gov + tech = integrity ?

governments use technology to tackle corruption - does it work? no cheerleading, no doomsaying - just distilling the hard evidence wherever good research finds it

[govtechrity]

Un-control the crowd control- the ambivalence of crowd-sourcing citizen complaints

We have come along way since the heyday of enthusiasm with online crowd-reporting mechanisms that were meant to be a game-changer in the fight against corruption. They might just work in certain, very limited contexts, but more sinister applications have emerged since then. Increasingly there is a sense that it is governments, primarily of the authoritarian type, that are the prime beneficiaries of such crowd-reporting mechanisms. They can be deployed to give the surveillance state an excellent window into what their citizens think and get worked up about, while also offering new plays on old ways to monitor the behaviour of lower level officials. But not so fast.  .  As a fascinating new empirical paper on Chinese complaints systems shows, even the most sophisticated reporting systems are prone to subversion by lower level monitoring agencies. Glass half full or half empty? On the one hand, the gaming of the reporting protects low-level officials from the over-reaching gaze of an authoritarian state. Yet then again, it layers corruption onto corruption, allowing corrupt officials to activate their cronies to conceal reports about them and thus shielding them from accountability for their behaviour. 

Pan, J., & Chen, K. (2018). Concealing Corruption: How Chinese Officials Distort Upward Reporting of Online Grievances. American Political Science Review, 1-19.

[govtechrity]

SMS notification for more water accountability - which SMS?

A fascinating experiment/study in India that set up a system to inform citizens in areas of intermittent water supply with timely sms notifications from the water works when water would become available. The idea was that besides helping water  consumers to adjust their behavior more effectively such a system would also encourage citizens to feel more empowered, complain directly to the water works if things go wrong and overall develop a more favourable sense of government responsiveness and performance. So did it work?

Surprisingly no significant outcomes to that effect could be measured and one root cause seemed to be that what at the outset looked like a straightforward information transmission pipeline from valve operator to customer via a non-profit intermediary broke down on several points with many valve operators  not complying or sending wrong information and many clients at the receiving end simply not receiving or consciously seeing the text message. So the devil is in the (upstream detail) even for simple tech information interventions that on paper appear quite straightforward.

Kumar, T., Post, A. E., & Ray, I. (2018). Flows, leaks and blockages in informational interventions: A field experimental study of Bangalore's water sector. World Development, 106, 149-160.

image cc by  Kiran Jonnalagadda

[govtechrity]

A tale of two ports - and how electronic workflows do make a difference

Durban container terminal, cc Xeon

E-customs, the use of ICT’s to make customs and thus trade work better,  has been one of the earliest application areas (ASYCUDA anyone?) for novel information technologies even predating the Internet revolution. And quite predictably these technologies are also expected to work their magic and help root out corruption. Yet, precious little in-depth research has been done to probe how this has played out. One very notable exception is the seminal work by Sequeira and Djankoff on ports in Mozambique and South Africa documented in a 2014 paper and now included in a book by Ackerman and Lagunes. 

The authors examine in astounding detail through minutiae ethnographic observations and lots of data crunching how and for what purpose specific ports are being selected and to what extent bribery and other types of corruption play a role.Among their many fascinating finding also surfaces the insight that online filing systems in Durban are associated with lower levels of bribe paying compared with Maputo where no such online systems were available and more human interaction was required. This is certainly not an unexpected insight and there are big question marks related to the broader external validity of this finding (e.g. an attempt to install even a very similar system in Maputo might play out very differently), yet great to see meticulous in-depth research validating this conjecture.  

Sequeira, S., & Djankov, S. (2014). Corruption and firm behavior: Evidence from African ports.Journal of International Economics, 94(2), 277-294.

Sequeira, S. (2015) Corruption and trade costs. In: Lagunes, Paul and Rose-Ackerman, Susan, (eds.) Corruption: Global Influences, Politics and the Market. Edward Elgar.

