Monday, March 26, 2018

We Must Understand Policy to Change Policy

John Lennon says, ‘We all want to change the world’. If we specify what exactly we want to change, two very common targets are the private sector and government. Many advocates devote their lives and careers to influencing these two sectors.  However, commensurate with this dedication, we need greater conceptual understanding regarding the conditions under which government policy and corporate behavior change.  In other words, advocates need to grapple with when that change happens. 

In this post, I focus on my favorite book on government policy changes.  The single best resource I have found to date for understanding the conditions under which government policy changes is John Kingdon’s Agendas, Public Policy and Streams (though a colleague coauthored another great article here).  This book, which is primarily based on interviews with federal government politicians and officials in the late 1970s, helps to organize the many aspects of policy change into a basic conceptual framework. 

Stage 1: Agenda ----------à  Stage 2: Alternatives ------à            Stage 3: Policy
Problem & Politics                 Solutions emerge                             

The diagram roughly sketches the main elements of the conceptual framework of Kingdon.  The main idea is that there are three streams - problem, politics and policy alternatives – that must align for a policy to change.  In other words, each is a necessary but not sufficient condition.

In stage one, he first describes two principal ways that issues move to the top of the government policy agenda, which involves the problem stream and the politics stream.  The problem stream relates to the perception of the importance and urgency of a particular issue.  When politicians consider an issue as a national crisis, as follows a large terrorist attack or the spread of a disease, then they are generally moved to action. On the other hand, sudden political events can drastically shift agendas, most obviously swings in political party control of the government.  Either one of these events can bring about what Kingdon calls a ‘policy window’, in which an issue may see movement.  However both of these streams must at least support the issue in order for it to move forward.  

In stage 2, Kingdon describes a process by which a few solutions addressing the particular policy issue emerge from a policy community.  The community includes academics, policy wonks, civil servants, advocates, lobbyists and others, who jointly develop a large variety of potential policy alternatives.    Through debate, discussions, advocacy and many other actions, a few solutions will rise to the top.  For a solution to be seriously considered, it shouldn’t just be technically appropriate, but also politically acceptable, especially to the elected officials in power.

The main reason I like this simple framework is that it breaks down the messy policy process into more coherent phases, and offers insight into how each phase works.  Using this framework and identifying the dynamics of each phase, advocates should be able to work more effectively – for better or for worse.  Here are a few specific ways that the basic ideas of Kingston help.

Policy organizations aim to influence all three of these streams, though different organizations are more capable of influencing particular streams. An organization can seek to raise awareness of a problem, providing evidence of a public crisis and communicating that message either to the public at large or within the policy community.  Many non-profit advocacy groups focus here, highlighting social issues and emphasizing the urgency to address the problem, without in-depth analysis of solutions or a massive political coalition to push the issue.  Second, an organization can work to change the political makeup of the government, through mobilizing membership to vote a certain way or through threatening negative press.  Mass membership organizations such as teachers unions or hospital associations are effective here, given their widespread membership in many congressional districts, but so are media savvy organizations.  Third, an organization can engage with the policy community to develop solutions.  This area of work is the primary function of Washington think tanks, though universities are often on the cutting edge of innovative solutions. 

If a community of concerned actors cannot influence all three of these dynamics at once, then absent a sudden crisis, it is likely that policy change will not happen in the short term.  But at least the framework can help identify the main impediment for a particular issue to move forward, whether on the problem front, the politics front, or the policy alternatives front; and the community of concerned actors can focus greater efforts to address that impediment.

Moreover, if a policy organization or a researcher is flexible in their issue area and want to find relevance in a relatively short period of time, then understanding these dynamics and choosing an area of expertise accordingly can help.  For example, foresight of a surge in crime or an imminent housing crisis would suggest that policy makers will be looking at these issues closely soon.  Of course, in an environment in which a policy community has developed, a newcomer might cause tensions with the organizations that have spent years on a particular issue. 

On a more personal level as a researcher, Kingdon’s book also provides a sense of how research makes a difference.  I often lament that, no matter how good the ideas, research usually does not find its way into policy.  Kingdon recognizes the difficulty in discerning a direct path by which an idea comes to influence policy, but discusses how this community of scholars, practitioners and others is essential to developing policy proposals that are ripe to respond to the agenda that emerges. 

In sum, while none of these are revolutionary ideas, the framework helps organize and prioritize factors involved in the policy process in a useful and even actionable way.

