Section: ARTICLES

This article explores issues related to the size and complexity of the White House Office (WHO). Overall numbers from the Roosevelt through the Carter administrations showy uneven growth (and decline) in the White House staff and point to the need for more disaggregated analysis. Attention focuses on three offices (congressional relations, speechwriting, and public liaison) in the Nixon, Ford, and Carter administrations as illustrations of the varying patterns in and explanations of changes in staff size. One implication of this analysis is that efforts to restrict the size of the WHO may undermine presidential capacity and accountability.

Questioning the size of White House staffs dates back at least to Herbert Hoover's introduction of a noncareer, professional staff of four: "In bygone days, the President had one secretary...now there is a whole machine-gun squad to handle the work" (A Washington correspondent 1929, 385). His successors, from FDR through Johnson, confronted similar criticisms. Even so, complaints mounted in volume and intensity with the presidency of Richard Nixon (e.g., Bonafede 1973; Cronin 1980), often hypothesizing a direct link between staff growth and involvement in mischief. Although John Hart (1995a, 1995b) and others have effectively challenged many such claims, concern about size persists. Members of Congress and journalists monitor and criticize allegedly bloated White Houses, while presidential candidates routinely pledge, and presidents typically fail, to slash their staffs (see, e.g., Bedard 1998; Robertson 1997).

Academics also continue to weigh in with their concerns. Looking at the national executive branch, for example, Paul Light (1995) contends that the increased size of any agency often signals "thickening," which in turn exacerbates the diffusion of governmental accountability. Matthew Dickinson (1997) focuses more specifically on the impact of the growth and thickening of the White House staff on the presidency, arguing that larger staffs hinder the bargaining president by increasing management costs and tying the president too closely to the interests of others. Both scholars suggest remedies that are quite similar to Cronin's (1980): downsize and "thin" the White House and turn to "spot contracting" for those functions that cannot simply be shed.

Provocative as such analyses are, they are not fully convincing, at least for so unusual an organization as the White House Office (WHO). Frequently missing in much of the debate over the size of the White House staff are systematic empirical analyses both of exactly what is being counted and of how and why the numbers of aides have changed. Nor has the normative assertion that bigger and thicker are always worse than smaller and thinner often been subjected to much careful scrutiny.(n1)

In what follows, we seek to further explore issues related to the size of the White House staff.(n2) First, we establish the context for the discussion by examining overall staff numbers from the administrations of Franklin Roosevelt through Jimmy Carter (when the size of the WHO mostly stabilized and strictly comparable data ceased being collected) .(n3) This examination also underscores the difficulty of developing generic accounts of the growth of the staff: there has been neither constant expansion nor stable trends. After advancing some possible explanations for the uneven growth (and decline) of the WHO, we stress the need for more disaggregated analysis. Deserving particular attention when looking at changes in size are the component units of the WHO. Below, three offices in the Nixon, Ford, and Carter White Houses (congressional relations, speechwriting, and public liaison) serve as illustrations. We conclude by briefly revisiting the normative arguments over size, contending that efforts to restrict the size of the WHO on occasion may undermine presidential capacity and accountability.

Estimating the size of the White House staff long has bedeviled observers. Yet, the first systematic study of the subject seems not to have been done until Richard Nixon's first term.(n4) The numbers reported in the Nixon administration analysis were repeated in documents in the Ford and Carter White Houses, with appropriate updates but without the full set of comparisons. Table 1 and Figure 1 include the Nixon data as well as the numbers used in the Carter White House, which incorporated the Ford staff's calculations.(n5)

Table 1 and Figure 1 point to several notable variations. The number of full-time White House staffers declined between 1941 and 1942 and stayed at the lower level through 1946, presumably because the White House per se had relatively less to do with the onset of World War II.(n6) These numbers then recovered under Truman and Eisenhower and remained relatively stable until rising under Nixon.

Perhaps the most dramatic change in staff size occurred between 1970 and 1971, when the full-time White House and special projects staffs (columns 3 + 6 in Table 1) jumped from 345 to 555. In some accounts, this is evidence that Nixon radically increased the size of the White House staff.(n7) Yet, at the same time, the number of detailees plummeted from 287 to 17, resulting in an aggregate decline in staff numbers (although still a slight increase over 1969 figures). The explanation is that the Nixon administration decided, insofar as possible, to eliminate reliance on detailees, to submit an "honest budget," and to ask Congress to pay for the actual size of the staff.(n8) Congress responded positively, and the number of detailees remained low for the rest of the 1970s. The Ford administration vowed to limit detailees to no more than 25, and came close to succeeding. The Carter White House held the line as well until 1980; then, the approaching campaign combined with firm presidential insistence on capping authorized permanent staff at 351 generated significant backsliding. The number of short-term detailees toward the end of 1980 was vastly greater than in prior years, jumping to 205 in 1980 (Hart 1995b, 119; cf. Hugh Carter Files, Box 24, "Detailees," Carter Presidential Library). Many of these were short-term replacements for regular staff who took leave to campaign.

A third point requiring comment is the clear shrinkage of the "special projects" category after 1970 and its disappearance after 1973. This was a budgetary line that Congress had funded since the Eisenhower years; it was by no means a ruse to inflate the staff, since it was passed like any other appropriation. Nonetheless, the line item was virtually eliminated at Nixon's initiative as part of his honest budget. Congress did away with the residue of special projects in one of its early reactions to White House involvement in Watergate.(n9)

Overall, although the writings critical of the size of the White House staff frequently give the impression that its growth was steady and inexorable, inspection of Table 1 dearly shows that the growth instead was discontinuous. The total staff count (column 7), for example, jumped fairly sharply in 1947-48, again in 1956, once more in 1961, and yet again in 1969. In 1978, it dropped sharply. These discontinuities require explanation.

In a previous work (Walcott and Hult 1995), we proposed that structural changes in the WHO can be explained in terms of three kinds of independent variables: environmental factors (acting through the mechanism of a "rational," or strategic, president protecting advantages), presidential preferences (independent of particular environmental influences), and organizational influences such as inertia, the "differentiation dynamic," and "partisan learning."(n10) Using these variables to help account for changes in staff size may add insight into the dynamics of staff growth and decline.

