Printer Friendly

Gubernatorial use of the item veto for narrative deletion.

In the policy science literature and in political discourse, a common misconception about the line-item veto in the states is that it is primarily (if not always) used to eliminate or reduce items of expenditure that appear in appropriations bills as specific dollar amounts. However, studies from Georgia (Lauth and Reese 1993) and Wisconsin (Gosling 1986) indicate that sometimes, if not frequently, governors use the line-item veto to remove narrative sections of appropriations bills. Such narrative may prescribe, proscribe, or otherwise delineate spending requirements; it defines the purpose of the expenditure. (1) In other instances, narrative may be a policy statement or rider that is largely or totally unrelated to any appropriation. The vetoing of narrative sections does not necessarily (though it may) result in the reduction or deletion of specific dollar amounts from appropriations.

The potential use of the item veto in protecting the executive budget from legislation in the appropriations process is not a new concept. In 1901, in Commonwealth v. Barnett (48 A. 977), the Pennsylvania Supreme Court noted that an objectionable practice for which the item veto was a remedy concerned the placement of riders in appropriation bills so as to "coerce the executive to approve obnoxious legislation or bring the wheels of the government to a stop for want of funds." Has the remedy worked? That is, is the item veto a restraint on the use of appropriations bills for the purpose of legislating rather than just appropriating? In addressing this question, we rely on data obtained from (1) a review of court cases, (2) a cross-sectional study of state appropriations acts for two recent fiscal years, and (3) two surveys of executive and legislative budget directors from the 50 states. (2)

The Role of Narrative

The use of narrative within budget bills and the use of the item veto against it varies from state to state. Even so, Rosenthal (1998, 315-16) reports that states increasingly are including policy in appropriations bills: "[H]uman services, education, and experimental voucher programs all find their way into the budget." He notes, "It facilitates passage of legislation under the budget umbrella. Wisconsin has pioneered in this area, with its budget bill running to 1,557 pages in 1993 (having been 281 pages in 1973), filled as it was with nonbudget policy items." A current example comes from the budget conflict in the state of New York in 1998. After the Senate and House took the unprecedented action of adopting the state's budget without the involvement of Governor Pataki, he vetoed more than 1,000 items, including more than 50 "language bills." The leadership of the House, controlled by the opposition party, challenged the legality of vetoes against the language items but accepted the constitutionality of the governor's actions against dollar amounts. The legislative leadership complained that the veto of language in budget bills was unique. The language items outlined "how, when and where money is to be spent for a variety of programs and state agencies" (Hernandez 1998). In one instance, the governor vetoed a provision that detailed how a community college child care program would be conducted. In another instance, Pataki struck a program intended to help welfare recipients find jobs and a section that set eligibility requirements for disabled workers applying for Medicaid (Hernandez 1998).

Criticism of the use of narrative in appropriations bills has occurred mainly at the national level. Such bills have become mechanisms for legislating new policy, for amending current policy, and for directing the administrative behavior of the executive branch, as well as for the traditional purpose of appropriating funds to previously authorized programs. In describing the congressional process, Schick (1995, 154) notes that "although the rules bar legislation in appropriations acts, many legislative provisions are included each year. (A provision is legislation if it changes existing law or makes new law.) In some cases entire laws have been enacted in appropriations measures; the more common practice has been to insert particular provisions." Schick (1995, 154, 160) adds that limitations are another matter. "These provisions which typically begin with the phrase `provided none of the funds shall be used for' have become the stock-in trade of members who want to make policy through the appropriations process. Hundreds of limitations are enacted in appropriations measures each year."

A major criticism of narrative in the appropriations process is that it infringes on the role of the authorization committees in the legislative process (Oleszek, 1996, 61-66; Schick 1995; Hamilton 2000). For example, in 1995, the Republican leadership in the House used riders in appropriations bills as the main method for enacting its legislative agenda, rather than the traditional method of the authorization process. This strategy involved rule changes designed to enhance the use of riders (Aldrich and Rohde 1996, 12-13). As a consequence of eliminating the authorizing committees from the legislative process, Devins (1987, 464-65) claims, Congress fails to give policy issues sufficient consideration because the appropriations process does not provide adequate time for study.

Besides undermining authorization committees, narrative in appropriations bills increases the difficulty of passing appropriations measures. For example, consider the following account of the belated progress (seven months into the fiscal year) of an omnibus appropriations bill in 1996: "House and Senate conferees have closed many of their dollar differences with the Clinton administration over the VA-HUD portion of an effort to finish fiscal 1996 appropriations. But they remain miles apart on the wording of the bill--specifically the presence of Environmental Protection Agency (EPA) riders the administration finds unacceptable" (Carney 1996, 1035). Riders had led to a presidential veto of a freestanding appropriations bill in December 1995. In April 1996, Congress passed the appropriations as part of an omnibus appropriations bill. The president succeeded in reducing the number of riders, but at the cost of a long delay in funding the government. The use of the appropriations process for legislative purposes contributes significantly to the inability of Congress to conclude its business in a timely fashion.

Devins (1987, 464-65) provides additional criticisms against the practice of adding narrative to appropriations bills. For example, he notes that such measures generate contentious debate between House and Senate. As these measures are enacted in annual appropriations bills, agencies do not know whether to view these statutory changes as temporary or permanent. Finally, Devins complains that legislative attention to policy matters in the appropriations process diminishes deliberation about fiscal matters.

Legislators at both the state and national level have strong incentives to use appropriations bills for legislative purposes, thereby making the continuation of such use likely. These riders or provisions allow legislators to take credit with groups or establish policy credentials to enhance their opportunities for re-election. The appropriations process with bills and continuing resolutions is an annual one, allowing regular opportunities for legislators to affect policy. Rosenthal (1998, 315) notes that large percentages of state budgets are fixed, allowing legislators little discretion; lacking the ability to define appropriations, legislators find adding policy within these bills attractive.