Thursday, August 18, 2016

Anti-Corruption Organizations Require Systems Thinking

When asked why their country is not more developed, or why there is so much poverty, most people in the world blame corruption (see this blog).  Corruption is present in every nation, and the struggle against it is never complete. In this post, I’d like to make the case for bringing systems thinking into corruption work.

First, let’s start with an obvious point that is often overlooked.  Corruption is defined as the abuse of public office for private gain, but it manifests itself in extraordinarily diverse ways, in terms of the individuals involved, the kind of corrupt act, the potential damage to the public, and the sector in which it takes place, amongst others.  Here are a few types of corruption:

  • Mid-level officials steal stocks of life-saving drugs and sell them on the black market
  • The executive requires a large, fixed monthly payment from the police force, as a means of sharing corrupt earnings
  • Passengers pay an intermediary – and indirectly the immigration office manager – to jump to the front of an immigration line
  •  A local official, at the behest of the MP, adds names of political supporters to a list of safety net beneficiaries
  • Various officials take a portion of all construction contracts
  • Teachers skip school to work elsewhere, but still earn a salary

One simple conclusion emerges upon describing the various forms of corruption. Because it varies widely, both simple explanatory theories and broad generalizations about the effect of corruption tend to be inaccurate.  Corruption is not just cultural, or only the result of a lack of supervision.  Similarly, not all forms of corruption are horribly damaging for the public. 

Some have tried to argue that corruption is either good for development, or not a top priority (Huntington, 1968; or other blogs cited here). These arguments fail to acknowledge the variety of forms of corruption. Indeed, some forms could potentially help development under certain conditions. Other forms may be relatively benign.  However, one need only conduct a simple exercise to comprehend the pervasiveness of damaging forms of corruption: name a development priority area in which corruption is not the greatest, or among the greatest impediments to advancement.   It would be difficult to find such an area in the majority of countries. 

Given that reducing some forms of corruption is critical for development, the question is, what can be done about it?  Unfortunately, I think that many anti-corruption organizations suffer because they fail to adapt to the dynamic social systems that support corruption.  Systems thinking is an essential step in developing, adjusting and executing a strategy to fight corruption.

Systems thinking as an integral part of organizations is not new.  It was popularized by Peter Senge in his classic 1990 book, The Fifth Discipline.  Its essence is that organizations need to incorporate an understanding of the complex, dynamic systems in the world in order to fulfill their mission.  Corruption plays a role in many socio-political systems that must be understood to make corruption efforts effective.  (For an excellent study in some such systems, see The Wire, my favorite TV show, which provides an analysis in terms of the interlink between the drug trade, policing, and politics.)

Here is a simple example of a system in corruption.  In the majority of countries of the world, patronage is required in the lead up to an election, which usually involves distributing some free items (such as money, rice, or jobs).  In order to spread wealth, politicians must raise funds.  Most of these funds will be collected through corruption while in power, either in the form of money raised or favors earned.  Corruption therefore forms a foundation of power for politicians, via patronage networks; and political power then allows them to use the government apparatus to conduct corrupt acts.  Impeding the various forms of corruption that support the patronage networks would therefore threaten the political survival of those who manage them.

                                                         Political Power
                                                         |                      ^
                                                         |                       |
                                                         |                       |
                                                        V                      |
                                             Corruption --------> Patronage

This very basic system diagram – which represents systems thinking - has important implications for fighting corruption.  One advantage is that it lets you know what you’re up against.  In this example, it suggests that political leaders have a strong interest in maintaining some forms of corruption, because their political power – and therefore their ability to be corrupt - depends on it.  As a result, an anti-corruption organization that focuses on the forms of corruption that support patronage can expect a strong reaction from political leaders.  Activists need to develop a contingency strategy to respond to the various tactics that politicians may take to undermine their activities.  Or, if an anti-corruption project seeks to reinforce checks and balances that limit the types of corruption that are needed for patronage, the effort will likely face opposition from politicians and their allies.

The system diagram also indicates that, for many forms of corruption, real reform would require a weakening of the strong, dynamic systems that rely on corruption, such as these patronage networks.  In the short or even medium term, this is unlikely.  If corruption is reinforced by other elements of the system, then a transformation of the system may be necessary.  Therefore, the fight against corruption should be seen as a long-term struggle that requires a multi-pronged strategy. 