Obviously relevant to any explanation are environmental factors. For example, as noted earlier, during World War II the number of permanent White House staff actually decreased; the modest overall growth was in detailees. What was growing during this period was the larger EOP, which contained the wartime agencies. The WHO caught the tail end of this expansion in 1947 when the remnants of the Office of War Mobilization and Reconversion were placed in it; even so, a decrease in detailees limited overall growth. The following year, however, as the White House geared up for the 1948 campaign, one sees an overall jump--much as increases appeared in 1956, 1976, and 1980. These were all, of course, years in which an incumbent president ran for reelection. Clearly, the rhythms of the political environment were eliciting the kind of strategic response from the incumbents that Kernell (1989) or Moe (1993) would have predicted. Nonetheless, one must note that the campaign effect for Eisenhower seems to have been modest, since seventy-eight new positions were added in 1956 by the first special projects appropriation. These slots were not campaign-related, although they might have freed up others for electoral pursuits. In the 1960s and later, by contrast, as the parties continued to weaken as key players in presidential selection, environmental pressures and incentives to place campaign direction in the White House mounted (cf. Tenpas and Dickinson 1997, 63ff).

The jumps in 1961 and 1969 evidently reflect new presidents--Kennedy and Nixon, respectively--gearing up the White House for their own purposes. Kennedy, for instance, added special projects units to achieve such goals as trade and agricultural policy reform. Nixon put in place a "staff secretariat" system along with an expanded emphasis on staffing for domestic policy under Daniel Patrick Moynihan and Arthur Burns. These increases appear to have been driven by presidential policy and administrative preferences. Even so, Nixon's adoption of the staff secretariat system (originally introduced by Eisenhower) likely also was an example of partisan learning and a consequence of the differentiation dynamic, triggered by the emergence of new staff units needing coordination.

At the same time, by this point in the evolution of the presidency, the task of staffing the executive branch had fallen squarely on the White House. Both 1961 and 1969 were years when the White House changed partisan hands, and thus some of the staff growth---especially in the case of JFK, whose staff went up eighty-eight in 1961 but then shrunk by forty-four the following year--reflected this change in the political environment. The main source of the change was the decline of the parties, which left the White House to absorb ever more of the work of filling executive branch vacancies (cf. Weko 1995).

Apart from 1961 to 1962, instances of shrinkage in White House staff size tended to be both modest and temporary, with two exceptions. One decrease, in 1973, was sizable (sixty-four slots) if temporary. It reflected Nixon's and Chief of Staff H. R. Haldeman's determination to cut the staff, along with the relaxing of campaign imperatives. Yet, it did not last, mainly because Watergate soon compelled the addition of legal talent to the extent that the WHO budget would permit. The other exception, in 1978, flowed from an even greater determination by Carter and his top aides to cut back, perhaps prompted in part by the congressional, press, and academic concern about staff size that then was in full flower. The sharpness of the Carter cuts, however, went well beyond any external demand, likely in part mirroring the "small staff" philosophy that Carter shared with most of his Democratic predecessors (Neustadt 1990, 219ff) and his Watergate-induced insistence on "cabinet government" (Moore 1981, 20). Yet, just as Nixon's "honest budget" produced increases that were more apparent than real, some of Jimmy Carter's "cuts"--for example, shifting many careerists to the new Office of Administration in the EOP and using numerous detailees on short details--led to decreases that were mostly illusory (cf. Dickinson 1997; Hart 1995a).

As the distinction in Table 1 between "permanent" White House staffers and more temporary detailees suggests, fuller understanding of the dynamics of staff size requires that attention be paid to more than total numbers. Other dimensions such as noncareer/career and nonprofessional/professional(n11) probably also need to be examined. Presumably, the staff growth that most critics fear, but relatively few stress, is that in the permanent, noncareer, professional ranks. Yet, in most recent White Houses, such aides--those at the level of staff assistant and above--probably number more or less one hundred.(n12) Here, of course, is where thickening may be of greatest concern, where presidents (or in actuality their chiefs of staff) confront both wide spans of control and deep hierarchies. Considerably less clear, however, is the extent to which even many of these aides have jobs with much discretion. More critically, there seems to be rather little basis for assuming that problems often linked to growth and thickening are apt to be constant over either time or space.

This last point in turn points to a final dimension along which the WHO might be disaggregated: organizational task or unit.(n13) Too frequently, those who explore White House staffing or who urge downsizing risk committing various versions of the ecological fallacy--for instance, assuming staff reductions were, or should be, uniform across units or that Government Manual listings of (some) senior staff capture the structural dynamics of the entire WHO. The WHO, however, is a complex organization composed of many units that handle numerous diverse tasks, respond to different parts of the environment, are located at various levels in the WHO hierarchy, and have variable significance depending on, inter alia, administration objectives, the point in the electoral cycle, and staff skill. One might well expect, then, that many WHO units would be rather loosely coupled to each other and to the larger organization. The size dynamics of any individual unit would not necessarily follow the dynamics of overall White House size.

To illustrate the need for more task- and unit-eye views of the WHO, we focus below on three tasks and three administrations: congressional relations, public liaison, and speechwriting(n14) in the Nixon, Ford, and Carter White Houses.(n15) Structuring for each task emerged and became institutionalized(n16) at different points over the past several decades, exhibited variable patterns of growth and decline, and requires somewhat different explanations of the changes in staff size. Two of the tasks--speechwriting and congressional relations--have been lodged in the White House for most of the "modern presidency." Aides for presidential writing can be traced back to FDR, although writing became fully institutionalized under Richard Nixon (Hult and Walcott 1998). Although it had been embraced covertly by Herbert Hoover, Roosevelt aggressively resisted placing congressional relations in the White House. Legislative liaison was partly assigned to lower ranking aides later in the Truman administration and institutionalized under Eisenhower (Davis 1983; Walcott and Hult 1995). Finally, although public (or group) liaison long has been a task performed by White House aides, it only was placed in a separate unit under Nixon and received an official title under Ford (Pika 1991).

The pattern of growth in congressional relations is striking. From early 1969 to mid-1978, the picture was one of relative stability; then it suddenly became one of explosive growth (see Figure 2). The explanation lies in the very different circumstances the three administrations faced and their responses.

Nixon began with a congressional relations office staffed and structured in a manner similar to that under previous presidents, going as far back as Eisenhower.(n17) The unit grew between mid-1970 and early 1971, in part due to dissatisfaction with its performance. Original head Bryce Harlow (who had a similar job under Eisenhower) was replaced by William Timmons; then former House member Clark MacGregor was brought in to oversee Timmons. During this period, the congressional relations staff grew markedly larger than any of its predecessors.(n18) This expansion coincided with many of Nixon's major legislative initiatives. The growth, driven mainly by Nixon's priorities, preceded that of other WHO units, notably public liaison.