Attaching conditions to funds is probably a more effective way for legislators to influence administrators than statements in authorization bills or conference reports. Members of legislative bodies have become aware of the significance of administrative decision making resulting from the delegation of legislative authority to the bureaucracy. This perception has increased legislators' desire to participate in administrative decisions (Rosenthal, 1998, 16-17). The use of the legislative veto prior to and after the Chadha decision (Fisher 1993) reflects that concern. Indeed, the use of riders is one way legislative bodies can continue to impose legislative vetoes (Kaiser 1980). Riders have required agencies to get the approval of Congress and sometimes committees prior to spending money (LeBoeuf 1992, 469). Legislators can be more direct; Parnell (1980) has shown how limitations in appropriations bills have been used to block IRS enforcement of the tax code without changing the law.

Another factor favoring legislators' support for narrative is that appropriations bills are more veto resistant and subject to control by the legislative leadership. Authorization bills are more likely to be single-subject bills susceptible to an executive veto. Appropriations bills tend to be omnibus bills containing funds necessary to keep government functions operating. Chief executives may be more hesitant to use the veto or to negotiate as vigorously when confronted with an appropriations bill containing unfavorable items (Ross and Schwengel 1982, 76). Finally, when legislation is handled by narrative in appropriations bills, the leadership of each legislative body has more control over the contents of the legislation and may expedite policy making more effectively than acting through the committee system. For example, Aldrich and Rohde (1996, 12-13) maintain that in 1995, House Speaker Newt Gingrich and Republican leaders in the House used riders in appropriations bills to bypass unsympathetic authorization committees.

For these reasons, legislators are not likely to discontinue the use of narrative in appropriations bills. (3) Legislators are not the only supporters of narrative instruction in budget bills. In defending the placement of the power of the purse in the legislative body, Madison proclaimed in Federalist 58 that the power of the purse is the "most complete and effectual weapon with which any constitution can arm the immediate representatives of the people, for obtaining a redress of every grievance, and for carrying into effect every just and salutary measure." On the other hand, George Mason and James Wilson warned during the debates at Philadelphia that the use of indispensable money bills would force the executive to accept objectionable legislation. (LeBoeuf 1992, 460). Nevertheless, even some critics of the use of narrative have accepted Madison's point and are unwilling to condemn completely the use of legislative instructions or riders in appropriations bills. For example, after saying that the use of narrative is a mechanism that Congress uses too frequently and often abuses, LeBoeuf (1992, 493) declares, "It is a useful weapon with which to ward off a despotic Executive." In particular, supporters point to restrictions on the war-making powers of the president through limitation provisions as a necessary use of the appropriations process.

Chief executives may also support the use of budget bills for legislative purposes. They may find the normal legislative process too slow or ineffectual. If the legislative leadership is in the hands of a friendly majority and leadership, the opportunity to move quickly with a policy agenda is enhanced by changing the course of government in appropriations bills. That route was chosen by Ronald Reagan upon becoming president. Similarly, a president facing a hostile Congress may find it more effective to work directly through appropriation committees controlled by the leadership of the opposition party than to work with authorization committees under more decentralized control. For example, in 1999, President Clinton sought to use the appropriations process (Frestel 1999, 2520) to expand the law in regard to hate crimes.

Despite the advantages, chief executives frequently confront unwanted legislation in budget bills. The executive veto is the authority provided in American constitutions to enable chief executives to block undesirable legislation. Its effectiveness in regard to appropriations bills is questionable, even though presidents from Hayes to Clinton have vetoed such legislation. Also, some presidents, such as Reagan and Bush, have claimed that particular narratives are unconstitutional, and therefore are not enforceable, while signing the bill. The major limitation to the executive veto in challenging appropriations bills is that the consequence of its use may be to provoke a financial crisis and lead to a shutdown of government services. This concern is particularly true at the state level, where the legislature often goes home shortly after passing appropriations bills. Faced with this dilemma in 1982, Governor Thomas Kean of New Jersey declared that, although he had considered a veto of the entire state budget, he decided against the action because the resulting fiscal crisis would have seriously damaged the state (Karcher v. Kean, 479 A.2d 405). On the other hand, President Clinton successfully used the veto and the threat of the veto to gain major concessions on monetary and narrative features in appropriations bills. Even so, that process did not increase the efficiency of the government in completing the appropriations cycle by the beginning of the fiscal year; the chief executive must wait on another bill from the legislative branch.

The Law, the Item Veto, and Narrative Deletion

Supporters of the item veto as a mechanism to suppress narrative at the national level have a positive perception of its potential. Spitzer (1988, 124) reports that "the single most important factor spurring the call for the item veto power [at the national level] has been the persistent practice by Congress of attaching riders and other nongermane amendments to impel enactment and avoid a veto." In her critique of the executive veto, Best (1984, 188) says the framers failed to anticipate the use of riders and non-germane amendments to legislation. She calls for the use of the item veto at the national level to solve this failure. Despite such calls, we know little about how the item veto might work. More important, we know little about the use of narrative and the use of the item veto against it at the state level.

In his examination of state law, Briffault (1993, 118889) explains that the item veto as a legislative power of the executive is quite different from the ordinary executive veto. Using the latter, "a president or governor can prevent a legislative measure from becoming law, but cannot change any aspect of the measure that the legislature has already approved." The ordinary veto is a negative power; no new law comes into existence except that passed by the legislature. On the other hand, "every time the governor wields the item veto, even in the most non-controversial setting, the governor is acting affirmatively. To veto an item and approve the remainder of a bill is always to enact a piece of legislation that the legislature has not approved. That is, a law missing an item found in the enrolled bill produces a different result than that approved by the legislature. The `negative' instances of the item veto are a null set" (Briffault 1993, 1189). If the item veto is an affirmative legislative power in removing monetary items, then presumably it can be used affirmatively to remove narrative from appropriations measures. State constitutions do not limit the veto to dollar amounts, but, in general, to items within appropriations bills.

Using a negative-affirmative continuum, Briffault classifies states according to how courts have characterized the effects of the item veto on narrative. Some state courts have perceived the item veto as a limited legislative power and require that it be used only in a negative sense. In these states, courts have held that a veto of purposive language that conditions an appropriation creates a new piece of legislation, thereby giving the governor an improper affirmative power. Briffault maintains that limiting the item veto to a negative role undermines its purpose. According to his argument, proponents of the item veto sought to enable governors to battle the harmful effects of logrolling in legislative bodies, which results in compromises that are often favorable to parochial interests rather than the broader interests represented by the governor. Therefore, for Briffault, states that limit the use of the item veto against narrative language fail to understand its role. Nevertheless, many state courts are concerned about executive encroachment on legislative power.

A frequently cited court case that favors the item veto's use solely as a negative instrument is State v. Holder (23 So. 643), decided by the Mississippi Supreme Court in 1898. In that case, the governor approved of the appropriation amounts but vetoed provisions requiring the adoption of rules by executive officials controlling their future behavior in accordance with guidelines suggested in the legislation. In finding the veto unconstitutional, the court (at 645) said:
   Every bill of the character in question has three essential parts. The
   purpose of the bill, the sum appropriated for the purpose, and the
   conditions upon which the appropriations shall become available. Suppose a
   bill to create a reformatory for juvenile offenders, or to build the
   capitol, containing all necessary provisions as to purpose, amount of
   appropriation and conditions, may the governors approve and make law of the
   appropriation and veto and defeat the purpose of the conditions, or both,
   whereby the legislative will would be frustrated, unless the vetoed
   purposes or conditions were passed by a two-thirds vote of both houses?
   This would be monstrous. The executive action alone would make that law
   which have never received the legislative assent. And after all and despite
   the pragmatic utterances of political doctrines, the executive, in every
   republican form of government, has only a qualified and destructive
   legislative function, and never creative legislative power. If the governor
   may select, dissent, and dissever, where is the limit of his right? Must it
   be a sentence or a clause or a word? Must it be a section, or any part of a
   section, that may meet with executive disapprobation? May the governor
   transform a conditional or a contingent appropriation into an absolute one,
   in disregard and defiance of the legislative will? That would be the
   enactment of law by executive authority without the concurrence of the
   legislative will, and in the face of it.

When courts have limited the use of the item veto against narrative language, governors have not been able to strike conditions or provisions attached to monetary items without vetoing the dollar amounts themselves. That is, if the legislature specifies its desire that the appropriation be contingent upon the proviso, then the governor cannot veto the condition without also vetoing the appropriation. The condition cannot be separated from the funding, as this would result in the governor having an affirmative legislative power. In the most extreme cases, conditions need not involve long narratives, but may be sub-items of appropriations with words indicating purpose. For example, the Oklahoma Supreme Court in State University v. Trapp (113 P. 910) held that items detailing specific expenditures to be made by the state university were not subject to veto. Essentially, the court saw these specifications as directions that were interrelated and a function of a larger sum. The New Mexico Supreme Court outlined the dangers of this approach in Fairfield v. Foster (214 P. 323):
   If we follow that line of reasoning, the Legislature may simply make a
   separate appropriation in any lump sum for each department, or by proper
   language in the general appropriation bill, consolidate the funds for
   almost the entire state government, and, under the guise of `directing' the
   expenditure of the money, limit its application to matters and amounts
   which the Governor believes to be highly injurious in part to the best
   interests of the state, practically compelling him to choose between
   abandoning the veto power, or suspending the operation of the government,
   thus nullifying the provisions of the Constitution under consideration, and
   going back to the very conditions its makers sought to avoid.

In most jurisdictions, a sub-item of a lump sum is not considered a condition and is subject to the veto power. However, in some instances, legislatures have achieved the same effect by attaching narrative language to make the sub-item a proviso or a condition.

In a second group of states, the governor may veto nongermane legislation or riders, but not conditions that are germane to the appropriation. Courts in these states judge the narrative to determine whether it is germane to the item, and therefore a condition that is not subject to veto.4 Briffault reports that a major limitation to this condition/rider test is that courts lack objective standards to determine germaneness. In Iowa, the courts have sought to examine whether the narrative qualifies the amount, subject, or purpose of the appropriation; if a condition does not have this effect, then it is a rider subject to veto. As a result, this test "generates few predictable outcomes and instead results in considerable subjective judicial resolution of political conflicts" (Briffault 1993, 1195). In Virginia, the Supreme Court in Brault v. Holleman (230 S.E.2d 238) considered whether the condition is "tied up" with the appropriation to determine if a veto could be utilized; as applied, this test makes use of the veto against narrative so limited as to place Virginia arguably in the first group of states.

In the second type of states, when the narrative is found to be nongermane to an appropriation, it is usually subject to a gubernatorial veto. However, a distinction may be made between a rider (nongermane to the appropriation but relevant to the subject of the appropriation) and narrative that is considered to be general legislation. Interestingly, the governor's ability to veto these, especially general legislation, may be limited. Legal reasons for opposing the item veto against general legislation include (1) state constitutions that limit the item veto to appropriations within budget bills, thereby excluding its use against general legislation placed within such bills; and (2) the governor and item veto may not be the appropriate office and means to enforce constitutional limitations on the use of general legislation in appropriations bills. Even so, the item veto is used against riders and substantive legislation in many instances. A common legal reason cited is that to allow legislatures to include general legislation in appropriations bills would place governors in the difficult position of risking fiscal crisis in order to reject the legislation through an executive veto. For that reason, for the Massachusetts Supreme Court, the use of the item veto against general legislation becomes a means to maintain the balance between the two branches in the appropriations process (Opinion of the Justices to the House of Representatives, 428 N.E.2d 120).

Briffault (1993, 1196-99) finds a third approach to the courts' treatment of the item veto in narrative language cases from Wisconsin's executive-centered veto. Here, the court allows governors to strike not only narrative sections, but sentences and individual words. The result may be a meaning that is totally different from what the legislature intended. The only limitation imposed by the courts is that what remains after the veto must be a complete and workable law, and the result should be germane to the original bill passed by the legislature. Briffault believes the advantage of this approach is that it allows governors to undo logrolling deals and limits questionable decision making by the courts about riders and germaneness. Although Briffault views the law-making powers of Wisconsin's governor as too broad, he believes this is the result of the nature of the item veto rather than the interpretation of the Wisconsin courts.

Although no other state allows such broad use of narrative deletion as Wisconsin, some states have come to a similar conclusion--that the item veto can be used to change legislative intent, and that all conditions are subject to veto. In abandoning the affirmative-negative test, the Supreme Court of the State of Washington held, "The Governor is free to veto ... without judicial review. The `check' as it has always been, will be the Legislature's two-thirds override. While it may be argued it is difficult for the Legislature to obtain the two-thirds concurrence to call itself into a special veto session or override a veto, these constitutional arrangement are for the people to determine, not this court." (Washington Federation of State Employees v. State, 682 P. 2d 869 at 875).

