Steps Toward an Enhanced Shared Services Strategy
The Biden Administration is building on the work of its predecessors in the area of shared services, which in turn was built on ideas from prior Administrations. In line with the Quality Service Management Office (QSMO) strategy announced in 2019, the current administration is turning the QSMOs introduced at that time from concept to operation, including the one that I was involved in pushing forward, the Grants QSMO at the Department of Health and Human Services (HHS).
Nevertheless, despite this helpful incremental work, most analysts would agree that the promise of shared services in the federal government remains unrealized and that the QSMO strategy, while important, cannot – by itself – achieve a shared services transformation. QSMOs are “marketplaces” supporting technology, and even when fully implemented, agencies will still perform most of these functions “in-house” and not gain efficiencies from scale and specialization. Agency adoption of a discrete set of common platforms enables potential consolidation, but it does not bring it into being.
In this post, I will reflect briefly on two lessons learned from the success (albeit limited) that HHS achieved in standing up the Grants QSMO. I’ll also suggest two additional critical reforms that will ultimately be necessary for the shared services model to become “the new normal” for federal agencies.
Lesson 1: Shared Services Should Start Within Agencies
Even if we are concerned about shared services across the entire federal government, shared services consolidation should ideally be at least well underway within the different Departments. Many subagencies – and this was certainly true at HHS and its 13 grantmaking subcomponents – have redundant support offices. and substantial efficiencies can be achieved through Department-level shared service models. HHS’ change management initiative, ReimagineHHS, identified grants as a priority area and implemented a program beginning in 2017 that built on several prior efforts, termed Reinvent Grants Management. In brief, this required business process analysis, stakeholder engagement, and not least leadership direction, for the various grantmaking entities within HHS as preparation for common platforms and systems. This in turn laid the groundwork for HHS to be predesignated as the Grants QSMO in 2019, and to achieve formal designation in 2021.
Understanding variation in processes, extracting common elements, identifying consolidation opportunities, and finding best-in-class feasible solutions takes time. But it also develops buy-in and enhances expertise, a feature applicable to services beyond technology. Although it might be attractive to imagine skipping over the Departmental stage of a shared services journey and going directly to a government-wide provider, in most cases that simply has not been and will not be feasible. A putative whole of government “shared services office” (SSO), whether limited (like the QSMOs) or more general, needs the credibility, substantive knowledge, and change management experience accrued by demonstrating results as a value-adding SSO at the Departmental level.
A good example of successful implementation along these lines is the Defense Financial and Accounting Service (DFAS), which performs payroll functions for several other agencies outside the Department of Defense (DOD). My own paychecks from HHS were issued from there. DFAS started as an internal DOD initiative in 1991, combining the separate organizations of the different military branches, and then leveraged its consolidated processes to later begin serving external customers, reducing headcount by over 50% and the number of systems used by over 75%.
Moreover, if and when a government-level SSO is set up, its interactions with government partners will be far more effective and efficient if those partners have already consolidated their functions, so the SSO can address itself to whole-of-agency solutions. There needs to be broad acceptance by leaders across government that the shared services journey will be incremental. This means (1) dedicated resources to enable (and require) all Departments to assess the level of fragmentation in their amenable support functions, and (2) for those agencies with high opportunities for consolidation, a program of carrots and sticks from the Office of Management and Budget (OMB) to push shared services models forward over a three to five-year window. These points lead directly to the second lesson.
Lesson 2: Interagency Shared Services Should Remain Within a Lead Program-Delivering Agency
This principle may be more counterintuitive than the previous one. Why not have an SSO be a specialized entity for delivering government-wide services? This looks well in theory, but in practice, it is usually better to maintain a close connection between a support function and the programmatic mission it supports. The point of an SSO, after all, is delivering higher quality at lower cost; learning about whether this is being achieved almost certainly occurs sooner if at least some mission clients are in the same agency as the provider, under the same executive authority. Relatedly, the host or lead agency provides a baseline level of work for the SSO, an opportunity for piloting new solutions, and a sophisticated organizational infrastructure (legal, analytical, acquisition, congressional liaison, and so forth) that a new specialized agency needs but rarely would have.
Perhaps even more importantly, using a Cabinet agency as the vehicle for an SSO helps overcome political and financial realities that have long stymied robust adoption of shared services in the federal space. The most trenchant argument for grants reforms at HHS was always a variant of the claim that the savings derived from standardization, automation, or consolidation of grants administration would in some form return to the grants programs, allowing them to expand the size or number of the grants they made. Although making this promise real requires working closely with budget experts, it can be feasible within an agency and allows the SSO to demonstrate the kind of concrete mission-enhancing results that make for good adoption incentives.
Moving to the interagency context, no Secretary likes to see, all else equal, budget move from his or her Department to another one, and similar attitudes prevail in congressional committees with authority over agencies giving up unshared activities. But the demonstrated prospect of returning money to programs can make this not “all else equal,” and mitigate these external concerns when the SSO seeks to expand its reach to the interagency.
Also, on the other side of the ledger, the Secretary of a lead agency hosting an SSO, and the associated committees overseeing that Department, all experience an increase in budget, authority, and importance. This creates a countervailing force against political resistance, which can include questions from agency leadership about the value in providing shared services that differ from the agency’s core mission (as has sometimes been the case with the Interior or Agriculture Department shared service provider offices). Overcoming such obstacles can generate what shared services models need most -- powerful advocates and allies with real incentives beyond merely good governance. However, for these incentives to become sufficiently strong in empowering a shared services transformation, consistent leadership is needed over multiple administrations who understand the value of their agency in delivering shared services, along with at least two additional critical reforms.