By the time of the 1972 election, however, the congressional staff had shrunk to eight professionals, only slightly larger than it had been in mid-1970 (or under LBJ). Clearly, the emphasis on legislative accomplishment had lessened, while the demands of campaigning had enhanced competing staffing priorities (cf. Collier 1997, 136-37). The congressional relations staff contracted slightly more after the election, in accord with Nixon's and Haldeman's plans to downsize the WHO sharply.(n19) Those plans not only responded to external criticism of staff size but also reflected a sense among senior White House officials that the operation needed to shrink.

Gerald Ford's Office of Legislative Affairs inherited the Nixon operation and did very little to change it structurally. Early on, the unit had seven professionals and grew only to eight by the end of the administration. This is scarcely surprising in light of the difficulties Ford confronted with a Congress controlled by the opposition party and permeated by Watergate-induced suspicion of presidential initiative.(n20) Moreover, the president preferred to deal with members of Congress personally. Again, a combination of environmental factors and presidential preference worked to keep the staff from growing significantly.

The Carter congressional liaison office started with only five professionals, but it more than doubled within the first year. As in the Nixon case, widespread complaints about ineffectual relations with Congress seem to have prompted this response (cf. Davis 1979). Even the 1978 reductions in the White House staff failed to slow the growth of congressional liaison, as its importance grew in an administration dedicated to winning legislative victories in a Congress dominated by the Democrats. At the same time, those Democrats had become more independent, and the 1970s congressional "reforms" had produced a far more fragmented yet formidably staffed institution.(n21) Meanwhile, both Congress and the president faced novel policy demands (e.g., stagflation and energy shortages) and a restive public. The interplay of presidential strategy, congressional opportunity and constraint, and policy complexity ultimately led to an innovation that pushed the congressional relations numbers far higher: issue task forces dedicated to pursuing legislative ends. The peak number of thirty-five professional aides in late 1979, for instance, encompassed not only the normal liaison operation (which alone included more conventional House and Senate liaisons plus coordinating staff than either the Nixon or Ford units) but also personnel detailed to congressional relations from around the executive branch for two task forces: energy and the budget.(n22)

That the congressional liaison office did not collapse back to its earlier, smaller size even in the midst of the 1980 election campaign seemingly has a more idiosyncratic explanation. To an unprecedented degree, Carter staffers took official leaves during this period to campaign. Often they were replaced temporarily by individuals detailed from departments and agencies. This rotation of staff produced a large roster of people formally assigned to the various White House staff units, doubtlessly overstating their actual, working size at any particular time.

Overall, the patterns evident in these three administrations differ in explainable ways. The Nixon pattern is perhaps closest to what one might expect: growth during a period of legislative activity relatively early in the first term then decline as legislative defeats mounted and electoral issues rose to prominence. That there was no replenishment of congressional relations staff in 1973 evidently flowed, first, from Nixon's "administrative presidency" strategy going into the second term and, second, from his determination to cut the staff as a whole. The unit's subsequent lack of growth apparently reflects the decision to respond to Watergate primarily by adding lawyers, not congressional relations (or most other) staffers.

The Ford and Carter patterns contrast more sharply, indicating the radically different legislative opportunities each administration faced. Ford, seeing little possibility of passing legislation, kept the staff smaller than Nixon's peak numbers. Carter and senior aide Hamilton Jordan, frustrated but still possessed of large legislative ambitions, fashioned an approach that dramatically increased the size of the staff committed to congressional lobbying, much of it focused on individual priority issues. In all three White Houses, then, one sees the interplay of environmental conditions and strategic response, but the combinations differ rather strikingly.

Finally, it should be noted that the congressional relations operations under Nixon, Ford, and Carter only partially support our expectation that the size dynamics of individual White House units would not necessarily parallel those of the WHO. In the Nixon case, the general WHO pattern and that of the congressional relations office are quite similar. Ford's Office of Legislative Affairs grew slightly while the White House staff was shrinking somewhat, but the patterns are not widely divergent. Under Carter, however, even as the official size of the WHO decreased and far firmer controls were placed on the use of detailees, the size of the congressional liaison unit skyrocketed.

Formal provision for public (i.e., group) liaison in the White House had a modest beginning. Charles Colson's operation under Nixon started with a small staff in April 1971 but grew rapidly. Clearly focused on partisan, often campaign-oriented work, the office might have been expected to show sharp swings in size over the term as is characteristic, for instance, of the White House personnel office. To some extent, such a dynamic appeared in the three administrations considered here. But the patterns are not so simple. Under Nixon and Ford, the public liaison staff did not shrink the way an electoral cycle-based account (i.e., rising before elections, falling afterward) would suggest. Under Carter, as in the case of congressional liaison, the pattern is even more idiosyncratic. (See Figure 3.)

Colson's staff grew from an early complement of four professionals to as many as seventeen at around the time of the 1972 election. As is well known, the staff also diversified its activities beyond its original mandate of reaching out to potentially supportive groups such as labor unions and white ethnics. The activities of the Colson unit overlapped considerably with---and even dominated--those of the Office of Communications, especially the "attack" efforts directed by Jeb Magruder (Maltese 1995, 38, 89-94). These actions at the very least sometimes involved staffers in walking rather narrow legal and ethical lines, as they "rigged" opinion polls and created what now are called "astroturf" protest groups (cf. Maltese 1995, 72-73).

Given Nixon's interest in "hardball" public relations and campaign strategies, it is not surprising that the Colson operation flourished. What perhaps is more surprising is that it proved so robust. Rather than returning to its original size or even vanishing altogether after the 1972 elections, the liaison staff (headed by Colson's successor, William Baroody) remained substantial, even growing for a time.(n23) In part, this can be explained by the Watergate-induced need for the Nixon administration to mobilize its supporters. Yet, that does not fully account for why the unit, now formally rifled the Office of Public Liaison (OPL), grew even larger under Ford--weeks before the new president himself had committed to running in 1976.

Over the course of the Ford administration, OPL's size gradually declined--from seventeen in November 1974 to twelve in December 1976. Still operating under Baroody, the unit's campaign-related activities were greatly curtailed compared to Colson's, so even the president's decision to seek election in 1976 does not seem sufficient to explain OPL's continuing relatively high staffing levels. Indeed, some of Ford's top advisors saw OPL as an appealing target during their efforts in 1975 to cut the staff below five hundred.(n24) More likely, the ability of OPL to survive downsizing attempts was grounded in the expectations of its constituents. Interest groups, once granted formal access to the presidency through a White House unit of their own, were loath to relinquish the channel and lobbied hard to keep it. In short, while presidential strategy was the predominant factor in OPL's creation, the political environment largely sustained it.

Things might have been very different under Jimmy Carter, whose transition team, in a study of the Ford White House, proposed eliminating the public liaison staff altogether and transferring its functions to an "appropriate federal agency."(n25) But Carter ultimately did not go that far. The administration began with a public liaison staff, headed by Midge Costanza, whose size approximated Ford's OPL at its smallest. The president evidently intended his OPL to pay "particular attention...to enhancing communication with groups and interest[s] which traditionally have not had access to the White House, Cabinet, or sub-Cabinet levels of government."(n26) This is scarcely surprising, since one could not have expected a Democratic administration to ignore interest groups if its Republican predecessors had not, pressures reinforced by a rapidly fragmenting political system and the exploding number of groups. Although Carter hoped to increase outreach to the poor, elderly, disabled, and racial and ethnic minorities, Costanza broadened her mandate to include supporters of draft amnesty, abortion rights, and gay and lesbian rights. That in turn sparked considerable controversy in and out of the administration. After a period of growth, OPL began to lose staff, crashing to extinction when Costanza left the White House in mid-1978 after Carter's White House cutting had decimated her staff.(n27)

Yet, if Costanza had gone, the work of public liaison remained. Indeed, as Figure 3 shows, the number of White House aides devoted to this task rose sharply, creeping above the highest of the Nixon and Ford staffing levels and staying there. This likely happened as a result of the impending elections and the addition to the White House of experienced political operatives such as Jerry Rafshoon and Anne Wexler, who strongly advocated group liaison work to pursue both political and policy ends.

What did change were the orientation and organizational strategy for public liaison. Instead of a single office, a half-dozen units emerged aimed at diverse parts of the volatile Democratic coalition (blacks, women, Hispanics, white ethnics, Jews, and senior citizens); other aides, in Wexler's political affairs office, also concentrated on group liaison.(n28) Thus, the number of staff dedicated to group liaison burgeoned, abetted to a degree by organizational arrangements likely both to produce some redundancy and to amplify and broaden external demands for White House access. Although plans existed to eliminate at least four of the special assistants for "outreach" in a second Carter term,(n29) whether the administration would have been willing or able to withstand angry group objections can be debated.

Public liaison, something that presidents prior to Nixon approached with great caution, thus became a predictable part of the White House landscape between 1969 and 1980, although its organizational form did not stabilize. Born of presidential political strategy, it persisted because of group expectations and electoral needs. Under Nixon, it grew when the WHO grew (if far faster than the White House as a whole) and shrank, although not by very much, when the White House contracted. Similarly, under Ford, OPL's size dropped gradually as, more or less, did that of the entire WHO. During the Carter years, essentially the reverse occurred. At the time the WHO was being downsized, the public liaison staff, seemingly cut sharply, actually was just on the brink of growth.

In contrast to the sharp shifts in size in the congressional relations and public liaison staffs, changes in the size of the speechwriting staffs in the three administrations discussed here were more gradual and relatively moderate. (See Figure 4.) Once again, the Nixon and Ford patterns resemble one another more closely than either resembles Carter's. Here, however, it is the Nixon and Ford staffs whose numbers fluctuated, while Carter's speech staff remained virtually the same size throughout his administration.

The Nixon speech staff began with about a half-dozen writers (plus several researchers and other members of the Editorial Department whose numbers are not reported in Figure 4). This number stayed stable into mid-1971, then rose to twelve in early 1972, as the campaign season loomed. The electoral cycle is one probable explanation of the increase. A second is the heightening interest that the administration was taking in trying to guide the activities of several of its members (including the vice president, Cabinet members, and friendly members of Congress) by writing their major political and policy speeches for them. Although not always allowed to do this, the White House devoted resources to the task and thus developed a need for more writers.

After the 1972 election, the speech staff was downsized along with the rest of the WHO, although it never returned to pre-1972 levels. Indeed, the staff probably would have shrunk even more had the Watergate crisis not required Ray Price, who had planned to assume other duties, to devote most of his time to writing Nixon's Watergate-related speeches.

The Ford speech staff exhibited a similar pattern. With five writers at the end of 1974, the staff gradually rose to ten in the midst of the 1976 campaign. Increased demand accounted for some of this increase,(n30) but organizational changes also contributed. Steady complaints about the quality of the writing that emanated from the Editorial Office (which operated under the general guidance of Counsellor Robert Hartmann) led Chief of Staff Richard Cheney to encourage development of a small, alternative speechwriting group working under David Gergen, who became director of communications at the beginning of 1976 (see Hartmann 1980). At the end of the term, the election over, the writing staff dropped back nearer its original size.

Far fewer changes took place during the Carter years. The administration began with four writers plus a researcher. For most of the first two years, the staff numbered five writers but no researchers (in contrast, Nixon and Ford each routinely employed almost as many researchers as writers). The number of writers ultimately climbed to six, but there was never more than a single researcher, and commonly there were none. This evidently was a presidential decision, in spite of apparent environmental demand for more, not less, presidential elocution. The small size of the staff, and especially the lack of researchers, elicited persistent requests from the writers for more assistance,(n31) but help never arrived.

The long-term trend toward a more "public" presidency helps account for the emergence of specialized writing staffs in the White House (see Hult and Walcott 1998), and this environmental factor partially explains the size of those staffs under Nixon and Ford. In addition, both administrations clearly were influenced by the election cycle. Carter, yet again, defied the trends of his immediate predecessors, deliberately resisting any push toward a larger staff while also mainly eschewing his predecessors' ambitions to control their administrations by writing speeches for cabinet members and others. In Carter's case, presidential preference, a clear sensitivity to the prevailing environmental critique of staff size, and, perhaps, partisan learning effectively counteracted environmental shifts toward a more public presidency.

If the sizes of the Nixon and Ford writing staffs responded to the electoral cycle, growing as elections approached and contracting after, they also were roughly consistent with the overall staffing patterns of the two administrations. Once again, Carter's was not. His 1978 effort to shrink the WHO did not touch the speech staff, which could hardly have been reduced further than it already was.