The Supreme Court in New Jersey has similarly given governors significant authority to veto narrative. In Karcher v. Kean (479 A.2d 416), this court explained, "In our opinion, the selective veto power may be exercised with respect to any subject matter that is included in the appropriations act and is broadly related to the State's fiscal affairs as reflected in the act." The New Jersey Supreme Court noted that if the legislature required the governor to accept general legislation in an appropriations bill or to risk forfeiture of funds for a department or departments, then the delicate balance of powers between the two branches would be largely undermined.

The legal status of the item veto as an instrument to remove the various types of narrative in budget bills varies greatly among the states. Even within states, the status of the item veto in this regard is often undefined by the constitution and by court decisions. Issues concerning the item veto may not have been litigated. In Georgia, for example, gubernatorial use of the item veto remains unchallenged, despite gubernatorial use of it. Even where there has been litigation, it is difficult to classify gubernatorial power because the facts in a particular veto may not match the facts in a veto that has been litigated. Furthermore, practice does not always conform to judicial decisions. Because of these limitations to legal analysis, the best way to determine the role of the item veto may be to examine what is happening within the states. (5)

Use of the Item Veto in Narrative Deletion

To understand how the item veto is used, we twice surveyed the heads of legislative and executive budget officers in the 50 states to get a view of state practices. How widespread is the use of narrative in state appropriations bills? We asked the respondents, "in what manner does the legislature indicate to agencies how incremental appropriations beyond the base budget are to be spent?" Seventy-one percent of the respondents (n = 99) indicated the legislature uses narrative instructions in budget bills to communicate with the executive agencies. At least one respondent in 44 of the 50 states indicated such use. Furthermore, we asked the respondents if the legislature includes substantive legislation in the appropriations bills. Sixty-one percent of the respondents answered positively; at least one respondent from 29 of the states indicated such use of narrative.

In order to explore the question of governors' vetoing narrative, we examined appropriations bills for the years 1993 and 1995 in the line-item veto states. Governors in several of the item veto states did not use the item veto in those years, reflecting the fact that usage is greatly affected by nature of the item veto. However, we were able to examine line-item vetoes in 18 states for 1993 and 22 states for 1995. Line-item vetoes were classified into four types.

Category I: Item vetoes that reduced or negated only a specific dollar amount. A vetoed item in the 1995 South Carolina appropriations act, Judicial Department section, "Other Operating Expenses--$880,000," is an example of a Category I item veto. A reduced item in the 1995 California appropriations act, "For capital outlay, Department of General Services. I reduce this item from $2,233,000 to $1,737,000 by deleting: Rehabilitation of Capital Library and Courts Building-Preliminary plans, working drawings and construction ($496,000)," also is an example of a Category I item veto. This category is the type normally associated with the item veto.

Category II: Item vetoes that negated a narrative passage containing a specific dollar amount. A vetoed narrative passage in the 1993 Kansas appropriations act, "On July 1, 1993, or as soon thereafter as moneys are available, the director of accounts and reports shall transfer $26,725,500 from the social service contingency fund to the social welfare fund" is an example of Category II item veto. A vetoed passage in the 1993 Michigan appropriations act, "From the funds appropriated in section 101 for the state emergency relief program, the department shall make a grant of $90,000 to the food bank council of Michigan for the purpose of supporting the mobile food bank program," also is an example of a Category H item veto. Such narrative may be used in some states to disguise items as conditions to avoid a veto.

Category III: Item vetoes that negated a narrative passage related to appropriations, but not containing a dollar amount. A vetoed narrative passage in the 1993 Georgia appropriation act (for FY 1994), "The Department [of Transportation] is authorized to amend its Travel Line Item with agency funds and other income as needed to accomplish its Transportation program responsibilities," is an example of a Category III item veto. A vetoed passage in the 1993 Iowa appropriations act, "In addition, state departments, agencies, and offices receiving appropriations under this act shall reduce expenditures for dues for organizational memberships and travel costs associated with the organizational memberships which are payable from the operations budget of the state department, agency, or office by a total of 10 percent during the fiscal year beginning July 1, 1993," also is an example of a Category III item veto. These vetoes relate to conditions placed on items of appropriations; whether these conditions are germane to the appropriation is an important question in some states and is determined by the legal standards of those states.

Category IV: Item vetoes that negated a narrative passage unrelated to appropriations and did not contain a dollar amount. A vetoed narrative passage in the 1995 Ohio appropriations act, "The Bureau of Employment Services shall provide on a quarterly basis to the Department of Human Services in a format agreed to by the department, a report containing W-4 information on all new hires in Ohio. Upon receipt of the quarterly report, the department of Human Services shall certify to the Bureau whether the information provided is complete. The information in the report shall be used by the Department of Human Services for effective child support enforcement. The Department of Human Services shall issue a progress report to the President of the Senate and the Speaker of the House of representatives by June 30, 1996," is an example of a Category IV item veto. A vetoed narrative passage in the 1995 Wisconsin appropriations act, "No person shall operate on a highway any school bus having a passenger carrying capacity of 16 or more persons in addition to the operator unless such bus is equipped with at least one mirror which is 7 inches in diameter so located as to enable the operator to see a reflection of the road from the entire front bumper forward to a point where direct observation is possible," also is an example of a Category IV item veto. These vetoes clearly involve riders or general legislation in appropriations bills.

These four types represent a continuum ranging from use of the item veto to reduce or eliminate specific dollar amounts, to use of the item veto to eliminate substantive legislation with no explicitly associated dollar amounts. In 1993, the governors of 18 states cast 790 line-item vetoes. As table 1 shows, only 18 percent were Category I item vetoes, while 82 percent of vetoes included narrative passages. Sixty-one percent of the item vetoes cast in that year negated narrative passages that did not contain specific dollar amounts but were relevant to the appropriation, and 25 percent negated a narrative passage that was unrelated to appropriations (a Category IV item veto). In 1995, the governors of 22 states cast 1,496 line-item vetoes. Similarly, only 23 percent were Category I item vetoes, while 77 percent were vetoes that included narrative passages. Again, 61 percent of the item vetoes cast in that year negated narrative passages that did not contain specific dollar amounts, and 27 percent negated a narrative passage that was unrelated to appropriations (a Category IV item veto). These data belie the conventional view that the line-item veto is used in states primarily to strike monetary items. Clearly, it is also used in major policy battles between governors and legislatures, which are expressed in the narrative found in appropriations bills.

