Steve Kelman argues that cross-agency coordination is not quite as quixotic as he once believed.
When I was a PhD student in political science, so many decades ago that I care not to say, one of the themes in my courses about American government was the utopian foolishness of attempting to get different government agencies to work together. One popular book at the time, Harold Seidman's Politics, Position, and Power, had a whole chapter called "Coordination: The Search for the Philosophers Stone."
Seidmans basic point was that feel-good talk of the need for "more coordination" ignored the fact that agencies operating in a policy area frequently had conflicting missions and interests, which could not "be reconciled through the magic of coordination." Coordination, he continued, "is rarely neutral. To the extent that it results in…a decision on some policy, course of action, or inaction, it inevitably advances some interests at the expense of others."
In these accounts, interagency coordinating councils were mainly a forum for what political scientists referred to as "bureaucratic politics," fights among different organizations to get more resources for themselves and more policy decisions to go their way. Members of these groups were "instructed delegates" there to defend their own organizations interests. They either fought with each other instead of collaborating or -- as a result of fear of controversy and a desire not to reveal any information that could be used by other agencies present -- accomplished nothing.
A few decades after I was in grad school, my colleague Gene Bardach of Berkeley, who in the early 1990s was starting work on a book on agencies working together, recounted that when he mentioned his project to a colleague, the quip came back: "Short book, huh?"
As a young graduate student who wanted to consider himself worldly wise rather than naive, I embraced without hesitation the message about the foolishness of trying to get agencies to work together.
I thought of these messages from decades ago as I read a recent Government Accountability Office report called Managing for Results: OMB Improved Implementation of Cross-Agency Priority Goals. (Blog readers who are familiar with GAO reports will doubtless not be surprised at the continuation of the title: "But Could Be More Transparent About Measuring Progress.") The report discusses performance goals, first established in 2012, where the same goal is shared by multiple agencies -- seven of them in "mission" areas such as cybersecurity and improving infrastructure-permitting processes, eight in "management" areas such as shared services and category management.
The message of the GAO report was that implementation of these cross-agency goals is proceeding fairly well. Mainly, an infrastructure has been put in place to facilitate cooperation. For each goal there is a "work plan that include(s) a brief description of planned milestones, milestone due dates, and the agency or interagency group responsible for implementing milestones." For each of the eight cross-cutting management goals, three meetings a year are scheduled, chaired by the deputy director for management, to discuss goal implementation. These are a counterpart in a cross-agency environment to the agency-level meetings that have become standards for tracking individual agency priority goals, a change in practice I blogged about last year.
Much of the infrastructure involves ways that either staff in governmentwide organizations, or informal groups among experts from the agencies involved, can share knowledge and learn from each other. For several of the goals, communities of practice have been established among experts inside each agency to share experiences and best practices.
A newly established White House Innovation Fellows program has assigned fellows to work supporting a cross-agency priority goal, to give the emerging generation of government leaders experience in cross-agency work. And Congress authorized agencies, with OMB approval, "to transfer up to $15 million" from their regular budget accounts "for purposes of improving coordination, reducing duplication, and other activities related to the implementation of the (cross-agency priority) goals.
As I read the GAO report, I paraphrased in my mind the message from an old Bob Dylan song, "I was so much older then, Im younger than that now." Because, lo and behold, my worldly wise cynicism from another era seems perhaps to have been overtaken by events. While collaboration across agencies has hardly taken over Washington, its tide seems to be coming in.
If there is something actually changing here (and I think there is), what is causing it? Probably a range of things. I wouldnt underestimate the influence of 9/11 and the message it sent about collaboration across bureaucracies that had been suspicious of each other. I also would note broader cultural changes that promote collaborative styles over king-of-the-jungle machismo. At a certain point, new phenomena start feeding on themselves as people begin to get experience acting in different ways and find that as they try it, they like it.
Finally, while the interagency work discussed in my grad school reading mostly involved perhaps-impossible efforts to get agencies to agree on policy priorities in a situation where the different participants had different missions and constituencies, it seems like more of the collaborative efforts one hears about these days involve doing something operational, rather than debating which policy to adopt.
We have also learned things about the potential benefits of collaboration that were missing from the books I read in grad school. In these writings, the case for coordination mainly involved avoiding program overlap, duplication of effort, or going at cross-purposes. Now there is increasing attention to a mission-related case for collaboration -- that the problems government is trying to solve transcend the boundaries and mission of one particular agency and might be dealt with more effectively by cross-agency effort.
And, in terms of developing solutions, an argument for cross-organizational collaboration is that the jostling of ideas within groups having members from many different perspectives, professional cultures, and experiences can produce more creative ideas and solutions than people from a single agency, who are more inbred.
A second argument is that, often, multiple agencies working on a problem may each have some tools or resources available that the other agencies do not, allowing them to be more effective in dealing with the problem if they work together. (When the departments of Veterans Affairs and Housing and Urban Development were tasked with working together on a joint goal of reducing veterans homelessness, they discovered that HUD could pay rent subsidies but had no ability to pay a new tenants security deposit, while for the VA it was the opposite.)
Related to this, cross-agency collaboration can give people working in an area in one agency information about best practices that have been developed in other agencies. Finally, once agencies genuinely get into a cooperation mode, they may discover that they can increase political support for their joint efforts because different agencies have different sources of constituency support.
There is beginning to be some academic , especially practitioner-oriented academic, literature on cross-agency collaboration that has moved (way) beyond the worldly wise cynicism I encountered when I was a graduate student. The work of Jane Fountain of the University of Massachusetts at Amherst, and a one-time Kennedy School colleague, is particularly noteworthy. Start with her 2013 report for the IBM Center for the Business of Government, Implementing Cross-Agency Collaboration: A Guide for Federal Managers.
NEXT STORY: After an 11-month vacancy, Census names a CIO