Our descriptions and tentative explanations of the changes in the number of White House aides devoted to three tasks under three presidents clearly suggest three conclusions. First, our initial proposition that, at least in terms of size, the units for these tasks are only loosely coupled to each other and to the WHO as a whole is partially supported. Although congressional relations, public liaison, and writing followed roughly the overall White House trends in staff size in the Nixon and Ford administrations, they did not track them precisely, and the explanations for each presidency differ somewhat. Indeed, under Nixon, even though the staff at the end was smaller than it was at the beginning, all three of the components we examined were larger. In addition, as a visual inspection of Figures 2, 3, and 4 underscores, patterns differed discernibly across tasks, even within administrations. Even more strikingly, the patterns in the Carter years neither resembled those of comparable units under Nixon and Ford nor at all mirrored the overall Carter staff numbers (whether or not one puts much credence in the claim that the staff was cut sharply in 1978). Overall, then, the expectation that the WHO must be decomposed to be more fully understood appears borne out.

Second, in terms of explanation, we have argued that environmental factors, presidential preferences, and organizational variables like partisan learning influence decisions about staff size, often in interaction with each other. The staff components did not grow (or shrink) because of inexorable external factors. But the blame (or credit) does not wholly rest with presidents, either.

Third, and related, some effects on staff size can be traced to political party. Partisan learning appeared in the Nixon use of the staff secretariat system and in Carter's emphasis on keeping the WHO relatively small but relying on detailees as the reelection campaign heated up. In addition, party mattered simply in the sense that whether the president's party controlled Congress affected the ways that all three presidents--but especially Ford and Carter--approached congressional relations. Party was relevant as well to the public liaison efforts of each of the administrations, ranging from Colson's "northern strategy" to the Carter White House's attempts to reach out to the diverse parts of the fraying Democratic coalition.

Another, perhaps more important, question has not yet been addressed: so what? That is, what difference does the size of the unit or the collection of staffers devoted to any particular task make? Our answers at this point must be tentative, given the limitations of the data and the analysis. Still, the findings are, we think, suggestive.

Of course, at the outset the obvious must be stipulated: the specific impact of size (larger, smaller, change per se) on any unit or staff group will depend a great deal on the aides' tasks and on the circumstances surrounding them at any given time. However one accounts for changes in size, it seems clear that shrinking, say, a policy staff would have different consequences than reducing, for example, the number of writers or congressional operatives. Yet, our data may suggest more. In particular, we want to address the normative issue that weaves through so many critiques of staff size. Is large size necessarily bad for presidents or the presidency? Is downsizing always good?

We can approach such questions by looking at two of the units examined here, speechwriting and public liaison. In the former, the example of the Carter staff indicates strongly that downsizing may well have had negative consequences for the attainment of administration purposes. Elsewhere (Walcott and Hult 1998, 14-17), we have argued that Carter's reliance on a writing staff with considerably fewer personnel (researchers as well as writers) than its predecessors probably helped weaken his presidency. His approach both reduced staff capacity to respond to external demands for presidential speech and diminished presidential capacity by removing speechwriting as a vehicle for integrating policy substance, political tactics, and public relations.

Certainly, however, staff growth is not necessarily the path to an effective presidency. As the activities of the Colson staff under Nixon so dearly demonstrated, sometimes adding staff can, as Cronin (1980) argued, build in a capacity for mischief and misdeed that overwhelms any good that may have been accomplished. Not all presidential purposes are laudable, and staffers, once set on the wrong track, can move beyond presidential control. Again, Light's (1995) worry about the loss of accountability surfaces.

Even so, some staff growth appears to have been salutary. For instance, Carter aides, looking back, tend to credit the task force system, which so radically expanded the congressional liaison office, as one of their prouder achievements.(n32) Likewise, sometimes staff shrinkage can be beneficial to presidents. The squeezing down, then closing, of the Costanza OPL under Carter probably improved the administration's overall image if not its relationships with particular constituency groups.

Although each of the units explored here evolved along different paths and varied somewhat across administrations, considered together they point to a final irony. One can argue both that the growth of White House commitment to public liaison under Nixon was at least partially a problem and that the staffing constraints placed on speechwriting and, on occasion, congressional relations were not always desirable. Yet, those constraints were in large part a consequence of the success of public liaison in cultivating constituencies who quickly embraced the idea of having direct access to presidents through staffers dedicated to serving their needs. In this instance, the growth that Cronin and his contemporaries decried was indeed a problem--partly because it precluded potentially more useful staff expansion elsewhere.

The complexity of this normative analysis reinforces a larger implication of our findings: to be fully understood, the White House organization must be studied not just as a whole but also in terms of its component parts. The results concerning variation in size across units and between the WHO and individual offices make that point clearly. Perhaps even more convincing are the explanations of those findings. Even though the same clusters of variables (environmental, presidential preference, organizational) account for size dynamics at both the White House and unit levels, the individual variables and their interplay typically differ depending on the level and the task being examined. Our point is not that there are no effects of overall WHO size on unit size or the reverse--there are. In the main, however, the explanatory narratives we have constructed for the size dynamics of each unit are specific to the task and circumstances of that unit.(n33) This sort of organizational complexity distinguishes the "modern" institutionalizing White House from its predecessors. It perhaps also poses the greatest challenge to both empirical research and evaluative commentary on the organization of the presidency.

(n1.) Key exceptions include Hart (1995a, 1995b) and Neustadt (1990).

(n2.) Our emphasis here is on size alone. Clearly, this simplifies the exploratory analysis that follows. At least as important, though, it reflects our concern that the concept of "thickening"--at least to the extent it refers to the "height and width of the hierarchy" (Light 1995, 6)--is problematic when applied to the White House Office (WHO). Not all of the formal structures are hierarchical, and many of the more informal ones are more collegial in nature. In addition, even more than most organizations, the WHO is not what organization theorists refer to as a "closed system": relations and structures frequently extend and intrude from beyond its boundaries, further complicating valid measurements of "height" and "width." For a similar view of the White House, see McDonald (1981, 105).

(n3.) The method of officially counting staff mandated by the White House Personnel Authorization-Employment Act, signed by President Carter in 1978, produces results not precisely comparable to the most useful data on earlier administrations, which we rely on here. (See Hart 1995b, 118-19.)

(n4.) With its customary meticulousness, the Nixon Staff Secretariat sought to produce as accurate and authoritative an account as possible of the growth of the White House staff since the days of FDR. The first evidence of such counting appeared at the end of Nixon's first year; the overall size of the staff was traced from 1934 through 1969 [Memo, no author, December 31,1969, White House Special Files, Staff Member and Office Files, Office of the Staff Secretary, "Personnel (April 1970)," Nixon Presidential Materials Project]. Subsequently, the desire for additional analysis and comparison evidently led to a more comprehensive study. Noble Melencamp, then a staff assistant in the staff secretary's office, transmitted his analysis to Staff Secretary Jon Huntsman in July 1971 (Melencamp to Huntsman, July 9,1971, White House Special Files, Staff Member and Office Files, Office of the Staff Secretary, Box 144, "Size: White House Staff," Nixon Presidential Materials Project). Melencamp's accounting made no distinction between career and noncareer staff. It did, however, distinguish among full-time employees, detailees from other agencies, and WAEs ("When Actually Employed"; i.e., part-time staff), although the latter were not included in the overall totals. For that reason, the numbers were somewhat understated. Melencamp's document also counted "special projects" personnel separately, although it did add them to the overall totals. The special projects category was created during the Eisenhower years to fund such "special" activities as Harold Stassen's arms-control project and the public works planning unit headed by General John Stewart Bragdon (Walcott and Hult 1995, chap. 9).

(n5.) These data contain only one discrepancy: the Ford/Carter numbers for 1970 differ from Melencamp's estimate. We tend to trust the later, higher figure.

(n6.) Even more striking, White House employment as a percentage of total federal employment fell from .0185 percent in 1938 to .0169 percent in 1940 after the creation of the Executive Office of the President. This percentage continued to fall through 1945:1941, .0124 percent; 1942, .0080 percent; 1943, .0058 percent; 1944, .0057 percent; 1945, .0056 percent. According to the Melencamp numbers, White House employment did not exceed .02 percent of total federal employment through 1970, when it was .0183 percent. The highest percentage was posted in 1961 under President Kennedy (.0195). See Attachment, Melencamp to Huntsman, July 9, 1971.

(n7.) Such a conclusion, for instance, can be drawn from the data presented in Ragsdale (1998, 266; cf. Ragsdale and Theis 1997, 1296). For additional figures, see Hart (1995b, 116).

(n8.) The specifics of and rationale for this budgetary strategy can be found in a memo (no author or date) outlining responses to potential questions from Congress (White House Special Files, Staff Member and Office Files, Box 167, "Appropriations--Fiscal Year 1972," Nixon Presidential Materials Project).

(n9.) Congress eliminated special projects funding as of October 1973. This caused a serious budget squeeze in the White House, which was under budgetary duress already. Some positions (but no units) were eliminated as a result. See Bruce Kehrli to Alexander Haig, November 8,1973, White House Special Files, Staff Member and Office Files, Box 189, "Miscellaneous Personnel [#1]," Nixon Presidential Materials Project.

(n10.) The differentiation dynamic resembles the standard concern about bureaucratization leading to further bureaucratization (i.e., creation of new WHO units requires more new units to coordinate all of the units), contributing to further thickening. Partisan learning refers to the tendency of presidents to repeat the patterns of prior presidents of their own party.

(n11.) Identifying "professional" staffers raises a host of issues. Typically, for example, aides who primarily perform clerical tasks are excluded from these ranks. Yet, the rifles of such staffers have varied. Moreover, there appears to have been a tendency in many administrations to routinely place women in secretarial or administrative assistant positions even though some had mostly nonclerical responsibilities (see, e.g., Moore 1981, 84).

(n12.) In late 1972, one estimate put the number of noncareer professional staff in the White House at 145 but envisioned cuts to around 78 (Kehrli to Haldeman, December 12,1972, White House Special Files, Staff Member and Office Files, Office of the Staff Secretary, Box 157, "WH Staff, [Memos: Written and Typed]," Nixon Presidential Materials Project). The actual number cut was 55, bringing the total of noncareer professional staff to 90 (Draft Press Release, January 5, 1973, "Ziegler Staff [#2], White House Staff Files, Staff Member and Office Files, Office of the Staff Secretary, Nixon Presidential Materials Project). Although comparable numbers are not available for the Ford and Carter White Houses, little evidence exists to suggest much variation, especially given the efforts of both to cut staff.

(n13.) Although task and unit often are coterminous, that is not always the case. After August 1978, for example, the Office of Public Liaison (OPL) officially disappeared. Yet, as we discuss below, numerous senior aides, most in distinct offices, handled liaison with key constituency groups (e.g, Office of the Assistant to the President for Women's Affairs, Special Assistant to the President for Ethnic Affairs).

(n14.) Data on staff size for these three areas appear in Figures 2, 3, and 4, respectively. The numbers were derived from White House staff lists, memos, and other materials archived at the Nixon Presidential Materials Project and the Ford and Carter Presidential Libraries. In some instances, the numbers were compiled from staff lists assembled by the authors from those materials. Further details are available from the authors.

(n15.) We emphasize the Nixon, Ford, and Carter administrations for several reasons. First, as mentioned earlier, numbers of White House staff are not strictly comparable after the Carter years. Second, units for all three of the tasks appeared in these three White Houses and have remained there since then. Moreover, writing and public liaison did not for the most part become lodged in separate units until Nixon. Third, key features of the legislative and political environments had begun to change dramatically as one moved into the Nixon administration. Still, where relevant (and to reduce selection bias), we consider other presidents' experiences with each task going back to Herbert Hoover.

(n16.) Here, we emphasize the "autonomy" dimension of institutionalization (Ragsdale and Theis 1997). Institutionalization increases as a task becomes the primary responsibility of a specific White House office and as that unit becomes more differentiated from its environment (including, most immediately, the rest of the WHO). Also see Burke (1992).

(n17.) See Walcott and Hult (1995, chap. 2). One major difference between Johnson's and Nixon's congressional relations offices was that LBJ's OCR included a small staff, headed by Robert Hardesty, that wrote speeches for friendly legislators. Nixon's aides did likewise, but this was handled by the speechwriting office.

(n18.) Numerical comparisons to previous administrations can be found in John Brown to Haldeman, February 18, 1970, White House Staff Files, Staff Member and Office Files, Office of the Staff Secretary, Box 127, "Personnel (February 1970)," Nixon Presidential Materials Project. Even in early 1970, the Nixon staff was judged to be slightly larger than LBJ's or JFK's, and it was to grow substantially in the next few months.

(n19.) See Press Release, January 5, 1973, White House Subject Files, Staff Member and Office Files, Office of the Staff Secretary, Box 192, "Ziegler Staff [#2]," Nixon Presidential Materials Project. Earlier memos document the key role of Haldeman in planning these reductions. See, for instance, B. Kehrli to Haldeman, December 12,1972, White House Subject Files, Staff Member and Office Files, Box 127, "[Memos, Written and Typed]," Nixon Presidential Materials Project.

(n20.) Administration officials were keenly aware of these environmental constraints. See, for example, Bob Bonitai [Assistant to the OMB director for congressional relations] to Max Friedersdoff [head of the White House congressional relations office], December 17,1974, "Some Thoughts on the Administration's Relations with Congress," John O. Marsh, General Subject File, Box 10, "Congressional Relations Office (1)," Ford Presidential Library.

(n21.) Again, administration officials, at least in retrospect, were cognizant of such constraints. See, for example, Cable (1981, 59); Wexler (February 12, 1981, 28). Likewise, they were greatly impressed by the relative resource deficit they ran vis-a-vis Congress. Staff Director Al McDonald recalled ruefully, for instance, that the Senate Judiciary Committee alone had 125 staffers (March 13, 1981, 56).

(n22.) McDonald and H. Carter to Moore and Thomson, November 1,1979, Hugh Carter Files, Box 16, "Congressional Liaison--1979 [1]," Carter Presidential Library. Even these numbers do not include all of the WHO staffers who worked to push bills through Congress. See, for example, Collier (1997) and Wexler (1981).

(n23.) Also in this period, Anne Armstrong came to the White House as counsellor to the president, with a portfolio that included liaison with women's groups. At least two members of her staff were primarily engaged in this and are counted here with the Baroody staff.

(n24.) See, for instance, Connor to Richard Cheney, September 18, 1995, James E. Connor Files, Box 21, "White House Staff Reductions (5)," Ford Presidential Library. "Baroody," wrote Connor, "is always a target for cutback."

(n25.) Harrison Wellford et al., White House Study Project, Report No. 1--December 7,1976, "Analysis of Present White House Office," Ford Presidential Library.

(n26.) Jo Ann to Sandy Adams and Seymour Wishman [all OPL staffers], November 22, 1977, "Draft, [OPL] Statement of Purpose and Philosophy," Carter Presidential Papers: Staff Offices, OPL: Costanza, Box 30, "Office of Public Liaison: Daily Routine, 11/77-12/77" Carter Presidential Library.

(n27.) The number of public liaison professionals never dropped below five because other offices, notably one serving minorities, were authorized at virtually the same time.

(n28.) Staff Director Al McDonald traced this strategy to Wexler's conception of public liaison responding to a fragmented public (McDonald 1981). Compare Pika (1991, 291). In addition, Wexler and the rest of her staff worked with interest groups to mobilize support for specific presidential legislative initiatives. See, for example, Collier (1997) and Wexler (1981).

(n29.) Document, no author, no date, Hugh Carter Files, Box 78, "Second Term Planning," Carter Presidential Library.

(n30.) For instance, Paul Theis (head of the Editorial Staff) complained that the writing staff was "overbooked" [Theis to Warren Rustand, August 6, 1975, Robert Orben--Paul Theis Files, Box 82, "Aug 1975 (1) (PT)" Ford Presidential Library].

(n31.) See, for example, Caryl Conner [writer] to James Fallows [head writer], August 22,1978, Staff Offices, Speechwriters' Subject File, Box 29, "Speechwriting Organization, 8/78JF"; Hendrik Hertzberg to Al McDonald, November 8, 1979, Staff Offices, Speechwriters--Administrative File, Box 1, "Administrative File--1979, September 1, 1979 to December 31,1979)"; Carter Presidential Library. These and other similar pleas made explicit comparisons to the Nixon and Ford staffs.

(n32.) See, for instance, Interview with Hamilton Jordan (including Landon Butler), Miller Center Interviews, Carter Presidency Project, vol. VI, November 6, 1981, Carter Presidential Library, pp. 53-62.

(n33.) This does not suggest that each individual White House unit is an independent "case." These are not cases but elements of a complex organization that interact with each other within the context of the WHO and the larger political system in which they are embedded.

Legend for Chart:

A - 1 Year
B - 2 White House Detailees
C - 3 White House Full Time
D - 4 Special Projects Detailees
E - 5 Special Projects Full Time
F - 6 Total Special Projects
G - 7 Total White House

 A            B      C        D     E      F      G

1934[a]      120     45      --    --     --     165

1935         127     45      --    --     --     172

1936         115     45      --    --     --     160

1937         112     45      --    --     --     157

1938         119     45      --    --     --     164

1939         112     45      --    --     --     157

1940         114     63      --    --     --     177

1941         117     62      --    --     --     179

1942         137     47      --    --     --     184

1943         148     46      --    --     --     194

1944         145     47      --    --     --     192

1945         167     48      --    --     --     215

1946         162     51      --    --     --     213

1947         27      190     --    --     --     217

1948         23      245     --    --     --     268

1949         26      220     --    --     --     246

1950         25      223     --    --     --     248

1951         40      257     --    --     --     297

1952         31      252     --    --     --     283

1953         28      262     --    --     --     290

1954         23      250     --    --     --     273

1955         28      272     --    --     --     300

1956         41      273     --    78     78     392

1957         59      271     --    93     93     423

1958         48      272     3     80     83     403

1959         29      275     2     79     81     385

1960         29      275     4     80     84     388

1961         112     270     22    72     94     476

1962         91      253     32    56     88     432

1963         88      249     23    69     92     429

1964         104     236     21    70     91     431

1965         131     235     23    59     82     448

1966         167     219     52    37     89     475

1967         179     209     67    42     109    497

1968         171     203     35    47     82     456

1969         232     217     --    --     97     546

1970         287     250     --    --     95     632

1971         17      547     --    --     8      572

1972         34      522     --    --     28     584

1973         24      483     --    --     13     520

1974         47      506     --    --     --     553

1975         27      533     --    --     --     560

1976         27      500     --    --     --     527

1977         17      446     --    --     --     463

1978         11      351     --    --     --     362

4/79         16      351     --    --     --     367

1/80         70      351     --    --     --     421

Note: Figures for 1934 through 1968 were taken from Attachment,
Noble Melencamp to Jon Huntsman, July 9, 1971, Nixon Presidential
Materials Project. Figures for 1969 through 1975 were drawn
from White House Study Project, Report No. 1 [no author],
"Analysis of Present White House Office," December 7, 1976, Ford
Presidential Library. The 1976 through April 1979 figures were
reported in Personnel Committee (Lipshutz, Jordan, H. Carter,
and Harden) to the president, May 17, 1979; Chief of Staff's
Office, Hamilton Jordan files, Box 57, "White House Staff [CF,
O/A 647]," Carter Presidential Library. The 1980 figures was
reported in "Employees, Consultants, and Detailees," [no author,
no date], Hugh Carter files, Box 24, "Detailees [1]," Carter
Presidential Library.

