Previously in Part One, I outlined some of the basic analytical questions that may help in analyzing a budget, and went into detail specifically on those related to understanding the content of the budget itself – how much is being spent (the budget scope), how is it being divided up (the budget priorities) and how does it compare to other budgets (the budget context).
But a budget is much more than just a series of numbers and how they stack up next to one another on an Excel sheet. The process of gathering resources, identifying priorities, and applying them to achieve goals is at the core of every organization’s strategy, and it is an inherently political process. Understanding how that process work requires understanding a budget’s sources and how they impact decision-making; the budget actors who set the priorities by which the money will be distributed; and knowing who will be responsible for budget execution in the end. I’ll cover all of these aspects in this post, again using the specific case of the U.S. involvement in Afghanistan for my examples. The last post sought to answer the relatively easy question of how much does the U.S. spend in Afghanistan? This one will cover the much more complex question of how does it get spent?
Part Two: How Does the U.S. Spend Money in Afghanistan?
Budget Sources: The way an organization gains resources and revenue will have a major impact on its structure and priorities; the top priority of any organization is usually to gain more resources or at very least to preserve access to those it currently maintains. Generally speaking, an organization is going to be faced with a tradeoff between maintaining a diverse sourcing of revenue, which increases resiliency but also potentially increases aggregate constraints, and single sourcing of revenue, which reduces the breadth of constraints but makes those that are in effect more severe. To use a business example, Amazon.com has a variety of different revenue streams, from its huge variety of online sales offerings to its new Kindle content ecosystem to cloud server hosting, each of which on its own maybe brings only a moderate amount of money to the company. Amazon can afford to lose some money on one part of the organization or another (Kindles are being sold at a loss right now in order to bring people into the service ecosystem where Amazon will hope to make back the money), but its margins of profit are thin and it doesn’t necessarily have a lot of freedom to concentrate a lot of resources at the expense of others and expand rapidly into new areas. Google, on the other hand, derives a massive share of its revenue just from its lucrative search ad business, which has given it a lot of cash on hand to throw into experimental projects that don’t necessarily need to pay off in the immediate term if ever; if the online ad market ever tanks, however, it will face serious problems, and it always has to prioritize the health of that revenue stream. Many organizations are going to be a mix of the two, and this is not a perfect example, of course; Amazon started with a concentration in book-selling and Google’s now diversifying into other sectors like mobile phones.
When analyzing a government organization, the most common form of diverse resourcing will be some form of taxation – income and sales taxes generally being the big ones, the former more progressive and the latter more regressive in terms of which parts of society feel the greatest impacts. Single source revenue usually comes from rents collected by virtue of the state’s exclusive control over access to various valuable assets – things like licensing fees, trading access (customs duties), and (often the big one) state-owned natural resources. The other big source for some countries and many non-governmental organizations is going to be donor aid; this can also be viewed as a form of rent, particularly when that aid is given in exchange for some form of strategic cooperation or access (as would be the case for U.S. aid to both the Afghan and Pakistani governments). Some governments also successfully generate revenue through state-owned businesses or sovereign investments, although the actual productive track record on these has not always been great. Finally, organizations can take out loans (usually on fairly favorable terms if they represent a state), although these too eventually require some form of resource income in order to repay them.
Organizations that depend on public tax revenue are more resilient in that no one single contributor bears responsibility for funding any more than a small fraction of the government’s operations. If taxpayer Colin Cookman happens to be unhappy with the exact ways in which the government is slicing up its cut of his paycheck between its various agencies, it doesn’t matter much (particularly since he lives in D.C. and has no representation). As a whole, however, the government is going to have to be concerned with public reactions to their budgeting priorities, since it doesn’t don’t have other large assets that can make up for the loss if a large portion of the taxpayer base revolts and stop contributing (or demands a change in priorities through an election). That taxpayer-constituent / government relationship represents the basic core of the social contract in a democratic state. Rentier economies suffering from the “resource curse” by contrast don’t have to give a damn about what the vast majority of their populations think, since state seizure of the major rent-producing assets is often enough to fund government operations (and allow it to coopt or marginalize any rivals) without any need for public input. The risk of this (besides becoming totally disassociated from the concerns of a large portion of the public) will be that if the asset price of your commodity of choice drops (or your big single donor gets upset with how you have decided to spend the money), you have very little alternative budget sources available (this is why all the major oil producers rely on cartelization to keep their commodity prices high). To use another example, non-governmental organizations that rely on a large endowment for their operations usually can afford to have very specific and not necessarily widely-shared priorities compared to those that rely on small-dollar donations, and have to devote far less of their energy to active grass-roots mobilization and cultivation.
Afghanistan Case: The U.S. government does of course generally depend on public taxation as the main source of its operating funding. When taxes got high enough and decision-makers disconnected enough from the priorities of those contributing that money, we fought a revolutionary war over the issue of who would get to control how our resources were going to be spent. Our elected representatives in Congress now control the budget in matters of both war and peace, and their need to be able justify those military operations to their constituencies does serve as something of a check on what we are willing or not willing to fund in Afghanistan. (Pres. Obama and U.S. Congress members primarily justify the ongoing intervention to the public on domestic U.S. counterterrorism concerns, so in this regard it’s not surprising that Afghan state-building has not been a big priority.) Again that’s not to say there’s always a direct relation between popular opinion and the ability to fund war operations; Congress has become increasingly deferential to the executive branch on the details of military operations around the world in the past half century, which probably not coincidentally coincides (particularly over the past quarter-century) with an increased reliance on external debt financing that further removes the direct costs of spending from the current generation of voters and taxpayers. (The other factor moderating constituent concerns is an all-volunteer military force that only draws from a small percentage of the overall U.S. population). As our national conversation has refocused on debt issues and our economic liabilities in the past few years, you’ve seen much stronger criticism of the Afghan war’s relative costs to benefits for U.S. security.
A pretty good mirror case to the U.S. system is that of the Afghan government, which is extremely donor-dependent for both its basic operating costs and any capital investments in its security establishment or other public services. I would argue Afghanistan definitely qualifies as a rentier state, providing a single high-value good (counterterrorism cooperation) in exchange for large-scale international assistance. The Afghan government’s primary constituency is U.S. (and to a lesser degree other international) donors, not the Afghan public; elected members of parliament may get to review and vote on budgets, but it’s ultimately the Afghan Finance Ministry officials who work with donors in an attempt to coordinate money flows that actually set the country’s priorities (and even their ability to do so is limited by their dependency on what donors want to do with their money). This dependency on international aid by an Afghan central government is not unique to the U.S. intervention, as Barnett Rubin’s history of the Soviet period notes.
Protip: The Afghan government’s chances of collecting large-scale tax revenues from its public are limited, given its capacity limitations, security challenges, and the country’s general level of impoverishment. Its priority then is to to capture control over as much of external donor money as it can in order to fund its own priorities (first and foremost, consolidating its position against rivals). Doing so requires satisfying the concerns of U.S. officials, not necessarily those of the residents of Kandahar or Mazar-e-Sharif. If you’re wondering why the Afghan government is not particularly popular with many parts of the Afghan public and where all the distortion and waste in the U.S. intervention in Afghanistan comes from, this dynamic is at the core of your answer.
— — —
Budget Actors: The process by which actors within an organization decide how resources will be split up amongst the various priorities of a budget account is the core of the political process. This is sometimes described as “governance” or “strategy” but it’s all essentially the same thing — you just have to remember that it is a political process and that the interests of any specific actor may not necessarily be identical to what we assume the interests of the whole would or should be. The actors who control the process of prioritization (either by actual statutory authority or their ability to broker agreements between the competing interests) are in most cases going to be the most powerful parts within an organization, so identifying them and understanding the ways by which they assign priorities is going to be one of the most basic tasks required of an analyst attempting to understand how another organization works and what it wants.