Similarly, an immediate approach that seeks to grab low hanging fruits should begin with a contextual analysis to determine the precise forms of corruption that are not supported by these strong and dynamic systems. If there are not powerful individuals opposing the anti-corruption effort, the likelihood of progress should increase.

The diagram also suggests that a success in reducing corruption in one area may actually increase the corruption elsewhere.  Patronage requires resources, so even if an anti-corruption efforts is successful, the necessity will push those involved to seek out other corruption opportunities.  Such an effect is frequently discussed in economics, with the analogy of stepping on a balloon.  The net effect may be the same for the public, thus neutralizing any gains won through the successful reduction of corruption in the particular area.

Instilling smart approaches to fight corruption are critical.  Too much of this work takes a standard, technocratic approach of ‘capacity building’ for an anti-corruption commission and supreme audit institutions, or generally promoting transparency and open data.  More targeted approaches with systems thinking is needed to increase the effectiveness of anti-corruption work.

Friday, August 12, 2016

A Bangladesh without Foreigners

Here is a recent editorial I wrote for The Daily Star in Bangladesh.

One of the natural results of the recent terror attacks in Bangladesh is contemplation.  Some are rightly focused on the immediate need to understand how these unprecedented acts of violence could happen, and how to stop them.  But the fact that one major attack targeted foreigners in Bangladesh makes me wonder, what would Bangladesh be like without the foreigner presence?

I first came to Bangladesh in 2002, and have lived in the country for several years and visited numerous times since then.  This is the first time in memory that many of my foreigner friends and colleagues in Bangladesh are hesitant to stay in or visit the country.   As we foreigners count our blessings that we were not among the victims at the Holey Bakery, we also recognize that it could have easily been us that night.  The fear is amplified by both the brutality and the utter lack of sense behind the attacks. 

In the recent past, few major areas of economic and social progress that would have happened without a partnership with a foreign individual or organization with a presence in Bangladesh, even if all progress was completed with Bangladeshi leadership.  The main markets for the garment industry are the US and Europe, whose buyers have large offices in Dhaka with other buyers visiting the country frequently.  The aid industry supports many government programs, as well as the enormous NGO system in Bangladesh.  Academic partnerships have similarly been crucial in creating life saving health institutions such as ICDDR,B, involved in life-saving interventions that include  ORS.  Foreigners, particularly from Sri Lanka and India, contribute greatly as skilled employees for a number of industries, particularly garments.  This is not to say that the record is perfect, with numerous disastrous aid projects and imposed, destructive policies, amongst other negative effects; but the partnership has been positive on the whole.

As we look to the future, these areas of progress have been firmly established in the country, so even if many foreigners would depart, it is difficult to imagine they would vanish.  But one wonders, which other partnerships for progress are endangered?  Where are the major new areas where a partnership with foreigners in Bangladesh can make further contributions?  I would argue that four areas are vital to maintain growth and ensure Bangladesh progresses further as a Middle Income Country.

New Businesses

While the garment industry is strong, sustained growth requires the development of new sectors.  Export of light manufacturing goods is one of the most promising sectors, but as with garments, this requires the presence of buyers in the country, who will provide business opportunities and work with manufacturers to improve their capacity.  Similarly, the growing technology sector of Bangladesh should be greatly enriched through international business and the exchange of ideas.  Reliance on the garment industry to sustain growth is a risky approach for the economy. 


To continue developing, Bangladesh requires major investments in infrastructure.  Everyone knows the current infrastructure in Dhaka is woefully inadequate.  Indeed, several of the victims of the terrorism were reported to be visiting Dhaka to advise the government on the construction of a badly needed metro system.  Foreign governments and other institutions are needed to both provide advice and financing for infrastructure. 

A Skilled Labor Force and Higher Education

Most large businesses in Bangladesh identify the lack of highly skilled, well-educated employees as a key constraint, leading many to offer such jobs to Sri Lankans and Indians.  These foreigners play a critical role in meeting this need.  At the same time, it is clear that Bangladesh needs to further develop its higher educational institutions.  While foreign investment in universities has been limited, exchange with foreign universities will play a critical role in producing critical research for the country, as well as improving the quality of higher education.

Climate Change

Bangladesh has received worldwide attention as one of the main victims of climate change.  While some progress has been made, adapting to this challenge will require large-scale programs, with cooperation on a number of fronts.  The West should rightfully bear the costs of many of these programs.  At the same time, their presence is required to offer these resources, as well as cooperate to produce technical analyses to understand the varied effects of climate change.

The need to prevent terrorism in Bangladesh is obvious, but it’s important to keep in mind what is at stake.  As if a general sense of fear and sadness isn’t enough, there are also grave threats to the partnerships that have yielded economic and social progress of the country.  Re-establishing a general sense of security in the coming months and years may prove to be pivotal to the future of Bangladesh.  The country may continue its current, admirable progress or see that progress grind to a halt.

Tuesday, July 5, 2016

What's the Use of Development Plans?

Nearly two years ago, I was part of a team that drafted the background paper on governance, justice and human rights for the current Seventh Five Year Plan of the Government of Bangladesh.  Bangladesh still drafts a Soviet/ Indian style Five Year Plan, as led by the Planning Commission of the government.

This experience presented valuable lessons on planning, and especially how to think about the value of a planning exercise.  Ex ante, one would think this planning document is the master plan, directing practically all government initiatives related to development.  This is incorrect, as is the case with most planning documents. This is not to say that the document is useless; instead, it requires a different conception of the function of a planning document.  To explain this, let me present some important contextual realities.

First, I quickly realized that there is a major issue in terms of implementation. I thought that, to provide some context and accountability, I should dedicate a section of the paper to a stocktaking of what had been completed under the Sixth Five Year Plan.  The results were not encouraging.  Of the 72 activities related to governance, justice and human rights that appeared, just 13 had been completely implemented, 38 had been partly completed, and 21 had not even begun to be implemented.    Granted, these topics – governance, justice and human rights – tend to be a difficult and contentious area.  But this is a rather mediocre track record.

Second, I realized that the process for generating the plan does not generate sufficient ownership.  After starting the paper, I learned that the main Five Year Plan is written by an outside think tank for the government, then goes through an extensive review by the government.  This process clearly raises concerns regarding ownership by the ministries involved.  One such concern is first mover advantage: as long as the planned action – planned being the operative word – was not somehow controversial, it was quite possible it would remain in the document.  But if the idea didn’t stem from the responsible ministry, it’s quite possible the action would be tolerated in the plan, but with no intention of following through.  Ideally, a given ministry composes the relevant section of the overall plan, to ensure ownership.  Simply having the ability to edit an outside product is not sufficient to build on genuine intentions. 

That said, we also realized that the plan is not the main driving force behind government action, but it can help to serve as a rationalization for action.  If a well-placed individual in government wants to implement an activity, then its appearance in the Five Year Plan can further justify implementation.  This became the priority in developing our recommendations: identify genuine champions within the ministries, and if their ideas had a reasonable possibility of generating positive results, strongly recommend these actions. 

The experience brings up serious questions about the value of a plan in development organizations in general.  In the technocratic development practitioner’s view, a plan is made to set things in stone, then a budget and implementation indicators are developed, the project is financed, and then executed. However, this becomes ridiculous once one acknowledges two realities.  First, the implementing organization – or the branch that implements - often feels little ownership over the plan.  Such plans are generally developed by an outside consultant at the request of the funding organization, with the funding organization hiring that person and financing the work.  The consultant usually has little contextual knowledge, and few days in which to develop the plan.  Moreover, because the funding cycle takes so long, this exercise may have been completed years before implementation begins; further learning may have emerged, senior staff been replaced, and country conditions transformed.

Second, unless the program is a routine development intervention, everything changes once implementation begins. The plan may be comprehensive with a substantial consultative process, but when things begin, challenges emerge, and the intervention may prove to be ineffective.  

To make matters worse, the financing bureaucracies face difficulties in allowing for flexible implementation plans, at times demanding that the implementing organization stick with it.  In such cases, the implementing organization often pretends to go through the motions, so as not to interrupt the flow of resources; they may do so, even if they realize that the plan will not achieve the targets.  Moreover, the lack of flexibility discourages innovation and the testing of a variety of approaches to see what works. 

All of this is to say that a more realistic planning approach is needed.  We shouldn’t simply scrap plans, though they are sometimes useless.  Instead, a greater focus on process is needed, one that is long-term, inclusive, and on-going, that provides the opportunity for revisions along the way and for learning by doing.  Ideas such as flexible planning and the development of a learning organization (i.e. the Fifth Discipline) should be taken up by both financing and implementing bureaucracies.  I haven’t found a good example in development of flexibility that appears in this article from Harvard Business School, but I’m curious to find one.