[a.] Figures are reported as of June 30.

GRAPH: FIGURE 1. White House Office, 1934 to 1980.

GRAPH: FIGURE 2. Congressional Relations.

GRAPH: FIGURE 3. Public Liason.

GRAPH: FIGURE 4. Speechwriting.

Bedard, Paul. 1998. Clinton legal team balloons, creating a new controversy. Washington Times National Weekly Edition, March 22, 1, 22.

Bonafede, Dom. 1973. Effect of Nixon reorganization remains subject of debate. National Journal 5 (March 10): 329-40.

Burke, John P. 1992. The institutional presidency. Baltimore: Johns Hopkins.

Cable, William. 1981. Miller Center interview. Carter Presidency Project. Volume IV, September 18-19, Carter Presidential Library.

Collier, Kenneth E. 1997. Between the branches: The White House Office of Legislative Affairs. Pittsburgh: University of Pittsburgh Press.

Cronin, Thomas E. 1980. The swelling of the presidency. Saturday Review 1 (February 1973): 30-36. Reprinted in Classics of the American Presidency, edited by Harry A. Bailey, Jr. Oak Park, IL: Moore.

Davis, Eric L. 1979. Legislative liaison in the Carter administration. Political Science Quarterly 94 (Summer): 287-301.

-----. 1983. Congressional liaison: The people and the institutions. In Both ends of the avenue: The Presidency, the Executive Branch, and Congress in the 1980s, edited by Anthony King. Washinngton, DC: American Enterprise Institute.

Dickinson, Matthew J. 1997. Bitter harvest: FDR, presidential power and the growth of the executive branch. New York: Cambridge University Press.

Hart, John. 1995a. President Clinton and the politics of symbolism: Cutting the White House staff. Political Science Quarterly 110 (Fall): 385-404.

-----. 1995b. The presidential branch: From Washington to Clinton. 2d ed. Chatham, NJ: Chatham House.

Hartmann, Robert T. 1980. Palace politics: An inside account of the Ford years. New York: McGraw-Hill.

Hult, Karen M., and Charles E. Walcott. 1998. Policymakers and wordsmiths: Writing for the president under Johnson and Nixon. Polity 30 (Spring): 465-87.

Kernell, Samuel. 1989. The evolution of the White House staff. In Can the government govern? edited by John E. Chubb and Paul E. Peterson. Washington, DC: Brookings Institution.

Light, Paul C. 1995. Thickening government: Federal hierarchy and the diffusion of accountability. Washington, DC: Brookings Institution.

Maltese, John Anthony. 1995. Spin control.' The White House Office of Communications and the management of presidential news. 2d ed. Chapel Hill: University of North Carolina Press.

McDonald, Al. 1981. Miller Center interview. Carter Presidency Project. Volume II, March 13-14, Carter Presidential Library.

Moe, Terry. 1993. Presidents, institutions, and theory. In Researching the presidency: Vital questions, new approaches, edited by George C. Edwards III, John H. Kessel, and Bert A. Rockman. Pittsburgh: University of Pittsburgh Press.

Moore, Frank. 1981. Miller Center interview. Carter Presidency Project. Volume IV, September 18-19, Carter Presidential Library.

Neustadt, Richard E. 1990. Presidential power and the modern presidents. New York: Free Press.

Pika, Joseph A. 1991. Opening doors for kindred souls: The White House Office of Public Liaison. In Interest group politics, 3d ed., edited by Allan A. Cigler and Burdett A. Loomis. Washington, DC: Congressional Quarterly.

Ragsdale, Lyn. 1998. Vital statistics on the presidency: From Washington to Clinton. Rev. ed. Washington, DC: Congressional Quarterly.

Ragsdale, Lyn, and John J. Theis III. 1997. The institutionalization of the American presidency, 1924-92. American Journal of Political Science 41 (October): 1280-1318.

Robertson, Stephen L. 1997. Executive office of the president: White House Office. In Cabinets and counselors: The president and the executive branch. 2d ed. Washington, DC: Congressional Quarterly.

Tenpas, Kathryn Dunn, and Matthew J. Dickinson. 1997. Governing, campaigning, and organizing the presidency: An electoral connection. Political Science Quarterly 112 (Spring): 51-66.

Walcott, Charles E., and Karen M. Hult. 1995. Governing the White House: From Hoover through LB]. Lawrence: University Press of Kansas.

-----. 1998. White House staff size: A new look at an old question. Paper presented at the annual meeting of the Western Political Science Association, March.

A Washington correspondent. 1929. The secretariat. American Mercury (December):385. Weko, Thomas J. 1995. The politicizing presidency. Lawrence: University Press of Kansas. Wexler, Anne. 1981. Miller Center interview. Volume I, February 12-13, Carter Presidential Library.

~~~~~~~~

By Charles E. Walcott and Karen M. Hult

Charles E. Wallcott is a professor of political science at Virginia Polytechnic Institute and State University. His primary research interests are theories of political organization in general and of the institutional presidency in particular. Among his recent publications are Governing the White House: From Hoover through LBJ (Kansas, 1995) and Policymakers and Wordsmiths: Writing for the President under Johnson and Nixon (Polity, 1998), both coauthored with Karen Hult. Karen M. Hult is a professor of political science at Virginia Polytechnic Institute and State University. Her primary research interests are the U.S. executive, organization theory and institutional design, and social science methodologies. Among her recent publications are Governing the White House: From Hoover through LBJ (Kansas, 1995) and Policymakers and Wordsmiths: Writing for the President under Johnson and Nixon (Polity, Spring 1998), both coauthored with Charles Walcott.