Skip to content

Whither “whole of government?”

March 10, 2011

Some first-rate commentators have questioned the “whole of government” ethos we are well-known for espousing. Gordon Adams at the Stimson Center says,

Seems there are too many agencies, too many programs, too many “turfs,” to do a coherent “whole of government” approach to intervention or foreign policy decisions in general.

Instead, why not bolster the capabilities – at USAID, the State Department – we always say we’re lacking?

There’s no doubt we need to do this. The State Department and USAID are underfunded and ill-equipped for their basic missions, and have been for quite some time. Meanwhile, the Department of Defense has more lawyers/musicians than diplomats in the State Department. Its budget takes up half our discretionary spending. Everyone knows DoD is the aircraft carrier in the lake in any policy debate or interagency meeting.

The main argument goes: if we strengthen the State Department and USAID, maybe these agencies can do more things, and take on DoD in the policy process. Then we’ll get more diplomacy and development and push back the militarization of foreign policy.

Of course, there are more reasons. Whole of government is a term that arose from the last decade of war where the United States had to “stabilize” and “reconstruct” countries at great human and financial cost. Whole of government doesn’t really work when the only way to keep all the departments and agencies working together is through NSC coordination, which even then is still anarchy without close Presidential direction. Whole of government works when you have a small government and everyone knows each other, such as in Sweden, the United Kingdom, and Australia, but the U.S. government is too big.

The answer is “drop the whole-of-government fetish and figure out a sensible structure with clear strategic, budget, and reporting lines of authority.”

Put that way, these arguments are right. When you’ve experienced the despair of interagency meetings, whole of government doesn’t look so great.

However, there is some confusion over what whole of government means and what it is about. In 2008, we saw whole of government as “an approach that fosters government-wide collaboration on purpose, actions, and results in a coherent, combined application of available resources to achieve the desired objective or end state.”

(Note that whole of government isn’t tied to stabilization and reconstruction operations. Whether you’re working on counternarcotics, counterterrorism, climate change, energy security, U.S. strategy toward China, you can use a whole of government approach.)

There is a lot about whole of government that is good in all this messiness. Most importantly, the challenges the United States faces today defy any attempt to put them under one agency. If you have one agency working on one issue, it will face constraints. Even worse, the problem will change so that whatever that agency is doing, its effectiveness will diminish, and the problem will grow in some other way. This is why we advocate for the whole of government approach. You can’t solve contemporary problems without it.

What about all these problems with whole of government described by Gordon Adams and Todd Moss?

What they’ve described are the systemic impediments to the whole of government approach.

One major problem, which they’ve pinpointed, is that the only place to really do whole of government on big policy issues is the NSC. Of course, this is not going to work. It’s overwhelmed, it has no real authority, it’s thin on staff, it’s dominated by department and agency turf battles.

This is why we don’t advocate doing it all through the NSC. In fact, we were worried when the Obama Administration put all these new departments and agencies on the NSC without other systemic changes that would help it handle these new players.

We advocate that the NSC delegate the day-to-day work, year-to-year budget details, and so forth, of national security issues to empowered interagency teams. Empowered: co-located, the best people, given real authority, possesses resources, and wields prestige. We’re talking about Holbrooke-style teams respected up and down, from the PRT to the President, and widely, in State, DoD, ODNI, Agriculture, DHS, wherever. It’s about addressing the underlying causes of turf wars, inertia, and bureaucracy. This is real whole-of- government.  In addition, this frees up the NSC to work on matters of national strategy – stuff they should be doing now but are too overwhelmed to do.

What about size? Is the U.S. government too big to do these things? With around 2.5 million people in the major national security institutions, and more than 4 million people in the U.S. federal government’s departments and agencies, it may very well be that it is too big. IBM has about 400,000 employees. General Electric has about 304,000. Cisco Systems has about 65,000. (All are larger than the State Department. At least we know the State Department is not too big to fix.) Google has about 23,000 employees. The UK Civil Service is about 530,000 strong.

People often say size and bureaucracy go together, but the companies listed above are some of the largest companies in the world, known as the most innovative companies in the world, and are talking about things like the Globally Integrated Enterprise, more than making box-and-wire diagrams pleasing to the eye, but exploring things like networks and non-hierarchical structures. They’ve also had decades to grow to their size today. Several of them had to go through wrenching changes to become the kind of organizations they are today. From rigid industrial age bureaucracies to the information age innovation engines they are now. Size might not matter so much. It may not be that eternal impediment to the U.S. government doing whole of government right.

To finish up, should we strengthen State and USAID? Yes. But today’s problems – their complexity, the pace of change – provide even more reason for a collaborative and agile whole of government approach. Without it, expect more incoherence, crossed-wires, and sprawl, and problems that State, USAID, DoD, or any other department or agency cannot solve on its own. Let’s not confuse the clunky whole of government approach in today’s system for what we aspire – empowered and agile teams. We need that powerful whole of government approach to deal with the challenges of our time.

Note: Todd Moss comments on “whole of government” again, here, along with some keen observations about where aid and development need to go: outcomes-based budgeting, private-public approaches, moving toward transnational aid and development issues, moving into more sectors (all of which call for an even greater whole of government approach).

  1. Todd Moss permalink
    March 10, 2011 7:10 pm

    Thanks for highlighting our criticisms of W-O-G. I still don’t believe that just because the world is complex that its better to throw more agencies and more bodies at it (this holds for CT especially, but also for development). I’m not in favor of a single US development agency, but i do think a lot of the small programs sprinkled all over the place could be consolidated with big efficiency and coordination gains — for instance many of the private sector tools could be put under an enhanced OPIC or some new entity (I’m working on a specific proposal here). That seems a better option than hoping agile interagency teams will cooperate.

    One other point: The comparison between govt agencies and big companies is wrong. Google and other big companies can be efficient and large at the same time because they have a single mission and the market provides a check on waste. Government agencies have multiple missions and they tend to only grow, in size and number. Not the same.

  2. Chris Paparone permalink
    March 11, 2011 7:51 am

    I would also add that “W-O-G” is a misnomer in that our system of government is not designed to be unified. Now, if we were to say “whole of executive branch” (or more commonly known as the “interagency”) we might be more accurate as there is a single “chief executive officer” (i.e. POTUS).

    W-O-G is not practical in a two-party republican government that is designed to NOT be unitary, particularly when the legislative branch is in constant flux. Also, the judiciary cannot be ignoreed as a check on power.

    We are not designed to be an “imperial power” when it comes to overseas operations, especailly those requiring occupation and subsequent nation-building. And we we will not reform to do so — it is antithetical to the design.

    What we may be able to do is a better job of “internal defense” (homeland secrurity) where there is more likely to be a sense of unity of effort when it comes to disasters and the aftermaths (and even that is questionable).

    So the concept of “National Security Reform” itself will always become mired in conflicting values (inherent to our form of government). A practical way to approach change may be more like Charles Lindblom suggested in his seminal article (I found a link at

    We (the national security community) just have to become adept at muddling through and find policy entrepreneurs who can develop small steps that make sense. Ultimately we asses through “successive limited comparisons” to evaluate how things are going and change/reverse as we go. John Kingdon’s book, “Agendas, Alternatives and Public Policies” is also a useful theory of how such incrementalism works.

Comments are closed.

%d bloggers like this: