"Micromanagement" of the U.S. aid budget and the presidential allocation of attention.

How presidents allocate their time and attention is a venerable part of the folklore of the presidency. Does a president like to spend time talking with people, or is he a loner? Does the president read a great deal, or not? Does he make a habit of attending National Security Council meetings, or does he rely on subordinates to conduct business without him and then receive a report on the proceedings afterward? How much vacation time does he take at Camp David or his ranch or estate? Even his biological functions are subject for discussion, or at least speculation. Was this president one who needed a lot of sleep? Was the president one who was in a great deal, and so unable to perform much strenuous work? Among journalists, biographers, and the general public, this type of discussion of the presidency is commonplace, and always seems to attract a large and attentive audience.

Aside from the common fascination with the great and the powerful and the desire to understand their daily lives in human terms, the interest in how a president allocates his attention across a range of activities is part of how many people come to evaluate presidential performance. Comments about such allocations of time usually have a normative message--that the allocation is somehow not the right one for the president, at least if the president is intending to serve the public interest rather than this own comfort or enjoyment or his personal political fortunes. Implicitly, presidents are supposed to focus on the "right" issues and activities, which presumably means the big ones. On the other hand, if people hear that the president spends a lot of time on "little" issues that they believe to be the proper concern of a subordinate, then they might come to believe that the president is not performing his duties properly.

Given this level of popular interest in the subject, it is a bit surprising to find that there has apparently been no systematic academic study of how a president or presidents spend their time. Indeed, there is apparently nothing in the literature on high-level government officials of any nation or office that is comparable to Henry Mintzberg's (1973) classic study of how business managers allocate their time and effort. We have an abundance of anecdotes, but no systematic knowledge of the presidential labor process, or that of other high officials.

This is not because the topic is frivolous. Aside from our intuition that how a president spends his time is important, we have some normative literature within microeconomics (e.g., Radner and Rothschild 1975) on how decision makers ought to allocate their attention, as well as a normative literature within political science going back as far as Plato that addresses how leaders ought to conduct themselves. From either theoretical standpoint, a normative concern with the allocation of presidential attention and effort is easy to justify.

From the more modest standpoint of merely explaining presidential behavior, one notices that sometimes presidents seem not to give much attention to projects that their own speech and conduct suggest is something that they monitor quite closely, while at the same time they focus substantial attention on topics the importance of which is hardly self-evident. Two examples from the Kennedy administration illustrate each extreme of this phenomenon. Although the Kennedy administration attached a great deal of importance to securing congressional legislation to enable the president to take actions to expand international trade, Kennedy delegated responsibility for drafting the bill that became the Trade Expansion Act primarily to private individuals (McKeown 1994). On the other hand, Kennedy was prepared to meet with the director of the Development Loan Fund (which after the 1961 Foreign Assistance Act became a part of the Agency for International Development) to discuss at length what that director described as "minute questions of internal organization and procedure," including a discussion of how to get enough Kelly girls for secretarial back-up because the civil service was too slow in filling secretarial positions (Coffin 1964). The juxtaposition of these two incidents suggests an apparently irrational allocation of effort and hence an interesting puzzle to be researched.

Why then has the subject been left to journalists, biographers, and the occasional editorial writer? One reason is common both to the study of corporate CEOs and presidents: the higher the organizational status or rank of an organizational member or employee, the lower the probability that their labor process will be studied in detail. Scientific management as articulated by Frederick Taylor was a way for managers and engineers to gather information about the labor processes that they supervised--it was not intended and seldom has been used to study the labor process of supervisory personnel (Edwards 1979). That surely is an important reason why Mintzberg could find so few empirical studies of the managerial work process, and why he could find only one study of the president-as-manager--Neustadt's (1960), which was based largely on secondary sources. Managers have the formal authority to order studies of the behavior of subordinates, but subordinates lack the formal authority to order such a study of their supervisors (it would be interesting to know whether such studies have ever been included in a collective bargaining agreement).

A second reason for the lack of studies of the labor process of high-level officials is that, based on what Mintzberg tells us, their labor is inherently difficult to study. An assembly line worker typically has a fixed location where the job is performed, a job that is highly routine and repetitive, and few if any organizationally authorized occasions for privacy or confidentiality in the performance of the job. High-level managers, by contrast, move around a lot, they have relatively few routine tasks, their individual tasks do not take much time (hence are easily missed if surveillance is not complete), and they have many authorized occasions for confidentiality. Of course, presidents are much more like CEOs than like assembly line workers.