[govtechrity]

E-filing taxes in Tajikistan: a shield against extortion, but not so much a sword against collusion

Palace of Nations, Dushanbe, Tajikistan  cc: RyansWorld

Tax collection is a hotspot for corruption offering multiple opportunities to lower tax burdens in return for bribes to inspectors. Electronic filing of taxes promises to reduce the human element - can it therefore also help reduce corruption? 

An interesting experiment from Tajikistan suggests so, however with some qualifications. Firms were offered the opportunity to file their taxes online and the ones that reported being victims of extortion from tax officials did so in greater numbers. However, firms that are prone to tax evasion appear to shun the system, suggesting that there will be some adverse selection unless e-filing can be made mandatory for all. 

Oyebola Okunogbe and Victor Pouliquen (2017):  Using Technology to Improve Governance: Evidence From the Introduction of Electronic Tax Filing in Tajikistan, abstract available here

[govtechrity]

Fighting ghosts successfully - electronic funds in social benefit schemes in India

A very interesting example of how electronic anti-corruption measures work in concert with other integrity measures to cut out corruption. Researchers teamed up with the government of Bihar, India to assess the effects of financial management reforms that saw the introduction of electronic funds disbursement alongside better tracking efforts and reducing middlemen for a workfare programme. The results: a reduction of fake households and less personal enrichment by local officials as evidenced by lower averages in asset/income disclosures. 

Banerjee, A., Duflo, E., Imbert, C., Mathew, S., & Pande, R. (2016). E-governance, Accountability, and Leakage in Public Programs: Experimental Evidence from a Financial Management Reform in India (No. w22803). National Bureau of Economic Research. 

[govtechrity]

mixed news for the big picture: no significant impact of e-procurement on bribes by business

I am always kind of skeptical about all-too-abstract macro-structural regressions that show or do not show how ICT related indicators correlated with governance attributes and then draw sweeping conclusions about a alleged (non)impact. This one however is an exception, since tightly focused on examining how the actual introduction of electronic procurement systems has or has not affected the reported bribe paying of businesses. And the result at this level is a sobering one: no significant effect could be discerned. Now it would be super-interesting to see the entire plot of countries analysed and zoom in on some positive outliers along the lines of the approach that we have proposed in joint research with DLP here. Are there any countries that have bucked the trend and actually seen a significant decrease in business bribery with the advent of e-procurement and from there digging deeper - what makes their stories different? Unfortunately this part of the story is not examined further in the paper.

Kochanova, A., Hasnain, Z., & Larson, B. (2016). Does e-government improve government capacity? evidence from tax administration and public procurement. Evidence from Tax Administration and Public Procurement (April 28, 2016). World Bank Policy Research Working Paper, (7657).

[govtechrity]

squeezing corrupt rents in medical e-procurement? the case of Chile

An interesting (and rare) pre-post, longer-range investigation into how procurement practices change with the introduction of an electronic tendering system. Chile made its electronic Chilecompra mandatory as early as 2004 and this study covers a five-year period of procuring medical supplies. While finding mixed impacts on competitiveness and frequency of tendering the authors do detect an 8% reduction in prices through e-procurement that they attribute to reduction in corruption. 

Raventós, P., & Zolezzi, S. (2015). Electronic tendering of pharmaceuticals and medical devices in Chile. Journal of Business Research, 68(12), 2569-2578.

[govtechrity]

e-procurement in Russia: making corruption more visible, but not eradicating it

A fascinating new paper uses the immense data footprint generated by e-procurement systems, in this case more than 3 million detailed records of procurement auctions in Russia to detect patterns of corruption. The result: an estimated quarter of procurement auctions analysed are affected by collusion and up to 8 percent of auctions suggest that information on other bids was leaked to the winning bidder, one of many types of corruption in procurement. 

And the bigger picture insight: e-procurement as deployed now might not be able to stamp out corruption and collusion but it makes it certainly easier to detect such patterns, paving the way for designing more corruption-proof systems in the future.

Korovkin, V. Corruption vs Collusion: Evidence from Russian Procurement Auctions. RSSIA 2016 Research Projects, 81.

[govtechrity]

biometrics identification and integrity of elections - perhaps a piece, but not the entire puzzle!

As mentioned in several earlier posts, high hopes surround the potential of biometric identification systems to help fight corruption. Emerging evidence is encouraging, but it also points to biometrics as being one piece of the puzzle, rather than the overall solution in a box. Here another very interesting research piece that confirms this picture by analysing how the introduction of biometric identification systems played out in Ghana in 2012. As it turns out, the new biometric machines happened to mal-function much more frequently in places with less unannounced electoral monitors and the data also showed that it was these very places that also experienced higher incidences of election fraud. So no biometrically-enhanced integrity where conventional electoral monitoring is not provided.  

Golden, M., Kramon, E., Ofosu, G., & Sonnet, L. (draft 2015) Biometric Identification Machine Failure and Electoral Fraud in a Competitive Democracy.

[govtechrity]

Egov to tackle corruption in land and trains: mixed evidence for Bangladesh

Very interesting synthesis of two e-gov cases from Bangladesh,were new systems for computerized land record management and electronic sales of coveted train tickets were hoped to tackle corruption in these settings. This overview paper just published by Hasan Baniamin of U4 nicely summarizes the corruption dynamics for both situations and shows how electronic sales helped cut service point corruption, while electronic land records have done little to reduce related corruption issues, since they have not dented the gate-keeping powers of land officials.

[govtechrity]

Toothpick: 1 - High Tech 0: Why cutting corruption via electronic benefit transfers turns out to be harder than we think

A fascinating new paper from Vivek Srinivasan and colleagues at Stanford’s  Freeman Spogli Institute that looks deeply into how direct cash transfers of social benefits in India supported by a very sophisticated tech architecture play out in practice. The result: high hopes of cutting out corrupt middlemen and making transfers instantaneous are not being realized. Front-end gatekeepers are finding insidious new ways to assert their power and continue to extract bribes by delaying payments, forging entries or demobilizing tech equipment (in one case using toothpicks). An excellent empirical word of caution that highlights that high-tech-direct-cash does not equal corruption free - and perhaps helps us design not only better tech, but also better  incentive systems for the future.

Srinivasan, V., et al. (2014):  “IS THERE SUCH A THING AS A “DIRECT” CASH TRANSFER?”, CDDRL Working Paper, December 2014.

[govtechrity]

Crowdreporting corruption on steroids - and guess who does it?

(image cc Glogger)

Imagine an app that let's you report corruption - quite common these days. But imagine, it also let's you record and send actual video snippets of corrupt acts - mmh this is getting pretty interesting. And now imagine that it let's you capture and upload footage in real time to a dedicated server, while you record, so that confiscating your phone would not destroy the evidence - now we are really approaching cutting edge!

But wait it gets even better: guess who is behind this initiative? Some techie NGO? 

Nope. 

It is your local government, doing it right now in Delhi, where the  Aam Aadmi Party just swept local elections, riding high on an anti-corruption ticket. That is pretty amazing. Now let's see, if promise follows action and action actually leads to impact, but even the prospect of having a local government come up with such a cutting-edge and hard-hitting tech initiative to tackle corruption would have looked really quite far-fetched until recently. And the idea to push local government into integrating crowd-sourcing elements into their conventional citizen feedback and performance monitoring mechanisms is exactly one of the promising future pathways for crowd-sourcing corruption reports that I flagged in my recent ideas paper on this topic. 

Zinnbauer, Dieter, (2014): Crowd-Sourcing Corruption: What Petrified Forests, Street Music, Bath Towels and the Taxman Can Tell Us About the Prospects for Its Future

[govtechrity]

Taking on the rice mafia with the help of tech - the struggle for food security in Kerala

Few other corruption challenges are as directly linked to people's essential livelihoods as corruption that stymies public food distribution geared towards the poor. Can technology help stem the massive diversion of subsidized food that is estimated to leave more than half of the low-income recipients without their entitled ratios and at dire risk of food insecurity?

(cc IRRI)

A very interesting new qualitative case study on Kerala explores this question, looking at a state that has been at the vanguard of e-government in India and that directly targeted its sweeping computerization of the distribution system for subsidized food at curbing those leakages. So how does this play out?

The results are mixed for two main reasons: First, built-in monitoring mechanisms only target the point of sale to the end user and leave out what is a very complex supply chain where diversion is also taking place. Second, even the highly monitored retail sellers are under immense pressure to squeeze additional profits out of the system, since their business models have all but collapsed when far-reaching reforms to the food subsidy system sharply reduced the number of entitled customers. 

Even the current roll-out of biometric ID cards in India that is meant to become another key accountability pillar in this food subsidy system is not expected to address these two major shortcomings. This means pressures and opportunities for corruption will not go away, unless the system will be re-designed to also monitor the long supply chain involved and unless new viable business models can be identified for the retailers involved. 

Extremely interesting food for thought, although mostly presented as a field-research informed outlook into the future, while unfortunately no hard data on changes in actual scale and scope of related corruption and leakage is included, one of the cardinal problems of many studies in this field.

Masiero, S. (2015).Redesigning the Indian Food Security System through E-Governance: The Case of Kerala. WorldDevelopment, 67, 126-137. (unfortunately behind a pay-wall)

[govtechrity]

Sousveillance instead surveillance: tackling police corruption and abuse the wearable way

- in the US more and more police departments are considering the introduction of wearable body cameras that have been found to result in more disciplined behavior

- in Russia, dash cams have become the weapon of choice for drivers to challenge shake-downs from corrupt traffic policy or other scams. It is estimated that more than a million car owners are already using them.

- and most recently the Metropolitan Manila Development Authority  announced a very interesting "public-private" crossover effort: it plans to heavily subsidize the installation of dashboard cameras by public utility vehicle operators in the hope of curbing police corruption and improve the safety of public transport. 

Many open questions remain about the efficacy, proper rules and further evolution of these systems for what is often termed sousveillance, monitoring from the bottom-up, rather than top-down surveillance.The use of dashcams, for example has been declared illegal for privacy reasons in several countries. Yet, sousveillance certainly is an interesting addition to the toolbox for tackling some particularly pernicious forms of corruption and abuse. Excited to see where this all will go... - and how more research can help us design this for maximum accountability with respect for privacy. 

(drawing above c Steve Mann, Christina Mann and Mike Zandwyk)

[govtechrity]

Biometric IDs - too successful in tackling subsidy fraud in India!

Great evidence in a new paper by Prahbat Bamwal who analysed data from more than 23 million fuel purchase transactions in India and found that biometric identification systems help curb fraud in the allocation of cooking oil subsidies. Yet, sadly also a compelling case study of how corrupt politics trumps sensible governance and related tech projects: the very success of the system may have ushered in its own demise and led to it being abolished by the politicians that it had deprived of some very lucrative and convenient patronage power.

(Prabhat Barnwal: Curbing Leakage in Public Programs with Biometric Identification Systems: Evidence from India’s Fuel Subsidies, working paper November 2014)

[govtechrity]

verification technologies to tackle corruption in timber and medicines

Two recent fascinating blog posts over at the Global Anticorruption Blog discuss the potential for new verification technologies to help put a stop to fake medicines and illegal logging. 

While actual evidence on effectiveness is still scarce the two blog contributions do a very good job in illustrating the crop of new technologies that are increasingly being developed and deployed to verify the origins of products in sectors rife with corruption. Importantly, they also discuss related challenges (cost, granularity in the timber example), as well as  some innovative applications in the future (community feedback and fraud hotspot-tracing in the drugs example).  Great reads for inspiration both for research and practice! 

(pictures cc Arpingstone and Sproxil)

[govtechrity]

Cities of egov deviance: why Yerevan and Bratislava are so much more interesting than Stockholm or Copenhagen

Where are the best places to study whether technology use by governments is actually  making a difference in boosting integrity and accountability? Going local, down to city level is probably a good idea, in order to make the scope of the inquiry sufficiently narrow and manageable. Yet from there on it gets more tricky. ICT-overachievers such as Scandinavian countries and cities typically also excel in good governance, public trust making it difficult to figure out what causes what. Picking cities that go big on tech in what is an otherwise rather mediocre governance environment offer a much better way to track down any possible transformational change. But where are these places?