While the data in table 1 are illuminating, they are also misleading. Fifty-four percent of the Category IV vetoes in 1993 were cast in Wisconsin, and 63 percent of such vetoes were also made in that state in 1995. As item vetoes against monetary items are concentrated in states where governors have the reduction power, (6) vetoes against narrative are similarly concentrated in states where governors have significant deletion authority and where there is considerable use of narrative within appropriations bills.

To better determine how the item veto is used in the states against narrative in appropriations bills, we surveyed chief legislative and executive budget officers from the 50 states about gubernatorial use of the item veto against narrative. We had responses from 76 of the 86 officers from states with the item veto states; the responses came from 41 of the 43 states with the item veto. We provided the officials with examples of narrative that might be found in appropriations bills and asked them whether their governors could veto the various lines; the hypothetical narratives presented to the respondents is shown in table 2. As might be expected, the two respondents from states where both respondents replied did not always agree. In part, this reflects the difficulty posed by the legal web surrounding the item veto. However, there was much consistency, and perceptions seem not to have been affected by branch. For example, of the respondents expressing a view about the use of the item veto against a condition (see line 1D in table 2) on an appropriation, 19 of the executives indicated the governor could veto the condition without vetoing the appropriation, and 20 of the legislative officials indicated a similar power.

Lines 1 A, 1 B, and 1C in table 2 might appear to be items subject to veto under traditional interpretations of the item veto; they concern dollars of appropriation. However, in some states they are viewed as part of the purpose of Program ABC and, therefore, to be a part of this larger item and not subject to a veto without a veto of the entire appropriation for Program ABC. In essence, they are also conditions on the appropriation for Program ABC. Of the respondents from states with the item veto, 16 or 21 percent said their governor could not veto one of these lines. In states where governors may veto these lines, the governor may lose that capability to do so if the conditional aspect is more explicitly indicated through the use of narrative.

Line 1D is an example of a limitation combining narrative with a dollar amount. When the respondents were asked about the ability of governors to veto this condition (line 1D), 39 of 67 respondents (58 percent) with knowledge about this issue and from states with the item veto reported their governor could use the item veto against this condition without vetoing the monetary appropriation for Program ABC. Altogether, at least one respondent from 26 states with the item veto indicated the ability to delete this condition.

Governors having the authority to veto conditions make use of it. We asked the respondents to indicate how often the governors had used the item veto against conditions, against substantive legislation, or against dollar amounts. Of the 39 respondents indicating gubernatorial authority to veto conditions (line 1D in table 2), 37 reported use of the item veto and at least one respondent from 23 of the 41 states with the item veto indicated that governors had vetoed conditions. Respondents from nine of these states said that half or more of all item vetoes made by their governors were of conditions as opposed to vetoes for dollars or for substantive matters.

In another example from table 2, we asked the respondents about the governor's ability to veto a restriction (table 2, line 1E) placed in an appropriations bill that stipulated how an agency receiving an appropriation should proceed in making regulations. This restriction did not specifically condition the appropriation of funds upon the agency's compliance (an example of Category IV above), but the restriction was placed next to an appropriation. Again, about half the respondents (34 of 64 with knowledge of this issue and from states with the item veto) reported their governor could veto the narrative without vetoing the appropriation. Altogether, at least one respondent from 23 of the 41 states with the item veto indicated the ability to veto this type of condition. A slightly larger percentage of the respondents reported the governor could veto a more explicit example (table 2, line 2) of substantive or general legislation. Altogether, at least one respondent from 26 of the 41 states with the item veto indicated gubernatorial ability to eliminate this rider or nongermane provision in appropriations bills.

Governors with the authority to veto substantive policy tend to make less use of it than governors with the authority to strike conditions. This result is probably the result of a greater number of conditions in appropriations bills than items clearly of a substantive nature. Of the 34 respondents reporting that the governor has the authority to veto substantive policy (table 2, 2) and who could identify veto usage, only four indicated that 50 percent of all vetoes were used for this purpose. Fifteen of the 34 reported that 10 percent or more of the governor's vetoes were to delete substantive legislation added to appropriations bills.

In table 3, using the perceptions of our respondents, we have divided the governors and their states by authority to cast vetoes against narrative in appropriations bills. These rankings reflect the ability of the governor in each state to veto various types of narrative found in appropriations, as reported by our respondents. In the event of a conflict, we used the ranking of the executive budget official, assuming he or she would have more knowledge of gubernatorial authority. Even so, in making this classification, we have also utilized case law on the item veto to check, and in some cases modify, the reports of the officials. Even with both sets of information (participant observation and the law), classification can be risky given the lack of legal clarity. However, we believe the classification in table 3 is useful and generally accurate.

As should be expected, the classification scheme in table 3 relates to usage; far more vetoes against narrative are cast in states where governors have the greatest authority. The data in table 3 indicate the average number of vetoes cast annually over a two-year period, and, of these, how many were cast in regard to conditions and riders in each state. These numbers are extrapolated from the total number of vetoes reported and the percentages indicated for conditions and riders. Numbers will surely vary in each state according to the circumstances in any given year, but the pattern is clear with usage against narrative increasing with authority.

Of particular interest is that the use of the item veto varies significantly among states where the governor has significant authority to use it against narrative. For example, the item veto is used significantly more often in Wisconsin than in any other Type 4 states. As indicated above, the governor in that state has the unique ability to strike words from a bill, providing the governor with strong incentives to use the authority and the capability to change the purpose of legislation by striking such key words as "not" or even the name of the recipient of an appropriation. While the party opposition in the Wisconsin legislature has sought a constitutional amendment to reduce this authority, the legislature has not found ways to limit effectively the governor's ability to legislate. At the same time, the legislature continues to include a significant amount of narrative in appropriations bills (Farney 1993).