A third reason is one that by its very nature is difficult to document, so its importance is difficult to assess. High-level officials are often indifferent and sometimes hostile to the preservation of a written record of their activities. For example, one is struck in reading Foreign Relations of the United States how common it is for declassified documents to reference other documents that the archivists cannot find. There is no law that says that government officials have to preserve all of their written work, so it is not surprising that some of it is simply destroyed. The destruction is not necessarily motivated by the desire to conceal some aspect of decision making, but of course it has that effect. It is also common to find that meetings of single officials or small groups of officials with the president did not lead to the writing of a memorandum of conversation. Presidents can also use social occasions to create a pretext for physical proximity to someone whom it would be better to keep off the president's official schedule. For example, when Lyndon Johnson wanted to meet with George Woods, then president of the World Bank, to discuss U.S. aid and World Bank loans to India during the peak of U.S. concern about a drought and famine that struck that country in the 1960s, he arranged to have Woods invited to a large dinner at the White House. (1) Because Woods was never invited to such social occasions before or after the period when Johnson's interest in coordinating with the bank was this intense, it is reasonable to suspect (but very difficult to confirm) that the invitation to a social function was a way of getting Woods into the White House without triggering the curiosity of reporters or diplomats about the purpose of his visit. Finally, restrictions on declassification constitute a last line of defense. Because many of those who evaluate declassification requests are retired career officials who might have been personally involved in the issues that are the subject of the documents, it would not be surprising if they sometimes choose not to declassify material that does not portray their actions in a flattering light.

Fortunately, attempts to study the daily activities of the president of the United States have one signal advantage over corresponding attempts to study just about any other high-level official, public or private: the presidency is probably the best-documented job in the world. This is partly because of the constant stream of paper that enters and leaves the president's office, the frequent preservation of written appointment books and guest lists, the availability (at least in principle) of Secret Service logs that detail the movements of the president and whom he saw at what times, as well as the coverage of the president by news media assigned to the White House. These sources are potentially of enormous value in reconstructing the daily activities of a president. (Unfortunately, "potentially" is the appropriate term because these sources are drastically underexploited, and seem to have escaped systematic study entirely.)

These official sources are sometimes supplemented by record-keeping activities established by presidents themselves. Perhaps out of a sense of history, or for more prosaic political reasons (in Kennedy's case, the immediate motive seems to have been to keep an accurate record of conversations so that others would have a much more difficult time misrepresenting the substance of those conversations after the fact; Doyle 1999), several presidents beginning with Franklin Roosevelt have resorted to tape recording some of their activities. Such taping was selective rather than comprehensive, so it is helpful primarily as a way of filing in gaps in the documentary record. Lyndon Johnson went one step further, commissioning every cabinet department to prepare an official history of its activities. He also commissioned histories of important foreign-policy decisions, not all of which have been declassified.

In his discussion of different methods of researching the activities of managers, Mintzberg devoted no attention to archival research, perhaps because it is usually irrelevant to the study of contemporary corporate managers. Business records from the distant past are sometimes donated to archives or even opened by the companies themselves to research (see Mosley 2003 for an example of what can be learned by exploiting such sources). If one is interested in more recent times, Mintzberg's work suggests that obtaining the cooperation of management for such studies is quite unusual. There is no reason to expect that it would be any easier to obtain the cooperation of high-level government officials.

Toward a Theory of Presidential Attention

The beginnings of a theory of presidential attention are available in the extant management science literature, extended by some consideration of principal-agent problems as they apply to the presidency.

* Presidential attention is a scarce resource, the supply of which is ultimately constrained by the number of waking hours in a day. This is true whether we adopt a conventional microeconomic approach that recognizes no explicit limits on human information processing, or adopt a more psychologically realistic assumption of bounded rationality (Simon 1995).

* Subordinates have their own objectives, so, as noted long ago by Neustadt (1960), they may deviate from presidential directives when it is in their interest to do so. This requires presidents to devise monitoring systems and reward or punish subordinates depending on their faithfulness to presidential directives.

* The constitution and statute law constrain presidential decision making and in some cases compel it. In the case of aid budgeting, a clear example of this is to be found in the effect of the various restrictions that have been placed on the disbursement of aid that require a presidential waiver in order for the restriction to be set aside. While Hinckley (1993) views such restrictions with waivers as merely a theatrical exercise in position taking, devoid of practical significance, the reality is not quite that straightforward. The requirement that the president sign a presidential determination that finds that aid to a given country is in the national interest personally implicates the president in the provision of such aid, and provides "cover" for members of Congress to vote for aid to countries that are unpopular with their constituents. The process of generating the finding and the waiver entails at least a minimal amount of presidential attention and a presidential signature on a document that typically contains budget figures for a given country program in a given fiscal year. The available evidence suggests that presidents and their staffs do not see these restrictions with waiver clauses to be significant constraints on their substantive policy choices, but if it were costless to remove them, they probably would. (2)

* Presidents have the formal authority to revise organizational rules and control systems. They do not merely play a principal-agent game with subordinates--they have some capacity to choose which game they play. Every president creates the White House staffing system that he desires, and in principle can change that system at will. He also has the formal authority as commander in chief to reshape military decision-making procedures and outcomes, and through his cabinet officers who serve at his pleasure he has some capacity to do the same throughout these departments of the executive branch. (However, in these latter cases, the principal-agent control problems are probably greater than those he encounters inside the White House.) In the short run, presidential allocation of attention is a function of how that system works.

