Skip Navigation

Looking Back: Why some OGP commitments don’t get implemented

Joseph Foti|

¿Quieres leer este blog en español?

Here in the open government community, we talk a lot about “failing forward” or “learning from failure.” This is part of what makes working for OGP so great. Governments and civil society can admit momentary defeat, and try again, maybe another way.

Being part of the Independent Reporting Mechanism (IRM), I see a lot of successes, but I also see a lot of failures. Recently, a number of different colleagues–civil servants curious about their peers, multilateral development banks, civil society leaders, technical experts–have asked me why it is that OGP commitments don’t get implemented. Similar questions were recently raised by Steven Adler and Nathaniel Heller with concern that implementation of OGP commitments is too low.

The staff at the IRM thought these are good questions, actually–worth a bit of time. We looked quickly at a number of IRM reports, trying to find how the IRM researchers (our colleagues at the country level who write the reports) explained away non-completion.

So, we present to you a decidedly unscientific list of all of the reasons OGP commitments do not get implemented. Some are funny, a few are depressing, and some are both. Most have workarounds or, at least, should give you an appreciation of the difficulty of implementing open government commitments. For those who have the time and energy to do a more in depth analysis, remember that the IRM always has an open door for researchers wanting to take the analysis a step deeper. (Hint:We would love some help in taking this analysis further.)

The examples below are illustrative, rather than direct references, to protect the innocent and the positive risk-takers. (Remember, OGP encourages ambition in its commitments, even if that means not every commitment is implemented in full. At the end of the day, it will remain a matter of debate as to what percent of OGP commitments in a plan should be achievable and what should be riskier “stretch” commitments.) So without further ado a first attempt at a list…


  • Vague or not really a commitment; no clear actions. For example, “We pledge to promote public participation,” is nice, but it doesn’t quite cut it as the “ambitious concrete actions” described in OGP’s founding documents.
  • Start date happens after the action plan ends (or IRM report happens). Believe it or not, some commitments have a start date in 2017 for a 2013-2015 action plan.


  • Commitment is against the law. This happens sometimes. There may be strong political will and the commitment might be great, but on closer examination, it may not work with pre-existing laws. Some open data and health commitments, for example, have run into privacy laws (often for good reason) and require a bit of rethinking for the next action plan.
  • Commitment is redundant with other actions carried out by the government in the past or in another agency. Sometimes, governments or agencies discover that another branch of government has already completed the work or that another agency is working on the issue so it might be dropped from an action plan.
  • Commitment does not fit the economic or social context. Take, for example, the utility of an open budget deficit clock in a country with lots of power outages and low literacy rates.

Institutions and coordination

  • Implementing institution has no mandate to carry out commitment. Similar to illegal commitments above, the agency may not have power under the law to carry out the commitment. Or it may simply not have the commitment implemented into its budget, work plan, or other key strategic documents (such as the national development plan).
  • No one is responsible. During the early days of OGP, few action plans had responsible agencies named for the activities. Fortunately, the norm is changing and many (or even most) new action plans name a lead agency.
  • No one is holding anyone else responsible. IRM researchers are regularly surprised at how few governments have internal tracking systems (like spreadsheets) to check in with commitment leads.
  • Implementing institution does not know that the lead institution has committed it to a commitment. This one is perhaps a little too common. Fortunately, there’s a cure: countries with interagency working groups tend to solve this issue.
  • Lead institution has no mandate to compel implementing institution. In some countries, agencies and legislatures are “volunteered” to carry out commitments, but the lead OGP agency may not be able to actually compel the other (often more powerful agency) to do things. In some countries, OGP is led by the public administration office which can do good work but which is unable to get the treasury or foreign ministry to implement any commitments. In other countries, the president may commit her or his country to pass a freedom of information or corporate transparency law, but this is actually in the hands of an independent legislature. Similar examples can be given for commitments on subnational governments, especially in federal states where subnational governments retain some sovereignty over governance issues. Finally, sometimes civil society organizations are listed as co-implementers and they lose interest. (Just this last point could inspire an entirely new “Failure Blog.”)


  • No budget. This is a common enough one that needs no example. Perhaps the more interesting question comes when one digs one level deeper to find out why there is no budget to begin with. In a number of African countries, the budget is closely tied to approval by the president’s cabinet. Without formal approval of the OGP action plan, commitments may remain unfunded mandates. This shows that budgeting for open government and its consequences are often very difficult to separate from political buy-in.
  • Commitment is too technically difficult. Skills and knowledge are one of many limiting factors. This is a critical area for peer learning, working groups, and multilateral organizations to help out on and is a big part of the value added of OGP.

Political support

  • Commitment or action plan does not have high-level support. In some countries, ministers or other politicians who are needed to get big reforms done are not engaged with the Action Plan. Sometimes, national action plans are out of sync with other key strategic documents like national development plans or legislative agendas.
  • Vested interests: the Commitment would disadvantage someone high in the administration (or legislature) who would suffer loss due to commitment being completed.
  • A change to a new administration that either (a) does not know about OGP or (b) does not like the commitments.
  • Change in lead agencies (or staff) which does not like the last lead agency’s commitments.
  • Civil society protest against a commitment. See some of the recent commitments on making health data public for examples.

Force majeure

  • Catastrophic event there (revolutions and natural disasters).

What are your thoughts? Did we miss an important cause of failure? Which of these have technical or financial remedies? Which are more political? What does that say about the role of peer learning and donor client relations?

Open Government Partnership