The presence of a Republican governor and a Democratic legislature or a divided legislature (depending on the time) in Wisconsin probably has contributed to its high usage rate. Among the states in table 3 in Types 3 and 4, the presence of partisan division increases the likelihood that the veto will be used for narrative deletion. Of the nine states (Alaska, California, Colorado, Kansas, Massachusetts, New Jersey, South Carolina, Wisconsin, and Wyoming) with the executive controlled by one party and the legislature by another at the time of the survey, all had five or more vetoes of narrative; of the other 17 states in Types 3 and 4, only six had five or more vetoes.

There are other explanations for differences among the states. Legal definitions, other than those used in table 3 for classification, matter. For example, in the state of Washington, with 20 vetoes reported in table 3, governors can veto sections in all bills, not just appropriations bills; however, governors may not veto less than a section unless the bill is an appropriations bill, and then they may veto items within the section (Constitution of the State of Washington, Art. 3, Sect. 12). Here, the item veto is a more powerful instrument.

In West Virginia, the state constitution and judicial interpretations make the use of narrative difficult for the legislature. The governor has significant authority to use the item veto. In this state, the legislature may only increase or decrease items in the executive budget (Brotherton v. Blankenship, 207 S.E.2d 421). Even so, the legislature does add items and narrative. The accommodation between the governor and the legislature in the 1990s probably accounts for the low level of use of the item veto (Kabler 1996).

In Pennsylvania, the governor does not need to use the item veto against narrative that is nongermane because the constitution prohibits its presence in general appropriations bills. While the governor could use the item veto against such general legislation, the governor may tell administrators to disregard it as it is unconstitutional. In Wesbury United Methodist Church v. DPW (597 A.2d 271), the Supreme Court of that state upheld the governor's admonition to the state budget secretary to disregard the conditional language in a 1988 supplemental appropriations bill, on the basis that it was substantive language in a general appropriations bill in violation of the state constitution; gubernatorial authority may produce an outcome without using the item veto.

In Arkansas, the need for narrative removal is much less than in other states because of a constitutional provision prohibiting omnibus appropriation bills. State constitutions generally require that substantive bills be single subject in nature; one of the reasons for this requirement is the same as the justification for the item veto, discouragement of logrolling among legislators and less political difficulty for governors in vetoing the bill. Traditionally, supporters of the item veto have sought to restore the veto that governors lost as the result of the omnibus nature of general appropriations bills. In Arkansas, that need for the item veto is not necessary.


Is the item veto a remedy for the increasing use of legislation within appropriations bills and the ineffectiveness of the executive veto? Several conclusions result from our survey. First, in many states, because of judicial interpretation and legislative prowess, the item veto is ineffective or at least restricted in its applicability to narrative in appropriations bills. The use of the item veto for narrative deletion has led to much litigation and lack of uniformity among the states in the level of gubernatorial authority. When courts continue to distinguish between narrative subject to the veto and that which is not, the role of the judiciary in the appropriations process is greatly increased and the influence of the governor is diminished. In such states, legislatures write bills so as to limit gubernatorial use of the item veto. Altogether, 35 (43 percent) of the 85 respondents from states with the item veto reported the governor had not used the veto or had threatened to use it because the legislature had written the appropriations bills to prevent its use. Of the respondents with Types 1 and 2 powers in table 3, 54 percent (n=28) reported this explanation for gubernatorial inactivity in the use of the item veto, compared to 38 percent of the other respondents (n=53) and 28 percent of those (n=32) with Type 4 (the greatest) powers of narrative deletion. Legislative bodies have apparently taken advantage of judicial decisions to restrict the use of the item veto.

Second, the ability of governors to use the item veto against narrative does not necessarily deter the use of narrative by legislatures. For example, legislatures in Wisconsin and Alaska continue to place considerable narrative in budget bills, even though governors actively veto much of it. Among respondents from Type 4 states (where gubernatorial authority is greatest) in table 3, 84 percent of the 32 respondents reported legislative use of narrative instruction in appropriations bills, compared to a similar report by 64 percent of the 28 respondents from Type 1 and 2 states. An explanation for the persistence is that legislatures like to use appropriations bills for symbolic or partisan purposes, regardless of gubernatorial authority. As noted above, gubernatorial need and use of the item veto for narrative deletion is greatest when the executive and legislative branches are controlled by different political parties. This persistence of use in Type 4 states may also result from the ineffectiveness of other mechanisms available to legislators in the appropriations process. The provision of directions in the reports of appropriations committees or in personal communications between committee chairs and agency heads tends to be ineffective, as these directives lack the force of law and are likely to be resisted when the appointment and removal powers of the governor are strong, encouraging bureaucratic loyalty to the chief executive,

Third, where governors have the authority to veto narrative, it remedies a weakness of the executive veto in the appropriations process and protects the executive budget. Governors who possess the power to veto conditions and substantive legislation use it. This conclusion is apparent from the data pattern in table 3. Furthermore, 72 percent of the respondents from Type 4 states said their governors had used the item veto for policy reasons, while 47 percent of respondents from other states with the item veto gave a similar response. On the other hand, 37 percent of the respondents from Type 4 states reported use of the item veto for efficiency reasons, while 39 percent of the respondents from other states made a similar response. The narrative veto is more about policy than about deletion of "pork barrel." Although spending matters have policy concerns, narrative often raises the policy stakes by changing the rules as to the recipients and purposes of money. The item veto apparently has given some governors the ability to contend with such challenges by legislative majorities. However, the need for the item veto may be reduced by constitutional provisions that either restrict the legislature's ability to include narrative in general appropriations bills or prohibit omnibus appropriations bills.

One thing is certain: Fashioning constitutional solutions is difficult. Consider the modern budget section of the West Virginia constitution, which seeks to protect the executive budget. The governor proposes a budget in a format dictated by law or in the absence of law by the governor. The legislature may only increase or decrease the budgetary items. If the legislature wants to pass supplemental bills of a single-subject nature following the passage of the general appropriations bill, it may do so. The governor has the capability to use the item veto against most narrative that is added by the legislature to appropriations bills. However, the legislature still finds means to bypass the governor. In West Virginia, following the passage of the general appropriations bill and its signing, the conference committee that considered that bill issues a document entitled the "Budget Digest." This document details how funds allocated to agencies should be spent, especially in regard to pork barrel projects. This document is not subject to gubernatorial veto and essentially is a political document with no legal basis. Even so, agencies must consider it for obvious reasons (Charleston Gazette 1998). No structure is likely to remove politics from the appropriations process. The original supporters of the item veto did not expect it would end politics; rather they sought balance between the governor and the legislature (Wells 1924). Arguably, the item veto has done that in some states in regard to the introduction of substantive legislation into the appropriations process.
Table 1 Item Vetoes by Category, 1993 and 1995

                          1993                     1995

                  Number     Percentage    Number     Percentage
                 of vetoes   of vetoes    of vetoes   of vetoes

Category I (a)     144            18        337            23
Category II        167            21        247            17
Category III       281            36        502            34
Category IV        198            25        410            27
TOTAL              790 (b)       100      1,496 (c)       101

(a) See text for description of categories.

(b) Based on vetoes cast in 18 states.

(c) Based on vetoes cast in 22 states.
Table 2 Hypothetical Appropriation

Mock Items of Appropriation

1. Program ABC                                            $20,000,000
   1a. Operation A                                         $5,000,000
   1b. Operation B                                         $5,000,000
   1cc. Operation C                                       $10,000,000
   1d. No Project Funded By Program ABC May Exceed           $250,000
   1e. In Administering Program ABC, the State Agency
       May Not Make Regulations Without Considering
       Costs and Benefits.

2. The Speed Limit On Class A Roads is Set at 75 Miles Per Hour.
Table 3 Perceived Gubernatorial Power to Delete Narrative by
State and Number of Vetoes (a)

Type 1 States (b)
Least power

Alabama (0)(0)
Maryland (0) (0)
Minnesota (5)(0)
Mississippi (2)(0)
Oklahoma (4)(0)
Utah (1)(0)
Virginia (1)(0)

Type 2 States (c)
Most power

Arizona (0)(0)
Connecticut (0)(0)
Delaware (0)(0)
Hawaii (0)(0)
Idaho (3)(0)
Illinois (25)(0)
Michigan (29)(3)
Oregon (0)(0)
Tennessee (0)(0)
Texas (10)(0)

Type 3 States (d)

Colorado (8)(8)
Florida (17)(11)
Iowa (11)(2)
Kansas (24)(7)
Mass. (100)(80)
Missouri (15)(0)
Montana (0)(0)
New Mexico (37)(12)
N. Dakota (1)(0)
Ohio (26)(16)
S. Dakota (2)(2)

Type 4 States (e)

Alaska (133)(27)
Arkansas (9)(6)
California (50)(10)
Georgia (2)(2)
Kentucky (7)(6)
Louisiana (38)(12)
Nebraska (5)(0)
New Jersey (18)(5)
New York (4)(4)
Pennsylvania (33)(2)
S. Carolina (17)(9)
Washington (28)(20)
West Virginia (2)(0)
Wisconsin (267)(227)
Wyoming (8)(8)

(a) The perception of veto power was based on responses to questions
posed in table 2, adjusted by case law an item vetoes in each state.
The number of vetoes was based on the average number of vetoes
reported in two previous sessions of the legislature (first number
shown) multiplied by the percentage of vetoes perceived to be of
conditions and of general legislation. The second number or darker
number is estimated number of vetoes of conditions and nongermane

(b) Governor is not able to veto sub-items containing dollar figures
or conditions (germane or nongermane) or general legislation.

(c) Governor is not able to veto conditions (germane or nongermane)
or general legislation.

(d) Governor is able to veto nongermane conditions and general

(e) Governor is able to veto conditions and general legislation.


(1.) Narrative in appropriations bills is not a recent phenomenon. In regard to limitation riders, Devins reports that in 1878, Congress attached a rider prohibiting the use of Army personnel at polls on election day. "By the 1940s, the use of riders was so widespread that the Joint Committee on the Organization of the Congress recommended that the practice of attaching legislation to appropriations bills be discontinued" (Devins 1987, 462). However, the practice continued; for example, in the period from 1971 to 1977, members of Congress offered 225 limitation amendments to appropriations bills (Devins 1987, 462).

(2.) The first survey received responses from 99 of the 100 legislative and executive officials. The second survey received responses from 85 officials, including 76 from the 86 officials in states with the item veto. The surveys were conducted in 1994 and 1996.

(3.) That national legislators prize their narrative-insertion powers is demonstrated by the wording in the Line Item Veto Act of 1996. In defining the president's ability to veto the dollar amount of discretionary budget authority, the statute excludes "any restriction, condition or limitation in an appropriation law or the accompanying statement of managers or committee reports on the expenditure of budget authority for an account, program, project, or activity, or on activities involving such expenditure (110 Stat. 1208-09 Section 1026(7)(B)(iv))." After examining the conference report on this legislation, Fisher (1997, 15) says the legislative intent was to be clear that presidential authority did not extend beyond deletion of dollar amounts.

(4.) For example, may the governor veto a provision in a state budget bill prohibiting the expenditure of state funds to finance the appearance on campus of state colleges persons who advocate the violent overthrow of the government? In Henry v. Edwards (346 So.2d 153), the Louisiana State Supreme Court held that this provision lacked "connexity" and was more in the nature of a general law.

(5.) Another legal factor that may affect the use of the item veto against narrative is the availability of the amendatory veto. This gubernatorial authority allows governors to return enrolled bills to the legislature with suggested amendments and depends on the legislature to approve the changes suggested. In effect, the veto conditions gubernatorial approval of the bill upon the changes suggested by the governor. Arguably, this power might make the item veto unnecessary for narrative deletion.

The amendatory veto does not obviate the potential use of the item veto. First, the amendatory veto is found in only seven states, and in one of them, South Dakota, it may be used only for technical corrections. Second, the amendatory veto requires the legislature to take action before the governor can be successful, whereas the item veto requires the legislature to take action to override the governor's position. In addition, governors may not have the luxury of presenting an amendatory veto to the legislature, as appropriations bills are not usually finished until late in the session. The problem of timing can even be present for the use of the item veto, though it is less of a problem than for the amendatory veto, which requires legislative action. Consider the law in Alabama that has both an item veto and an amendatory veto. The Supreme Court in that state maintained that the "pocket veto" only applies to the executive veto and the item veto must be presented to the legislature prior to adjournment in order to be effective. As this court noted in Hunt v. Hulbert (588 So.2d 848), the tendency of the legislature to submit important appropriations bills at the end of the session, even though the constitution requires that they receive" paramount" treatment, reduces the effectiveness of the item veto. The effectiveness of the amendatory veto is even more questionable.