* For both subordinates and presidents, interaction takes place under conditions of incomplete information. One does not have to accept the accuracy of all of the assumptions that underlie game theoretic models of interaction under such conditions (e.g., Banks 1991) to agree that incomplete information is a precondition for and frequently an incitement to signaling behavior. However, identifying just what is and what is not a signal requires substantive knowledge of how those involved in government attach meanings to symbols and actions, and that knowledge is an input to game theoretic treatments rather than a conclusion or implication of the theories themselves. The involvement of leaders in ceremonial functions has long been understood as a form of signaling (Edelman 1967). A more ambiguous but for present purposes more relevant case of possible signaling is observed in the Carter White House. Reading the documents at the Carter Library that were known to have been seen by him, one notices that Jimmy Carter would sometimes mark minor grammatical or syntactical errors in papers reaching his desk that were to be returned to a subordinate. Was this simply a mannerism suggesting that Carter was mildly obsessive-compulsive, or was it a method of signaling to subordinates that their material was being closely read by the president and that his approval of a proposed course of action was considered and deliberate rather than routine? Game theory cannot answer this question, but it can make sense of this behavior if one has reason to believe that such actions were strategic rather than habitual.

Presidential Involvement in Country Aid Program Budgets: A Case of "Micromanagement"?

If the internal meanings of acts and symbols are not easily recoverable by researchers, then how are they to study the allocation of presidential attention and effort? I begin my research with an assumption--that presidents have good reasons to allocate their time in the way that they do. If the reason is not self-evident, then one searches for an explanation for the allocation of attention and effort in terms of the achievement of known presidential goals. One might ultimately conclude that to some degree changes in the allocation of attention, particularly changes from one administration to the next, are the product of idiosyncratic factors not easily treated within the confines of a general model of attention, but as a starting assumption it seems more fruitful to begin by postulating that presidents have good reasons for the allocations that they choose. Thus, our approach to the allocation of attention and effort is analogous to Kagan's (2001) analysis of the strengthening of presidential effort to exert control over administrative agency rule making beginning in the Reagan administration and continuing through the Clinton administration. Her argument is that the change can be understood not because the Clinton administration or its immediate predecessors had an unusually strong desire for control of these agencies, but because the political situation facing the administration left the White House with little choice:

   Faced for most of his time in office with a hostile Congress but
   eager to show progress on domestic issues, Clinton and his White
   House staff turned to the bureaucracy to achieve, to the extent it
   could, the full panoply of his domestic policy goals. (Kagan 2001,
   2248)

Her point generalizes to other situations where the president's party does not control Congress and wishes to change the policy status quo, so it fits the Reagan and Bush I administrations as well as it fits Clinton's.

The presidential allocation of effort discussed here pertains to a narrower issue than the wide territory discussed by Kagan. However, like greater presidential attention to administrative agency rule making, this shift in presidential attention is revealing not just for what it tells us about the motivations and actions of the president and his advisers, but also for what it reveals about the political environment within which presidents work.

Presidential involvement in setting the aid budget is a useful focus of inquiry for three reasons. First, the fact of a shift in attention and the nature of that shift are relatively well documented. Second, the shift is counterintuitive in the sense that it involved presidential attention to a level of administrative detail that under normal working conditions would be supervised by officials in executive branch agencies at the subcabinet level or below. It thus is an example of what journalists term "micromanagement"--a presidential concern with administrative questions that would ordinarily be seen as much too unimportant to merit presidential attention. Third, the case sheds light on an aspect of the larger political questions at the national or international level: microlevel changes in the White House allocation of attention and effort are related to macrolevel phenomena in national and international politics.

The President as Budget or Loan Officer: Budgeting for Foreign Aid Under Kennedy--Johnson and Nixon

From the end of World War II until about 1958, the United States was able to conduct very substantial foreign aid programs without concern about the consequences for the balance of payments, U.S. gold reserves, or the stability of the Bretton Woods exchange rate system. This is because for that thirteen-year period the United States maintained a surplus on the balance of payments. That benign policy environment ended in the late 1950s, as balance of payments deficits replaced surpluses. As the deterioration in the U.S. balance of payments continued, President Kennedy informed the treasury secretary and AID administrator on June 20, 1962 that he wished to be notified of "all actions which affect our dollar drain or balance of payments" before a final decision is made, and that "before any loan stabilization or AID agreement is made I would like to have an indication of what effect it would make on our dollar position" (Kennedy Administration 1962a). At a meeting three days later, the White House staff pressed the agency for much more detailed information about the costs of specific AID projects:

   [Mr. Kaysen of the White House staff] believed that a priority list
   with costs of each individual item should be prepared so that AID
   could present to the President a meaningful schedule of expenditures
   and what those expenditures would accomplish. At a previous meeting
   AID officers were not successful in convincing the President that
   AID was able to scrutinize the expenditures in relation to what they
   accomplished....

      Mr. Janow [Bureau for Far East, AID] ... stated that AID officers
   had strenuously objected to the use of AID funds for what were
   essentially military and political purposes. ... FE had contested
   DOD, CIA, and State every step of the way; but that he had been
   overruled in every instance. Kaysen replied that perhaps AID had not
   stood up to the pressures and that ... if there was a substantive
   issue involved here the President should have been informed about it
   prior to the time a decision was made by the Secretary of State or
   the Secretary of Defense. Mr. Janow replied that Mr. Forrestal of
   the White House had been informed ... and that members of the White
   House staff had informed him that political and military
   considerations were of overriding importance....

      Mr. Hutchinson commented that [in the conflict between military
   and political considerations and development] AID would never win.
   Mr. Kaysen replied that AID must nonetheless make the effort and
   apprise the President prior to the time of a hard decision being
   made at Cabinet level.

      Mr. Amory [Bureau of the Budget] [said] that AID officers were
   correct in contending that DOD and State must also be made aware of
   the President's concern over the balance of-payments problem.

      Mr. Kaysen [said] that at any time AID funds are expended on
   programs or projects in which AID does not believe, the President
   should be informed. He intimated that the President is not always
   aware that AID funds are expended to meet commitments made by the
   Department of Defense and State. Kaysen then urged AID officers to
   use the White House staff to warn the President before the
   Secretar[ies] of State and Defense make decisions involving AID
   funds.

      Mr. Gordon [AID] commented that there was disagreement in the
   White House staff on these matters and he believed that this
   presented a problem. Mr. Kaysen replied that these members of the
   White House staff generally in touch with specific areas could be
   expected to present the arguments and debate the issues before the
   President.... [T]he essential problem was that AID did not have the
   internal procedures necessary to make clear its position and to
   inform the President of the effects on the balance-of-payments
   position of the commitments made for political and military
   consideration.... Mr. Kaysen indicated that it was not intended that
   the President should act as a loan officer. However, he insisted
   that AID should be able to present a schedule that will tell the
   President what the effects of each loan, grant, and agreement would
   be on the balance of payments position. (Kennedy Administration
   1962b)

This exchange has a number of remarkable features.

* Nearly eighteen months after taking office, the administration had not established administrative procedures at Defense, State, and AID that would provide the president and cabinet officers with information about the balance of payments implications of specific AID projects. Given the depth of Kennedy's concern about balance of payments questions, this is mildly surprising. (3)

* The White House understood that AID, as part of the Department of State, was subordinate to State and Defense, and that therefore AID's interest in development would be subordinated to the military and short-run political interests of the senior departments. However, it counted on AID's complaints to the White House to trigger presidential attention to resolve conflicts among these three organizations. AID was expected to "stand up to the pressures" and let the president know what State and Defense were going to do to its programs. (In a classified speech in March at AID headquarters, the president had told AID employees that "I think it is your responsibility to the maximum extent possible to keep your eye on the long range economic developments, to struggle with the Department of State and ultimately surrender ungracefully"; Kennedy Administration 1962d.)

* AID apparently had already done what Kaysen asked them to do--take their case to the White House--and it had not resulted in the decision reaching the president's desk (nor, apparently, even Kaysen's desk).

In reading through the files on foreign aid at the Kennedy Library, I did not encounter any documentation on presidential decisions that resolved conflicts between AID and State or Defense, save the decisions that had to be made in the course of resolving high-profile cases such as the Volta dam and aluminum smelter in Ghana. If Kennedy adjudicated such disputes (or delegated that role to one of his staff) it is not evident in the papers. We thus do not know much about how this policy worked in practice.

Kennedy's successor tightened this policy even further. By April 1966, the Johnson administration was requiring that any request for foreign aid above $10 million on project assistance or above $5 million on all other commitments had to be approved by the president (Johnson Administration 1966). Unlike his predecessor, he left a substantial paper record indicating that he had done exactly this.

The procedure that the White House created for screening these proposals placed demands on the time of many high-level decision makers within the administration:

   AID sends a memorandum to the President through the Bureau of the
   Budget. BOB passes a copy of it to Treasury and obtains a statement
   from Secretary Fowler on his assessment of the commitment's impact
   on the US balance of payments. The Budget Director also prepares a
   statement indicating his views on the commitment in terms of US
   objectives. The entire package goes next to the White House. In the
   past, Mac Bundy added his own views before passing the proposal on
   to the President. I assume you [i.e., Rostow, Bundy's replacement]
   will take over this step.

      In some instances new commitments are presented to the President
   for his approval because of special foreign policy issues, e.g., aid
   to Indonesia, the UAR, India, Pakistan, without regard to the
   arbitrary dollar limits mentioned above. When foreign policy issues
   are the sole or prime reason for going to the President, the
   memorandum is signed by Secretary Rusk and is delivered directly to
   the White House. In such cases, Mac Bundy obtained the views of Joe
   Fowler, Charlie Schultze and others as appropriate and necessary.

      The above covers the formal procedure. In addition, each proposal
   is discussed rather thoroughly among the staff levels of State, AID,
   Treasury, BOB and the White House before being presented to the
   principals for signature. So far this has proved to be a useful but
   time-consuming method of recognizing and resolving issues and
   problems. (Johnson Administration 1966)

While the level of scrutiny that Johnson and the high-level officials in his administration gave to the foreign aid budget was remarkable, it was not entirely unprecedented. On at least one occasion, President Truman cleared a State Department request for aid expenditures in the Far East that involved approval of a program of economic assistance for Indo-China of only $750,000 (he simultaneously authorized a program for Indonesia of approximately $3 million, and a program of military assistance to Indo-China and Indonesia of $13 million; Truman Administration 1950). Even in 1950 these sums were a tiny portion of the aid budget, let alone the federal budget. There is, however, no indication that Truman required presidential approval of such small aid programs as a matter of routine. (Nor could I find a comparable record of presidential involvement in budgeting for foreign aid programs during the Eisenhower years.)

The congressional reporting requirements for aid expenditures were comparably stringent. In the early 1960s, only $50 million out of an aid program of about $4 billion could be spent on an unvouchered basis (i.e., no accounting had to be made to Congress of the specifics of those expenditures). Even then, any such expenditure had to be accompanied by a presidential finding authorizing such expenditures (Kennedy Administration 1961a).

What could motivate such a level of scrutiny of programs with such modest budgetary implications? Certainly, part of the answer is to be found in the balance of payments problems that the U.S. government experienced after 1958, which became more acute after spending for the Vietnam War increased sharply. The Eisenhower administration began reviewing the balance of payments impact of its aid programs in the first half of 1959, after a substantial acceleration of gold outflows (Eisenhower Administration 1959). During the Kennedy administration, an estimate of the balance of payments impact of the aid program found that $1.25 billion out of total aid expenditures of $3.4 billion in 1960 and $1.3 billion out of $4.1 billion expended in 1961 constituted a dollar outflow from the United States. (These statistics count expenditures for Food for Peace and Export-Import Bank loans as aid, and deduct from the dollar outflow all aid expenditures made within the United States; Kennedy Administration 1962e.)

Given the acute pressure on the U.S. balance of payments in this period, it is natural to ask why U.S. officials did not respond to this situation simply by cutting back on aid. The answer is that they perceived the costs of doing so to be very large. As Walt Rostow argued to John Kennedy at the outset of his administration, despite balance of payments pressures, the United States would still need to operate a very ambitious aid program, because "we must sharply reverse our fortunes in the under-developed areas" (Kennedy Administration 1961b). Rostow later in 1961 described the U.S. position in the less developed world as similar to the Allied position in 1942; a former staff member of Nelson Rockefeller's Office of the Coordinator of Inter-American Affairs compared economic penetration of Latin America by Germany with Soviet penetration two decades later, and presidential adviser Arthur Schlesinger told Kennedy in September 1961 that "The west is preparing itself for the greatest crisis since 1939" (Kennedy Administration 1961c, 1961d; Humphrey 1962).

Strong anti-Western sentiment in the less developed world in this era, highlighted by the triumph of Castro in Cuba, caused widespread concern in Washington. As RAND economist and administration adviser Charles Wolf noted, Castro's victory in Cuba showed that Communist-type revolution could be successful "where there is no feudalism, little conscious misery, no general popular despair, no mass desire for a 'socialist' transformation and no strong Communist party" (Wolf 1964). (4) The White House in this period perceived itself to be caught between Scylla and Charibdis in deciding on the size of aid expenditures: too large a program would accelerate the hemorrhaging of U.S. gold stocks, while too small a one would weaken a key barrier to further Soviet gains in the Third World. It responded to this situation by paying extraordinary attention to even small aid allocations.