While this post argues for strengthening a lead agency model, central agency shared services models can emerge successfully in certain cases; the General Services Administration, for example, operates services for travel, real property, and government-wide acquisition contracts (GWACs). Similarly, broadly comparable governments in other nations have established effective shared services offices at the center of government, including Canada and Australia. But these efforts have either started at a central services agency without needing to consolidate existing effort, as with GWACs; or are done in parliamentary systems, which generally have fewer complexities in having a separate congressional funding and authorization process for each agency. Thus far, the logic is in favor of lead agencies as the better model for the delivery of services currently done by multiple agencies in the U.S. Federal government, relative to transitioning to only a central agency model.
Reform 1: Congress Should Develop More Flexible Funding for Administrative Operations
For stability and continuity, an SSO should have a baseline level of funding, but the bulk of its budget will track the extent to which its services are used by various government clients. The DFAS, for instance, operates as a working capital fund outside of regular appropriations. This is a useful and scalable mechanism for SSOs; however, the transferred money from various clients derives from appropriations and needs to be carefully structured by Congress to advance shared services. Even within an agency, it can be a challenge to translate operational savings into additional dollars for programs and missions.
Perhaps Congress’s first step, in line with starting shared services at the agency level, would be to enable the necessary reprogramming within an agency’s budget. This would include authorization for the use of unexpended administrative funds for additional grantmaking or mission activities (possibly in a “special projects fund” with goals specified by Congress). Ultimately, however, government-wide SSOs require overcoming interagency barriers, and must do so in a way that maintains the correct incentives – that is, one that allows the transferring agency to retain administrative savings, repurposing them for a desirable goal.
One possible approach might resemble the following:
- A budget line relating to a function amenable to shared services could be set up with no-year money (i.e., available until expended).
- OMB policy and appropriations law could confer transfer authority on all or a substantial portion of this money, so that (with notice) it is available for use in services or technology acquisition from a designated SSO or QSMO outside the agency for which it was budgeted.
- Such money would also be reprogrammable as discussed above, so that the unexpended funds remaining within the agency can be used to enhance the agency’s mission impact.
One potential use for these savings might be a set of agency-level technology funds (along the broad lines of the Technology Modernization Fund), in which worthy modernization and upgrading projects within each Department would recapture the benefit of shared services savings. This would offer an attractive prospect of aligning incentives as well as a virtuous cycle of governmental improvement in which shared service gains are reinvested in further efficiency. With care in project selection, system modernization will make agency operations even more amenable to future shared service consolidation.
Reform 2: The QSMOs Should Offer Services Beyond Technology Acquisition
The QSMO model is, as noted above, a step in the road to shared services. This model has the virtue of being relatively unthreatening to existing administrative staff, at least no more than technology adoption itself (which the QSMO seeks to accelerate and optimize). However, QSMOs have challenges in addressing many opportunities for savings that scale, consolidation, or sourcing to the most efficient provider could create. Nevertheless, this model is a suitable foundation for building the next stage of shared services. Additional QSMOs should continue to be designated over time, and in line with a new updated OMB strategic vision should expand the range of products and services in the marketplaces they offer.
It remains appropriate for a QSMO to begin with technology solutions. Technology will, for the foreseeable future, likely be the key driver for federal transformation in “back-office” administrative functions, including robotic process automation and implementable artificial intelligence. Getting the technology right, by assuring that it is standardized, secure, stable, and interoperable, is a critical step for government reform.
In conjunction with acquisition, however, the quality and efficiency of current operations and platforms should be transparently and regularly assessed by each agency – against common benchmarks to track performance in general and to identify suitability for shared service or technology adoption in particular. In other words, to have a sense of the value to be derived from solutions offered by a QSMO, the “buying" agency needs to know its baseline efficiency in supporting mission delivery to calculate the expected value of the new technology. This type of consulting and business process analysis could be offered to the agency by the QSMO itself, but a more effective approach may be for the QSMO to continue to serve as a marketplace -- in this instance, for vetted providers who can offer these analytic services according to guidelines and standardized methodology set by the QSMO.
Similarly, the performance of the various technology solutions should be regularly monitored and evaluated, to decide which solutions remain within the QSMO’s offerings and how best to advise agency clients. The actual post-adoption performance of technology is the key piece of information needed by potential later adopters, along with a comparable performance analysis of their legacy systems. This type of systematic and standardized evaluation could be done by the QSMO, but offering a competitive market of evaluators is likely to prove more useful. Introducing these analytic services before, during, and after adoption expands the concept of the QSMO substantially yet in a logical manner. The expanded marketplace for services could be supported further by OMB policy, requiring such assessments and evaluations while providing budget support.
The new data collected under the auspices of the QSMO would, notably, include information on opportunities within each agency for shared service provision – even if there is not currently such a provider. If identified opportunities cumulatively reach a certain dollar level, this may prompt serious discussions about shifting to a shared service provider, and government policy should intentionally seek to develop this evidence base and rationale. At some point, a “QSMO plus” of the type described here will have acquired knowledge of the relative alignment of different functions to shared services, as well as a sense of the highest-performing service providers. This, in turn, can lay the groundwork for one or more of these providers to begin to take over provision from less efficient agencies, a process facilitated by the relevant QSMO, which might continue to serve as a broker or intermediary. This evidence and market-based model, empowered by Congressional support and function-specific teams at OMB, could then be the next step in the evolution of shared services, with the potential to – finally – capture for the public sector the kind of gains shared services have shown in the private sector.
Image by rawpixel.com on Freepik