rd place (out of 168 countries), scoring 37 out of 100. 2016 reveals some improvement; China’s score rose by three points to 40 and the PRC is now 79th out of 176 countries (8 have been added since the 2015 table was published).  A journey of 1000 miles begins, as the Chinese say, with one small step.

We should nonetheless be wary of taking the data in the CPI too seriously. Coming up with one single number to encapsulate all of China’s corruption problems, for example, would seem like a heroically optimistic undertaking.  How, for example, do you compare petty corruption in rural Guangxi with the grand corruption that might well take place in Beijing? Many would say you can’t.

Be that as it may, if you’re going to get to where you need to go, then you have to be taking steps in the right direction. And this data makes it look like China may indeed be doing just that.  Chinese policy-makers are certainly likely to be pleased to see their score moving upwards.

If, however, we take a closer look at what tends to characterise low corruption countries, then Chinese policy-makers will find much less to be enthusiastic about.  As Transparency International’s Chairman, José Ugaz, noted this week three key things help states genuinely root out corrupt practices; freedom of expression, transparency in all political processes and strong democratic institutions.  Only then will civil society and the media be able to hold those in power to account and only then will the building blocks for a successful anti-corruption fight be in place.

China, of course, displays precisely none of those core traits.  Freedom of expression is strictly limited, decision-making processes are opaque and convoluted, and democratic institutions in the widely-understood sense don’t exist.  The civil society organisations that are permitted have only very narrow room for manoeuvre and the press is there to defend the Communist Party’s line much more than to hold it to account.

Ugaz may nonetheless be simplifying what is actually something that is very complex. Off the shelf anti-corruption tools don’t exist, and the ‘right’ way to tackle corruption will depend on the nature of the challenge to hand.  What works in one place may well have little impact in another and be counter-productive in a third jurisdiction.  The introduction of democratic institutions across Africa in the latter part of the 20th Century, for example, did little to help those states tackle endemic corruption.  Indeed, in some it actually made it worse.  Tackling corruption is subsequently messy, it’s time-consuming and it’s difficult.

Xi’s China will certainly have to find a ‘Chinese Way’ of tackling the corruption evident in the PRC.  That may well involve different ideas and different approaches to what we’ve seen elsewhere.  But that doesn’t mean that the baby should be thrown out with the bathwater. Much of what the CCP has been doing under Xi flies directly in the face of what anti-corruption thinkers tend to believe might work.  Small improvements in a global anti-corruption league table shouldn’t disguise that.




The Corruption Perceptions Index (CPI); much ado about nothing?

Earlier today Transparency International, arguably the world’s most well-known anti-corruption NGO, published the 2016 Corruption Perceptions Index (CPI). Despite taking plenty of criticism over the years, the CPI has become an integral part of the global anti-corruption discussion. Proponents and critics alike are rarely slow in coming forward to unpack its findings. SCSC Director Dan Hough analyses why the CPI manages to generate such controversy.

Some things are almost guaranteed to prompt a reaction. Transparency International’s Corruption Perceptions Index is one of those such things. Very few people sit on the fence and claim to be non-plussed by what has become the most well-known attempt to get a feel for how much corruption may plausibly exist. Some commentators use it as an indicator of how much needs to be done, others regard it as an (often forlorn) exercise in quantifying the unquantifiable. Very rarely is anyone ever neutral about it.

Before launching in to any critique of the CPI, it’s worth remembering what exactly the index is and where it has come from. The CPI is a composite index. A variety of data sources are used to create what is in effect a poll of polls on perceptions of corruption in around the world. Data is gathered from surveys of business people and country experts with the aim of measuring perceived levels of public sector corruption. TI provides a detailed account of where its data comes from and also how it uses it, and this is accessible via TI’s own website.

The Results

The CPI was first published in 1995 when it included 41 countries, with New Zealand achieving the best score (i.e. nearest to 10) and Indonesia the worst (nearest to 0). Over the years TI has changed the way it presents and indeed produces its results, with the range of scores now stretching from 100 (no corruption) to 0 (complete corruption).

This year’s CPI included 176 countries (although in 2011 it included as many as 183), with Denmark (90) and New Zealand (90) at the top of the pile and Somalia (10) at the bottom (for the tenth year running). Predictably, the best performing countries share a significant number of characteristics. They are generally open, liberal democracies with a free press. They embrace the notion of transparency, therefore helping citizens see where their hard-earned tax money gets spent. They have independent judiciaries, and all support long-held assumptions about increased accountability leading to lower levels of corruption.

The Scandinavians always do very well, as do countries in western Europe more generally. There are, however, always interesting outliers when compared to their regional peers; Singapore (7th in 2016) regularly appears in the top 10, whilst Botswana (joint 35th) leaves many of its African counterparts in its wake. The countries at the bottom, meanwhile, also have lots in common; leaving North Korea (12) to one side, they are war-torn and bordering on the ungovernable. The fact that South Sudan (11), Syria (13), Yemen (14), Sudan (14) and Libya (15) are immediately above Somalia (10) is evidence of that.

Criticisms of the CPI

The CPI’s prominence has certainly not shielded it from criticism. Indeed, criticising the methodology that underpins the CPI has become a veritable cottage industry (see here, here and here). These criticisms cover a number of issues. To start with, boiling down a country’s corruption troubles to one score is, to put it mildly, methodologically problematic. The type, scope and extent of corruption evident in, say, the city administration of Chicago is likely to be altogether different to that which you’ll find in, for example, rural Arizona. Computing one score to accurately cover such variety is always going to be very difficult. Plus, if a state were ever to register a score of 100, what precisely would that mean? What does a country that apparently has no corruption at all actually look like? Or, conversely, what would a country that scored 0 (totally corrupt) look like? Any discussion of utopias usually ends in disagreement, and it is very likely that that would be the case here.

Furthermore, measuring concepts such as democracy, justice, fairness and indeed corruption is hard at the best of times, but those who do it well acknowledge that their attempts are always approximations. Indeed, statisticians have developed their own language to discuss the problems in getting these measurements right. Even though the CPI’s methodology has undoubtedly got more rigorous over the years, none of this is overtly acknowledged. This has led some researchers to cast doubt on whether this data can be put to any real use at all.

One particular challenge concerns the problem of defining corruption. It is not always clear what respondents actually understand the term corruption to mean. The terms bribery and corruption often appear to be used interchangeably plus responses to the various surveys are very likely to be shaped by – whether directly or indirectly – the assumptions and attitudes of the western business community. That is the case for the simple reason that the majority of people asked have roots in this particular milieu.

The problem of perception

The CPI also measures perceptions of corruption rather than corruption itself. TI regularly and consistently acknowledges that this can be problematic. While knowing more about how citizens perceive a phenomenon certainly has its uses, it is also plausible that perception and reality might differ considerably. As Ritva Reinikka and Jakob Svensson succinctly note “perception indices raise concerns about biases”. These potential biases may well mean that the CPI is actually (and inadvertently) distorting reality, and simply reinforcing stereotypes and clichés.

A further limitation is that the CPI focuses on perceptions of public sector corruption. In other words, the corruption that takes place in and around governments and public servants. It says nothing about corruption in private business. The rigging of Libor (the rate of interest that banks charge each other when lending money between themselves) in Britain, for example, or the VW emissions controversies in the United States (and indeed elsewhere) involve private actors, but they have very real public impacts, whether on the interest rates that people pay on their mortgages or on public health.

Babies and Bathwater

These problems have prompted a significant number of analysts to be quite scathing about the CPI. Steve Sampson, speaking for many in the development studies community, is sceptical of what he regards as “corruption becoming a scientific concept”. Even fellow quantifiers such as Stephen Knack have criticized some of the statistical techniques that TI has employed in the past. Indeed, Anwar Shah and Theresa Thompson leave no one in any doubt as to how grave they think the CPI’s methodological shortcomings are when they state that “closer scrutiny of the methodology … raises serious doubts about the usefulness of aggregated measures of corruption” and “potential bias introduced by measurement errors lead to the conclusion that these measures are unlikely to be reliable, especially when employed in econometric analyses”. Knack’s careful dissection of the CPI raises further significant issues about the independence – in a statistical sense – of the data used, claiming that many of the ‘statistically significant’ changes that TI claims to have uncovered would not in reality be so if “appropriate corrections for interdependence” had been made.

Facing down the criticisms

For its part, TI has certainly tried its level best both to be open about the methodological shortcomings of the CPI (as well as its other corruption indices) and also to adjust them wherever possible. The founder of the CPI index, Johann Graf Lambsdorff, for example, is careful to acknowledge some of the methodological issues inherent in all composite indicators and he is always careful to describe changes in country scores from year to year as changes in perceived corruption rather in actual corruption levels.

TI has also tacitly admitted that the CPI has its limitations by the very fact that it has developed a whole host of other indices – such as the Bribe Payers Index and the Global Corruption Barometer – to look at both the perceptions and experiences of specific groups of stakeholders (ranging from businessmen to households).

And yet, all these criticisms not withstanding, the CPI has done one indisputable thing; it has put the issues of corruption and anti-corruption well and truly on the policy map. As Andersson and Heywood observe;

“We should not underplay its significance in the fight against corruption: its value goes beyond the stimulation of research activity, since the publication of the CPI each autumn has generated widespread media interest across the world and contributed to galvanising international anti-corruption initiatives, such as those sponsored by the World Bank and the OECD”.

Even staunch critics of the quantification of corruption have begrudgingly admitted that “whatever its limitations” the development of the CPI has “undoubtedly done much to promote the anti-corruption agenda”. It is also doubtful that any of the more nuanced indices that both TI itself and other organizations have developed would have seen the light of day if the CPI hadn’t existed before them.

The CPI certainly doesn’t represent the gospel in terms of global levels of corruption. There are plenty of problems with its methodology and subsequently with its findings. But anyone who takes the detailed numbers produced in the CPI too seriously is missing the point. The CPI, for all its sins, has a constructive role to play in helping us think just a little more about how we can better measure corruption and how the battle against corruption can subsequently be taken forward,

Dan Hough

University of Sussex

How much corruption is out there? Well, that depends …

In recent years attempts to measure how much corruption exists have blossomed. Some approaches are based on perceptions of corruption, some on the experiences of individuals. Others use a range of proxies to measure what they argue might be corruption. Subsequently we now have broad aggregate indicators such as Transparency International’s Corruption Perceptions’ Index (CPI) and the Control of Corruption variable in the World Governance Indicators (WGI), as well as much narrower and more focussed indices looking at, amongst other things, state capture, levels of financial secrecy and the quality of the business environment.

One of the most widely-watched of these attempts is TI’s Global Corruption Barometer (GCB), the latest editions of which were published late in 2016. The development of the GCB in 2003 was an attempt to move the previous emphasis away from seeking the opinions of experts and business leaders and on to ordinary people. The GCB is a public opinion survey that in 2013 asked 114,000 citizens in 107 countries about their “direct experiences with bribery” before going on to detail citizens’ “views on corruption in the main institutions in their countries”. The latest round of the GCB, published in stages through 2015 and 2016, is equally as expansive. The intention of the GCB is to bring ordinary people in to a field where their experiences and perceptions have traditionally been neglected.

Corruption is (apparently) everywhere

The GCB’s findings give plenty of food for thought. In 2013 more than one in four people around the world reported paying a bribe at some point over the last 12 months. The police and the judiciary were apparently the most bribe-prone institutions. Over half the people surveyed regarded their government as acting in the interests of preferred groups rather than the citizenry at large.

The picture was in many ways just as downbeat in the apparently less-corrupt western world. In the 2016 survey, for example, 53 per cent of citizens across 42 European and central Asian countries believed that their government was doing a poor job in fighting corruption. Perceptions were particularly gloomy in Spain (where 80 per cent of people thought their government was doing badly in this regard), Italy (70 per cent) and France (64 per cent).

Alongside data on questions like these, the 2013 survey also revealed that 65 per cent of New Zealanders – a country that is regularly vying for the top spot in the CPI – felt that over the last two years the level of corruption had increased in their country, whilst only 5 per cent felt that it had decreased. The picture was little better over the Tasman Sea in Australia; over the same time period 59 per cent of Australians thought corruption had worsened whereas, like in New Zealand, only 5 per cent thought it had got better. Things were not dissimilar in Europe; 57 per cent of Germans believed that there was more corruption in their country than two years previously whilst only 8 per cent thought there was less. Furthermore, 65 per cent of Germans thought political parties were in general corrupt, 54 per cent thought the same of the media whilst 49 per cent thought civil servants were either ‘corrupt’ or ‘extremely corrupt’. Sobering numbers.

Perception and reality

The GCB has now firmly established itself as an important part of the measurement mosaic. To TI’s credit, it reacted to criticisms of some of its other indices (principally the CPI) by introducing a detailed survey that helps bring people on the street back in to the discussion. Data such as this does nonetheless highlight two further potential problems.

Firstly, sceptics such as Claudio Weber Abramo have illustrated that opinions on corruption as revealed by the GCB “are strongly correlated with opinions about other issues”. It may well be that rather than picking up specific attitudes to corruption the GCB is tapping in to a much larger worldview. Indeed, Weber Abramo claims that the correlation is so strong that we could actually save ourselves the time and effort of measuring public opinion in this area by simply reading across from data on attitudes to things such as human rights and violence.

Secondly, many of the things that we think can help predict levels of perceived corruption (economic development, press freedom and so on) are not good predictors of how much corruption people claim that they personally have witnessed. In other words there is at times a considerable difference between the amount of corruption that people perceive to exist and the amount that they themselves experience. Only 1 per cent of Australians over the 12 months to 2013 paid a bribe, for example, yet 53 per cent think corruption is a problem. In the 2016 survey no Britons (!) reported paying a bribe, yet 57 per cent of people thought that the UK government was doing badly at fighting corruption.

This discrepancy could, of course, be a reflection of citizen awareness of grand corruption that they themselves don’t experience directly. The differences between what people think and what people experience are nonetheless often very large indeed. Furthermore, the size of these differences can differ noticeably from place to place. Such discrepancies are not simply the domain of the GCB; Eurobarometer surveys reveal similar patterns. As Paul Heywood has noted, 74 per cent of European citizens surveyed in the 2012 Eurobarometer surveys believed corruption to be a ‘major problem’ yet only 8 per cent reported any experience of bribery or attempted bribery.

Where to next?

Data on corruption is now in abundant supply. But it would be disingenuous to claim that analysis of it is an exact science. All of the organisations that produce indices (no matter whether they deal with perceptions, experiences or proxies of corruption) point out, to greater or lesser extents, the weaknesses in their respective approaches. TI and other organisations in the data-producing field may, in other words, talk up their respective efforts to quantify aspects of corrupt practice (and who can blame them), but they do also realise that they are dealing in imperfections.

The GCB offers a plethora of interesting information.  But, as with all corruption indices, we should be careful of reading too much into it. Someone, somewhere across the UK, for example, will have paid a bribe over the last 12 months, and just because the GCB didn’t reveal that statistically doesn’t mean that bribery doesn’t exist. The GCB is merely illustrating that everyday bribery is not a common facet of life in the UK whereas it is in other parts of the world.

Corruption is complex, multifaceted and riddled with nuance, and this makes quantifying it very difficult indeed. The GCB can help us spot trends and illustrate the scale and scope of particular types of corruption. But it is just an indicator not a statement of fact. If used carefully, the data can have a constructive role to play in helping us think just a little more about how the battle against corruption can be taken forward. What it shouldn’t be is taken as the gospel truth.

Dan Hough

University of Sussex