The budgeting process can be separated into three distinct phases which correspond to three different types of actor — requests, wherein members of an organization attempt to advance competing priorities for the distribution of resources; authorizations, wherein approval is granted for a certain level of spending among the priorities, usually based on certain conditions; and appropriations, wherein the final approval is granted to release funds. (The actual disbursal, or spending of the money, will usually be done by the original requesting agency, and is dealt with in further detail in following section on Budget Execution.)
Requesting agencies are generally going to be pushing for the highest levels of resources they can get; authorizers will attempt to control the process by which it gets spent by adding conditions and other requirements; and appropriators potentially exercise the ultimate authority by agreeing to dole out the money (or not). Some organizations involve multiple layers of authorization before the final appropriation is approved; others regularly grant budget requests with minimal debate over their particulars. Some actors play different roles at different times. Organizations with greater internal checks and balances will separate some or all of these functions into different offices or agencies in order to avoid a concentration of decision-making power in any one actor. Without this separation of powers, organizations can potentially end up with seriously distorted or unsustainable budgets — it’s why we generally don’t let employees independently set their own salaries or implementing agencies choose their own funding levels without some higher approval and oversight process.
Afghanistan Case: Since we’re dealing with the U.S. system here and how it spends money in Afghanistan, that means understanding Congress (which has control over the budget source in our system of government, public tax revenue) and how it works. There are a variety of different backgrounders floating around on this process that you can Google but, in brief, it begins with the various agencies of the executive branch — the State Department, Defense Department, USAID, CIA, etc — formulating their requests, which the White House reviews and presents as an annual budget proposal package, usually in February of the year. (With the exception of a very, very few programs, like Social Security or Medicare, Congress is extremely loathe to enact multi-year budget appropriations, and thus everything is up for renewal every year.) The budget proposal then goes to Congress (usually starting in the House of Representatives) where the action begins.
The two houses of Congress (the House and Senate) have a long-established committee system which divides up legislative jurisdiction over various issues facing the government. If you think that this system of jurisdictions might tend to encourage narrow parochialism among members and uneven standards of accountability across the multiple agencies that collectively contribute to American foreign policy, you’d be right! But in any case, these committees — like the Senate Foreign Relations Committee or the House Armed Services Committee — have the power to authorize spending at particular levels (which may be higher or lower than those sought by the requesting agencies), create new program accounts or agencies to spend money in the future, and introduce new laws related to their jurisdiction.
These can come in a big legislative package (the FY 2012 Defense Authorization Bill, for example) or in bills dealing with a particular country, conflict, or program (the Kerry-Lugar-Berman bill authorizing expanded civilian assistance to Pakistan, for example). Authorization bills for the major categories of the U.S. government budget are supposed to pass every year and given the Congressional logjam, most smaller initiatives tend to get glommed onto them by amendment. It’s worth noting that the foreign aid authorization process in particular has been broken for the past several decades, effectively forcing State and USAID to rely on the same basic program accounts since the 1960s or otherwise on irregular, narrowly-focused legislation (like the aforementioned Kerry-Lugar-Berman) that amends this piecemeal. The act, as amended through 2002, can be found here.
The authorizing committees’ powers are separated from those on the Appropriations Committees and their various subcommittees (which deal separately with Defense, Foreign Operations, etc) who hold ultimate decision on the release of funds, which cannot be higher than the levels authorized by the authorizations process but can sometimes be considerably lower. Although they can’t technically enact anything that doesn’t have a specific monetary angle attached to it, Appropriations committee members are the ones who release the funds and not coincidentally tend to be the longest-serving (in a seniority system) and most powerful members of Congress.
Any legislation, including appropriations bills, that passes through committee has to go on to pass the full House or Senate (where it is potentially subject to further amendment) and then has to pass in a matching version in the other house of Congress. Any discrepancies in the two final bills have to be ironed out in a conference committee appointed by House and Senate leadership and then pass again in both branches of Congress. The THOMAS legislative search engine will be your friend when tracking all of these bills.
Protip: The “power of the purse” is Congress’ biggest institutional prerogative, but once budget funds are officially released to fund a particular program or agency, its ability to actually control how money is used can be limited, particularly in matters of foreign policy where the executive branch has wide leeway (and where Congressional ability to conduct oversight has been poor). “Sense of Congress” statements or Congressional “findings” that express a preference on how the money gets used are common, but generally speaking are not going to be binding. Congressional committees can threaten not to fund a budget account in the future their members don’t like how the administration is using its money, but this makes for a rather blunt policy instrument.
Three of the most common mechanisms used by Congress in both authorization and appropriations bills in an attempt to place closer controls on how money is spent are notification requirements that require agencies to let Congressional committees know when money has been spent or transferred; reporting requirements where the administration has to testify or prepare a lengthier report on how it is spending or planning to spend money; and the stricter certification requirements, where the president or one of his deputies must formally certify to Congress that certain standards are being met or otherwise trigger a fund cutoff. Beyond forcing the executive branch to share information with Congress, requirements can potentially create a public trail of documentation on a particular program account that Congress (and outside analysts) can use to make a closer assessment of the administration’s activities.
— — —
Budget Execution: After the budget actors have divided up a budget’s resources among its various priorities and approved the money’s release, it still has to be spent. Here’s where the balance of power has the potential to shift back to the requesting and implementing agencies. As noted above, appropriators usually have limited direct control over these issues, and depend on their control over future funds (as well as whatever other non-monetary authority they are able to exert by means of law or custom) to compel implementing agencies to abide by their budget priorities.
Knowing whether or not those priorities are actually being met requires regular oversight and feedback mechanisms, which some organizations structure into their basic operations and others do not. Even organization that do review their own spending practices often don’t make this information public, which can make conducting outside analysis and oversight extremely challenging.
Is the money actually being spent at the levels appropriated in the original budget document? Agencies that fail to spend all of their appropriated funds are inviting other competitors for those resources to make the case that they would be better-spent elsewhere, so the pressure to spend all the money received even if it goes to waste will be high. Will money actually be spent in the ways specified by the budget framers? Appropriators sometimes give organizations considerable freedom to transfer money around between budget priorities, but most require a minimum a notification if this takes place.
Afghanistan Case: The U.S. government has a comparatively robust oversight framework in place to monitor how its budgets are spent, starting with the Congressional committees, which can compel testimony and reporting from the various agencies that they fund. As part of the budget process, agencies like the State Department and the Department of Defense prepare budget justification documents which are supposed to describe the activities they are carrying out and what future funding will be spent on. (This year’s State Department justification can be found here, for example.) Reporting or certification requirements enacted in other Congressional legislation are also intended to
Both the House and Senate also have dedicated Oversight committees which are supposed to carry out investigations and keep an eye out for waste or misuse of funding. Each U.S. government agency has its own in-house Inspector General, which in the case of Afghanistan is further supplemented by the Special Inspector General for Afghan Reconstruction‘s office (although after a tumultuous leadership history, there have been suggestions that SIGAR may be either folded into the State and DOD Inspectors General offices or merged with the Special Inspector General for Iraq Reconstruction to create a general wartime oversight body). The Government Accountability Office‘s investigators also conduct oversight when tasked to by Congress.
Protip: The last question to ask yourself when considering how a budget is spent is the most important one: what results did the money actually achieve for the organization? This is the hardest form of oversight to conduct because it requires tracking not just how money came or went (the outputs) but how the programs that were funded did or did not accomplish the goals set out for them (the outcomes). Conducting effective oversight requires identifying a desired outcome and identifying what metrics you actually need to look for in order to judge the success or failure of that effort; in Afghanistan, we have not done very well at this so far. Tracking outcomes is not only complicated but also potentially challenging to the interests of those who provide the information for assessment and who implement the programs. Tracking budget outputs is always easier and safer, but don’t confuse it for the reality on the ground.
— — —
Thoughts? Critiques? Alternative frameworks for approaching budget analysis that I’ve failed to cover here? I’d appreciate any and all feedback in comments below.