The advent of the Nixon administration brought a different approach to White House monitoring of aid requests. With the promulgation on April 11, 1969 of National Security Decision Memorandum 10, Henry Kissinger stated that the Johnson administration requirements that the president must approve all PL 480 agreements, all AID project loans exceeding $10 million, and all AID program loans exceeding $5 million were now rescinded (Nixon Administration 1969d). (5) The new president now required a detailed country memo on all aid programs, but only for a select list of countries: Korea, Indonesia, Vietnam, Thailand, Laos, India, Pakistan, Turkey, Brazil, Chile, Colombia, and Nigeria. (In this era, these were the countries that were usually the recipients of the largest aid programs.) Moreover, the new procedure permitted some country programs to escape presidential review if prior presidential decisions had already dictated the terms of reference for that country's aid program:

   If the Secretary of State and the Assistant to the President for
   National Security Affairs, in consultation with the other
   addressees of this memorandum, determine that the President has
   already considered the major issues involved in a country program,
   they may decide not to submit that program to the President under
   this directive. (Programs already negotiated for this year need
   not be re-submitted.) (Nixon Administration 1969d)

Why would the new administration take a slightly more relaxed view of the task of monitoring aid expenditures? Part of the answer is to be found in the balance of payment statistics. The deterioration in the U.S. balance of payments situation stopped in 1968. As the U.S. economy moved into recession in 1969, the balance of payments improved, as is normal when the home economy moves into recession more quickly or more deeply than its trading partners (U.S. Department of Commerce 1981, Figure 31.1 and Table 1492). Henry Kissinger advised Nixon that as long as the U.S. balance of payments situation did not become dire, the United States should avoid heavy-handed controls, partly because that would lead to increased calls for bringing home U.S. troops and cutting the aid budget (Nixon Administration 1969e).

A second reason for a more relaxed view of the balance of payments implications of aid is to be found in changes in the way that the United States administered aid between the late 1950s and the early 1970s. While in fiscal year 1969 only 50 percent of the goods and services financed by U.S. aid programs were procured within the United States, by 1972 that proportion had increased to between 85 and 90 percent. If only commodities and not all goods and services are counted, the percentage was 96 percent. The administration was even willing to permit a very limited amount of "untying" of aid purchases in September 1970, but that resulted in offshore commodity procurement of only $37 million out of $829 million in commodity purchases under the aid program (Eisenhower Administration 1959; Nixon Administration 1973). Thus, the marginal gains in reduction of pressure on the fixed exchange rate for the U.S. dollar were very slight compared to the costs to the U.S. international position involved in a substantial reduction in aid.

The impetus for providing aid to counter the Soviet threat in the less developed world was also weakening. CIA assessments in the early 1970s of the Soviet position in the less developed world were uniform in seeing few opportunities for dramatic Soviet gains, at least in the short run (U.S. Central Intelligence Agency 1970, 1971a, 1971b). The sense of crisis that pervaded the discussions of the early 1960s had by this time completely evaporated. That is not to say that the Nixon administration was unconcerned about events in the less developed world. As the history of U.S. involvement in Chile would later show, the United States could be quite concerned and quite involved in selected countries. However, that does not contradict the proposition that Washington was generally inclined to worry less about Soviet penetration during this period.

The return of balance of payments pressure in 1971 did not lead to a return to the intense White House monitoring of aid requests. The Nixon administration developed a simpler and more comprehensive solution to those balance of payments constraints--closing the "gold window" in August 1971. The subsequent de facto floating of the dollar effectively decoupled aid spending decisions from concerns about the balance of payments. In addition, for a variety of reasons the U.S. aid program, which began to shrink in the Kennedy-Johnson years, continued to shrink under Nixon and Ford. The pursuit of detente with the Soviet Union also was an alternative to reliance on a large aid program to safeguard the U.S. position in the less developed world. Taken together, the twin constraints that had so bedeviled the Kennedy and Johnson administrations seemed altogether much more tractable by the time of Nixon's second inaugural.

Conclusions

With federal budgets in the range of $80 to $250 billion and aid budgets in the range of $3 to $4 billion, why would a president "micromanage" the aid budget, devoting attention to the expenditures of $5 to $10 million? Aside from the possible requirement to do so in order to comply with congressionally mandated aid restrictions and their accompanying presidential waiver clauses, the impetus for such attention to detail can be found in the high perceived marginal costs of over- or under-budgeting aid expenditures. Too high an expenditure would create a balance of payments problem; too low an expenditure would lead to a foreign-policy problem. The two-sided nature of the budgeting constraints and the size of the stakes were what stimulated the high level of presidential involvement. If the constraint were only one sided, then the necessity for significant presidential involvement would disappear: if it were simply a matter of satisfying a balance of payments constraint, and cutting aid programs created minimal marginal costs, the president could simply order cuts. Alternately, if there were no balance of payments constraint, and the White House needed to be concerned only with the marginal political costs in the less developed world of reductions in U.S. aid, the president would simply enlarge programs as appropriate. And if the marginal costs of either error were low, it is unlikely that presidents would focus on the decisions.

Of course, if one or both of these two constraints were relaxed, then other constraints--the unpopularity of aid programs among the general public and Congress, other claims on federal revenues, the level of aid expenditures by U.S. allies, or macro-economic pressures on macroeconomic policy--might prove to be binding, and the White House would need to be attentive to them as well. The difficulty of persuading a reluctant voting public and Congress to go along with aid stimulated a great deal of intramural discussion in the Kennedy-Johnson administrations, but both the public opinion and congressional constraints and the pressures on federal spending arising from the need to pay for the Vietnam War pulled the aid budget in the same downward direction as the balance of payments constraint.

A more systematic theoretical and empirical treatment of the allocation of presidential attention helps to clarify the reasons for patterns of attention that sometimes seem quirky or even irrational. Available material in presidential libraries, supplemented with secondary sources, could be used to create a far more systematic and complete map of how presidents allocate their time and effort. Because it is axiomatic that the modern presidency is constantly beset with more opportunities for decision making than it possibly has the time to accept, an adequate understanding of presidential attention is important not only for predicting patterns of presidential action and inaction, but also for sharpening our rather inexact normative conceptions of how presidents ought to divide their time.

(1.) The Johnson Library maintains that president's Daily Diary on a card file that lists the dates and occasions for each visitor to the White House. Woods also conducted a lively correspondence with Chester Bowles, U.S. ambassador to India, outside of official channels. The material is in the Chester Bowles papers at Yale University. See, for example, Bowles papers, Group 628, Series I, Box 338, folder Wolff-Wriggins.

(2.) At its inception, the Nixon administration considered whether to seek from Congress the repeal of thirteen of these restrictions. The consensus that emerged was that the restrictions imposed no significant hindrance to the conduct of foreign policy, but were "barnacles" that marginally slowed and complicated the budgeting process. Because of this, the benefits of removing the restrictions were judged to be small, while the costs if a member of Congress wanted to make an issue out of the removal and portrayed the administration as abandoning the defense of the national interest might be significant. On this basis, Nixon decided not to press Congress for their removal (Nixon Administration 1969a, 1969b, 1969c).

(3.) The administration believed that the balance of payments problem arose directly from the "disproportionate share of the defense and burdens for the whole free world" that the United States was carrying (Kennedy Administration 1962c). The section of the Foreign Relations of the United States volume dealing with Kennedy administration financial and monetary policy shows very substantial presidential involvement in and concern with balance of payments questions.

(4.) Wolf's work was known to several officials within the Kennedy-Johnson administrations, and he had considerable contact with administration aid officials. Conversation with Wolf, Chapel Hill, NC, January 1994.

(5.) Cognoscenti of national security memoranda should note that this time it was the president's national security adviser Henry Kissinger and not the president himself who was the signer. In the previous Kennedy-Johnson administrations, such memoranda were issued under the president's signature. Thus, Nixon took a small step away from personal involvement with these directives.

References

Banks, Jeffrey S. 1991. Signaling games in political science. New York: Harwood Academic. Coffin, Frank. 1964. Interviewed by Elizabeth Donahue, March 2-3. Kennedy Library Oral History Program.

Doyle, William. 1999. Inside the Oval Office: The White House tapes from FDR to Clinton. New York: Kodansha America.

Edelman, Murray J. 1967. The symbolic uses of politics. Urbana, IL: University of Illinois Press.

Edwards, Richard. 1979. Contested terrain: The transformation of the workplace in the twentieth century. New York: Basic Books.

Eisenhower Administration. 1959. Status of the Mutual Security Program as of 6/30/59, NSC 5912, Declassified Documents Reference System 1984: 948.

Hinckley, Barbara. 1993. Less than meets the eye Foreign policy making and the myth of the assertive Congress. Chicago: Twentieth Century Fund and the University of Chicago Press.

Humphrey, Hubert. 1962. Letter, William Barclay Harding to Hubert H. Humphrey, March 6, Alliance for Progress, Box 150.B. 15.10F, Legislative Files, Hubert Humphrey papers, Minnesota Historical Society.

Johnson Administration. 1966. Memo for W. W. Rostow from David Bell, April 27, AID, Vol. III [2], Box 2, Agency, National Security Files, Johnson Library.

Kagan, Elena. 2001. Presidential administration. Harvard Law Review 114(8): 2253-385.

Kennedy Administration. 1961a. Letter, Deputy Director of Bureau of the Budget to Allen Dulles, 5/30/61 (marked "Personal"), K 4-1 [2], Box 28, Series 61.1a, Record Group 51, National Archives.

--. 1961b. Walt Rostow, "The Strategy of Foreign Aid," Foreign aid, 12/60-2/61, Box 297, National Security Files, Kennedy Library.

--. 1961c. Memorandum to the President from Walt Rostow, June 17, re: The shape of the battle, Rostow, W 6/61-12/61, Box 65, President's Office Files, Kennedy Library.

--. 1961d. Memorandum for the President from Arthur Schlesinger, Jr., September 18, Arthur M. Schlesinger, 8-10/61, Box 327, National Security Files, Kennedy Library.

---. 1962a. Memorandum for the Secretary of the Treasury and Administrator of AID from the President, June 20, Foreign Relations of the United States 1961-1963, IX, Foreign Assistance Policy, document 144, note 3.

--. 1962b. Funari to Hamilton, June 23, Foreign Relations of the United States 1961-1963, IX, Foreign Assistance Policy, document 144.

--. 1962c. Kaysen to the President, September 18, 1962, Foreign Relations of the United States 1961-1963, IX, Financial and monetary policy, document 60.

--. 1962d. Unedited transcript of Fowler Hamilton introduction and Kennedy off-the-record speech to AID employees, March 12. National Security Files, Kennedy Library.

--. 1962e. Memo for Secretary of State and the Secretary of the Treasury from Frank Coffin, April 6, AID, Box 68, President's Office Files, Kennedy Library.

McKeown, Timothy J. 1994. What forces shape American trade policy? In The NAFTA puzzle: Political parties and trade in North America, edited by Charles E Doran and Gregory P. Marchildon. Boulder, CO: Westview Press.

Mintzberg, Henry. 1973. The nature of managerial work. New York: Harper & Row.

Mosley, Layna. 2003. Global capital and national governments. Cambridge, UK: Cambridge University Press.

Neustadt, Richard E. 1960. Presidential power: The politics of leadership. New York: Wiley.

Nixon Administration. 1969a. Bergsten to Kissinger, May 20, Foreign Relations of the United States 1969-1976, IV, Foreign Assistance, document 8.

--. 1969b. Burns to Nixon, May 22, Foreign Relations of the United States 1969-1976, IV, Foreign Assistance, document 9.

--. 1969c. Kissinger to Nixon, May 22, Foreign Relations of the United States 1969-1976, IV, Foreign Assistance, document 10.

--. 1969d. National Security Decision Memorandum 10, April 11, Foreign Relations of the United States 1969-1976, IV, Foreign Assistance, International Development, Trade Policies, 1969-1972, document 7.

--. 1969e. Kissinger to Nixon, June 25, Foreign Relations of the United States 1969-1976, III, International Monetary Policy, 1969-1972, document 131.

--. 1973. Letter, Roy Ash to Sen. Milton Young, July 20, K 4-2, Series 69.1 (1973-4), Box 13 and 14, Record Group 51, National Archives.

Radner, Roy, and Michael Rothschild. 1975. On the allocation of effort. Journal of Economic Theory 10(3): 358-76.

Simon, Herbert A. 1995. Rationality in political behavior. Political Psychology 16(1): 45-61.

Truman Administration. 1950. Letter, Truman to Acheson, May 1, Folder 5, Box 7, Lot 58D258, Record Group 59, National Archives.

U.S. Central Intelligence Agency. 1970. National Intelligence Estimate 11-6-70, Soviet Policies in the Middle East and Mediterranean Area (supersedes NIE 11/6-67), March 5, NIEs on the Soviet Union, Record Group 263, National Archives.

--. 1971a. National Intelligence Estimate 11-9-71, Soviet Policy in Asia (supersedes NIE 11/13-69), April 15, NIEs on the Soviet Union, Record Group 263, National Archives.

--. 1971b. National Intelligence Estimate 80/90-71, The Soviet Role in Latin America, 4/29/71, NIEs on the Soviet Union, Record Group 263, National Archives.

U.S. Department of Commerce, Bureau of the Census. 198l. Statistical Abstract of the United States 1981. Washington, DC: Government Printing Office.

Wolf, Charles. 1964. Political effects of economic programs, unpublished report, RAND Corporation, Santa Monica, CA, quoted in remarks of Floyd E Feeney, Special Assistant to the Administrator, A.I.D., for discussion at Advisory Committee on Economic Development meeting, February 4, 1966, Political Development, Box 23, David Bell papers, Kennedy Library.

Timothy J. McKeown is professor of political science, University of North Carolina, Chapel Hill. His work on presidential decision making during the Cuban Missile Crisis was published in the Journal of Politics in 2000 and 2001.

AUTHOR'S NOTE: This research was assisted by an award from the Social Science Research Council of an Advanced Fellowship in Foreign Policy Studies with the support of a grant from the Ford Foundation.