The presence of an item veto raises significant constitutional questions about the applicability of the amendatory veto to appropriations bills or items; that is, why include both unless they are different, with the item veto for appropriations bills and the amendatory veto for substantive bills? The Alabama Supreme Court suggested such a difference in Hunt by reference to the proceedings of the Alabama Constitutional Convention. On the other hand, in Illinois, discussion at that state's constitutional convention suggested the amendatory veto might be a substitute for the item veto, although the convention included both in the Illinois constitution (Dillard 1988, 600). The courts of Illinois have not allowed the use of the item veto against conditions without a veto of an appropriation to which they apply, and it probably would reach the same conclusion about the use of the amendatory veto. Furthermore, in Illinois the failure of the legislature to adopt a bill as amended by the governor results in the legislation being nullified without a chance for the governor to reconsider; this prospect in regard to the state budget would probably be troubling to the state governor.

(6) Governors in 13 states have the ability to reduce as well as to negate monetary amounts. For the effects of the reduction authority on use of the item veto, see Abney and Lauth (1997).


Abney, Glenn, and Thomas P. Lauth. 1997. The Item Veto and Fiscal Restraint. Journal of Politics 57(3): 388-94.

Aldrich, John H., and David W. Rohde. 1996. The Republican Revolution and the House Appropriations Committee. Paper presented at the 1996 meeting of the Southern Political Science Association, November 7-9, Atlanta, Georgia.

Best, Judith A. 1984. The Item Veto: Would the Founders Approve? Presidential Studies Quarterly 14(1): 183-88.

Briffault, Richard. 1993. The Item Veto in the States. Temple Law Review 66(4): 1171-1206.

Carney, Dan. 1996. Conferees Close on Funds, Not Words. Congressional Quarterly Weekly Report, April 20, 1035. Charleston Gazette. 1998. Indigestible Odious Practice Continues. June 24, 4A.

Devins, Neal E. 1987. Regulation of Government Agencies through Limitation Riders. Duke Law Journal 1987(3): 456-500.

Dillard, Kirk. 1988. The Amendatory Veto Revisited: How Far Can the Governor's Magic Constitutional Pen Reach. Illinois Bar Journal 79(11): 598-603.

Farney, Dennis. 1993. When Wisconsin's Governor Wields Partial Veto, the Legislature Might as Well Go Play Scrabble. Wall Street Journal, July 1, A16.

Fisher, Louis. 1989. How Tightly Can Congress Draw the Purse Strings? American Journal of International Law 83(4): 758-66.

--.1993. The Legislative Veto: Invalidated, It Survives. Law and Contemporary Problems 56(4): 273-92.

--. 1997. Line Item Veto Act of 1996: Heads-up from the States. Public Budgeting and Finance 17(2): 3-17.

Frestel, Karen. 1999. CJS Compromise Speeds toward a Final Confrontation. Congressional Quarterly Weekly Report, October 23, 2517-20.

Gingrich, Newt, et al., 1994. Contract with America. New York: Times Books.

Gosling, James J. 1986. Wisconsin Item Veto Lesson. Public Administration Review 46(4): 292-300.

Hamilton, Lee. 2000. What I Wish Political Scientists Would Teach about Congress. PS: Political Science and Politics 23(4): 757-64.

Hernandez, Raymond. 1998. Assembly Democrats File Suit Challenging Vetoes by Pataki. New York Times, June 16, B6.

Kabler, Phil. 1996. Caperton Hesitant to Use Line-Item Veto. Charleston Gazette, April 12, 1A.

Kaiser, Frederick M. 1980. Congressional Action to Overturn Agency Rules: Alternatives to the "Legislative Veto." Administrative Law Review 32(4): 667-710.

Lauth, Thomas P., and Catherine Reese. 1993. The Line-Item Veto in Georgia: Incidents and Fiscal Effects. Paper presented at the annual meeting of Georgia Political Science Association, February 26, Savannah, Georgia.

LeBoeuf, Jacques B. 1992. Limitations on the Use of Appropriations Riders by Congress to Effectuate Substantive Policy Changes. Hastings Constitutional Law Quarterly 19(2): 457-93.

Mayhew, David R. 1974. Congress: The Electoral Connection. New Haven, CT: Yale University Press.

Oleszek, Walter J. 1996. Congressional Procedures and the Policy Process. Washington, DC: Congressional Quarterly Press.

Parnell, Archie. 1980. Congressional Interference in Agency Enforcement: The IRS Experience. Yale University Law Journal 89(7): 1360-94.

Rosenthal, Alan. 1998. The Decline of Representative Democracy. Washington, DC: Congressional Quarterly Press.

Ross, Russell M., and Fred Schwengel. 1982. An Item Veto for the President? Presidential Studies Quarterly 12(1): 66-79.

Schick, Allen. 1995. The Federal Budget. Washington, DC: Brookings Institution.

Spitzer, Robert J. 1988. The Presidential Veto: Touchstone of the American Presidency. Albany, NY: State University of New York Press.

Thompson, Pat, and Steven R. Boyd. 1994. Use of the Item Veto in Texas. State and Local Government Review 26(1): 38-45.

Wells, Roger, 1924. The Item Veto and State Budget Reform. American Political Science Review 18 (4): 782-91.

Glenn Abney is a professor and chair of the Department of Political Science, Georgia State University. Email:

Thomas P. Lauth is a professor and head of the Department of Political Science, University of Georgia. Email:
COPYRIGHT 2002 Wiley Subscription Services, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2002 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Abney, Glenn; Lauth, Thomas P.
Publication:Public Administration Review
Article Type:Statistical Data Included
Geographic Code:1USA
Date:Jul 1, 2002
Previous Article:Does regionalism beget regionalism? The relationship between norms and regional partnerships for economic development.
Next Article:The evolution of policy analysis.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters