Journal Issue

Who Controls the Budget

Ian Lienert
Published Date:
June 2005
  • ShareShare
Show Summary Details

Sometimes it is said that man cannot be trusted with the government of himself. Can he then be trusted with the government of others?

Thomas Jefferson, 1801

I. Introduction

This paper discusses the budgetary powers of legislatures under different forms of government. In presidential systems, where elections for members of the legislature are independent from the elections for the presidency, the legislature is a powerful agenda-setter and decision-maker. The annual budget encapsulates the legislature’s wishes with respect to taxation and expenditure policies. The United States is a prime example of this form of government. Both houses of the Congress have unfettered powers to amend the President’s draft annual budget. Also, by adopting laws relating to budget processes (as opposed to the laws it adopts on budget policies), the legislature is able to reinforce its budgetary control over the executive.

The United Kingdom is at the other end of the spectrum. In a parliamentary monarchy of the “Westminster” type, the effective head of the executive branch is both a member of parliament and the leader of a major political party—usually the party with the most seats in the lower house. Within the government, the budget-making powers of the Prime Minister and his/her Cabinet of Ministers are very strong. By tradition, the Head of State, the Monarch, follows the advice of his/her ministers. Although Parliament votes on the annual budget, if the government has a majority in Parliament, the government alone determines the shape and size of the budget.

Between these two extremes there are semipresidential systems (e.g., Finland, France), parliamentary republics (e.g., Germany and Italy), and non-Westminster parliamentary monarchies (e.g., Denmark, Netherlands, Norway, Sweden). Relative to strongly separate presidential systems or weakly separate Westminster parliamentary system, the budgetary powers of legislatures of these countries could be expected to occupy an intermediate position.

Can generalizations be made about these other forms of government? Can one rank the budgetary powers of the legislatures of semipresidential systems and other parliamentary systems in an unambiguous order—with countries with monarchies expected to be closest to the Westminster model and the budgetary powers of semi presidential regimes closest to pure presidential systems? Does it matter if there is a multiplicity of political parties, as opposed to only two dominant political parties?

This study is confined to an examination of the nexus between the separation of powers and the legislature’s budgetary authority in selected countries with different forms of government. The sample includes the Nordic countries and a few Asian and Latin American countries. The overriding objective is to discern whether the legislature’s budgetary powers are linked to the degree of separation of powers. If not, then what are the factors that explain differences in budgetary powers across countries?

The paper is laid out as follows. Part II briefly reviews relevant literature. Part III describes five different forms of government, particularly the degree of the legislature’s political control over the executive. In Part IV, specific areas of budgetary control by the legislature are identified— mainly those exercised prior to the adoption of the annual budget. An index of the power of the legislature to shape the annual budget is developed. Various factors that determine the legislature’s budgetary powers are identified. Part V highlights the importance of country-specific political, legal, and other factors that attenuate or strengthen the legislature’s budgetary powers.

II. Renewed Focus on the Budgetary Powers of the Legislature

Research on the budgetary powers of legislatures has been conducted in individual countries, particularly the United States (e.g., Schick, 2002a, Shuman, 1992), but also Germany (e.g., Wehner, 2001), France (e.g., French National Assembly, 1999), Japan (Meyer and Naka, 1998). Comparative research on these issues for parliamentary systems is less frequent (e.g., Krafchik and Wehner, 1998). Mezey (1979) classified legislatures into four groups: active, reactive, marginal and minimal. In a study of six European countries, Coombes (1976) noted the complexity of diverse systems and concluded that more discussion on fundamental difference of constitutional parliamentary arrangements and procedure was needed. Schick (2002b) addressed whether the legislature can regain some of its budget powers. More recently, tools for legislative oversight in a large number of countries have been examined (e.g., Pelizzo and Stapenhurst, 2004) and Wehner (2005) developed an index of the legislature’s budgetary power.

Since the 1990s, the relative roles of the legislature and the executive have been discussed in studies of the institutional constraints that contribute to controlling budget deficits and debt (e.g., von Hagen, 1992, 1994). One strong conclusion of this literature is that a strong central budget authority in the executive—and strong constraints on the legislature’s budget amendment powers—is necessary for disciplining legislatures and maintaining sustainable fiscal positions. A model of the interaction between political parties and fiscal policy performance was developed by Baldini (2000) and tested for 17 countries.

III. Separation of Political Powers under Different Forms of Government

A. Forms of Government

There is a diversity of forms of government (Laundy, 1989; Cheibub and Limongi, 2002) and different ways of classifying them (e.g., Pizzorusso, 1988). The main two forms of government are presidential and parliamentary systems. In this paper, four parliamentary systems are also identified.

Presidential systems

The main characteristic of a presidential system is that the head of the executive, the president, is directly elected by citizens.

“Pure” presidential systems. An elected President appoints a Cabinet composed of individuals that support his/her political agenda. There is no prime minister. A vice-President is common. In contrast to parliamentary systems, Cabinet members are chosen from outside the chamber(s) of elected representatives. The President is both the head of State and the head of the executive. He/she may usually appoint other political supporters to high-level councils and management positions in the civil service.

Semi-presidential system. This system features both a prime minister and a president. It differs from the “pure” presidential system in that it has a prime minister who is generally drawn from the legislature and who is responsible to it. The sharing of powers between the president and the prime minister varies considerably in semipresidential systems. Depending on the country, both are active participants in the functioning of government, although the Prime Minister often has more authority in budget matters. This is especially the case in semipresidential countries where the president is a ceremonial figurehead.

Other parliamentary systems

Parliamentary systems usually have a clear differentiation between the head of government and the head of state. Another important feature of parliamentary systems is that there is no clear-cut separation of powers between the legislative and the executive. This is mainly because the survival of the government is dependent on the support of the parliament. The government can be removed by the parliament through a vote of no confidence. On the other hand, the executive may have the power to dissolve the parliament and call elections. Parliamentary systems vary as to the number of parties within the system and the dynamics between the parties.

In parliamentary systems, the prime minister is the head of the government. Typically, both the prime minister and the Cabinet of ministers have their background in the parliament. The leader of the largest political party in the parliament is usually appointed as the prime minister. The position of the head of state is primarily ceremonial in Westminster and non-Westminster monarchies. This may also be the case in countries where the president is directly or indirectly elected by parliament (parliamentary republics). Differences between these three forms of parliamentary systems are outlined in Annex Table 1.2 In Westminster systems, Cabinet ministers must be members of parliament—not necessarily the case in other parliamentary systems. In all three systems, the head of state usually retains some reserve powers, for use during crises. In most cases however, such powers are exercised only upon the advice and approval of the head of government.

B. Separation of Political Powers of the Executive and the Legislature

Separation of Political Powers and limitations on executive power under different forms of government. The degree of separation of the legislature from the executive depends on a number of factors. Seven of these are presented in Part A of Annex 1. The notion of “separation” or “independence” of the two branches of government is closely linked with the issue of control of the executive by the legislature (or vice versa). Five factors of control of executive by the legislature are described in Part B of Annex 1.

For five different forms of government, there is a broadly linear relationship between the separation of powers and the control of the executive by the legislature (Figure 1). It is apparent that in presidential systems, the political separation of powers and the control of the executive by the legislature are particularly high. The Westminster system is at the opposite extreme. Between these two extremes, the differences between the characteristic of semipresidential regimes, republican parliamentary systems and non-Westminster parliamentary monarchies are not great. As expected, in semipresidential systems the legislature is slightly more separate.

Figure 1.Separation of Political Power and Control of Executive By the Legislature

Source: Indices described in Annex 1

It could appear tautological that the legislature controls the executive (or vice versa) in the case of non-separation of powers. However, the notion of separation of powers—as original conceived in the 18th century by Montesquieu—was based on two principles: independence and specialization of each authority, which does not necessarily imply that control is exercised by one or the other authorities. In later developments on the philosophy of the separation of powers, de Malberg (1922) introduced the notion of heirarchy, which could imply control by the higher authority (Troper, 1992). However, the hierarchy of organs is not necessarily matched by a hierarchy of powers. The Westminster system is a clear example of this: whereas in budget matters, parliament (the hierarchically superior organ) is in reality subservient to the executive (which is functionally the hierarchically superior body).

IV. Authority of the Legislature to Shape the Budget

If a strong linear relationship exists between separation of powers and political control of the legislature over the executive, does it necessarily follow that the separation of political powers is also accompanied by strong budgetary control of the legislature over the executive?

A framework for analyzing how legislatures influence budgets was proposed by Oppenheimer (1983). There are six dimensions, including the stages of legislative budget processes, which is the focus of the following sub-section.

Ex ante budget powers. An important question is “does the legislature formally approve a binding medium-term budget framework that guides the executive in preparing each year’s annual budget?” If so, does it concern merely the broadest aggregates—total revenue, total expenditure, the overall balance and the level of debt—or does it also include the allocation of spending. The latter question is particularly important for the legislature.

A second question—and perhaps the most important power the legislature has over the annual budget—is whether it can amend it without restriction? The most unrestricted legislatures are those that can change revenues, expenditures and the balance in any direction, without limit. The most restricted legislatures are those that always adopt the budget exactly as proposed by the executive.

A legislature that has limited time to examine the budget and propose changes—because it is does not have supportive legislative committees or a nonpartisan budget office—is clearly at a disadvantage relative to a legislature that has adequate time to formulate alternative tax/expenditure proposals, an elaborate committee structure and adequate research capacity.

During budget execution. The annual budget law provides legally binding upper limits for discretionary expenditures. The legislature may allow the executive to exceed individual items of spending, provided a more aggregate sub-totals are not exceeded. Rules for virement within expenditure categories are usually laid down in law. Very strong control by the legislature would disallow any virement. For non-discretionary spending, legislatures allow the executive to exceed indicative budget limits (e.g., for spending on social transfers, interest on debt, since a separate law may require such expenditure to be paid irrespective of the budgeted amounts). The legislature may also restrict the executive from reducing approved expenditures, even if revenue shortfalls should occur. This is more likely in presidential than in parliamentary forms of government, as an independent legislature would like to see its expenditure policies adopted.

Ex post budget authority. An independent external audit office usually provides a report to the legislature on the budget outcome and the annual financial statements of government. When the legislature has strong powers to follow up on audit recommendations, desirable changes in financial management practices by the executive may ensue. Ex post budgetary powers of the legislature are oriented to improving budgetary procedures, and are less likely to affect the shape and size of future budgets.

Index of the legislature’s budgetary authority

On the basis of the most important parameters discussed above, an index of the legislature’s budgetary powers has been established, based on five particularly important criteria (see Annex 2). Some criteria have been excluded from the index on the grounds that they are relatively less important, especially for the determination of the shape and size of the annual budget. These include: the legislature’s power with respect to interim and supplementary budgets; questionings of government officials in committees of the legislature; and the strength of budget-related parliamentary committees.3 These criteria are less important determinants of the size and allocation of revenues, expenditures, and the overall balance.

V. Separateness of the Legislature and Its Budget Authority

Relationship between separateness of the legislature and its budgetary authority

Although there appears to be a broadly linear relationship between the separateness of the legislature and the legislature’s control over the executive as measured by political variables (Figure 1), this 1:1 relationship does not seem to hold for budgetary variables (Figure 2). The selected budgetary powers of non-Westminster monarchies are higher than those of semipresidential systems, contrary to what one expect solely on the basis of separation of political powers.

Figure 2.Separateness of the Legislature and its Budgetary Powers

Source: Annex Table 1 (horizontal axis); Average of Annex Table 2 values (vertical axis).

The actual average budgetary power of the legislature, as depicted in the above chart, is influenced by the inclusion of two semipresidential countries—Ireland and Greece—whose legislatures have particularly weak budgetary powers. The shape of the chart would be different if these two countries were replaced by the other semipresidential countries (e.g., those of Eastern Europe).

Similar observations could be made for other countries in the sample. This suggests that, for each of the five forms of government, there are a number of country-specific influences on the budgetary powers of the legislature. These factors are now discussed in turn.

A. Political Variables that Influence Budgetary Powers

Different electoral laws result in differences in the number and strength of political parties. These in turn affect the formation and cohesiveness of governments. Four situations are considered:

  • Political parties and the voting system. Electoral laws may favor the formation of a two-party political system (e.g., a first-past-the-post voting system) or a system with two main blocks of parties (e.g., a two-round voting system). In these cases, the alignment of ideologies results in a government that is more likely to be cohesive compared to a coalition government made up of a multiplicity of parties of diverse ideologies. When there is little internal discord in the government, a strong centralized budget authority— usually headed a Minister of Finance—is likely to emerge (Hallerberg, 2004). Such centralization of authority in the executive is enhanced further when there are mechanisms for enforcing discipline on party members. Strong centralized governments may enhance their budgetary powers by proposing laws to that effect. Alternatively, they may withhold budgetary information from the legislature’s purview.

  • Majority versus minority governments—parliamentary systems. In order to ensure that the annual budget law is adopted, minority governments necessarily need to seek the support of political parties that are not included in the government. Dialogue, negotiation and agreement with non-government political parties are required in such countries. When there are minority governments, the legislature has relatively more budget powers than in countries with majority governments, since the parties not represented in the government can force the minority government to adopt policies or programs that change the allocation and amounts of budget spending.

  • Presidential systems with a multiplicity of political parties. In such situations, there is the risk of deadlock, which depends on the share of seats in the congress that are controlled by the party of the president, as well as specific institutional features, including whether the president possesses veto powers and the majorities needed for overriding presidential vetoes (Cheibub and Limongi, 2002, p.4). Additional complications arise if the vice-president belongs to a political party different from that of the president.4

  • Bicameralism. Provided it has some budgetary powers, a second chamber of the legislature unequivocally strengthens the budgetary powers of the legislature relative to those of the executive. If the members of the second chamber are independently elected, by a voting system different from that of the house of representatives, the two chambers may have differences of view over draft budget law proposals. Depending on the constitutional and legal powers accorded to it, the second chamber can veto budget legislation that has already been adopted by both the government and the first chamber of the legislature. Germany provides a good example of this (Box 1).

  • Politics and “veto players”. The above discussion of multiparty systems and bicameralism is illustrative of more-general theories of the influence of politics on budgetary powers. Some authors have shown that an increasing number of veto players can increase the size of fiscal deficits (Heller, 1997; Tsebelis, 1995). However, this is not necessarily the case. In Sweden for example, in the 1990s the fiscal deficit was reduced very markedly and changed to a surplus, despite the presence of political parties (opposition parties in the cases of minority government) that could have used their “veto” power to derail the decisions that resulted in rapid fiscal consolidation and debt reduction. Similarly, in New Zealand, where the essentially two-party “first-past-the-post” electoral system inherited from Britain was replaced by a German-style mixed member proportional (MMP) representation system in the 1996, it has been shown the government efficiency has not deteriorated—the country continued to maintain fiscal surplus under MMP, a radical departure from the large deficits that had been experienced under the majoritarian electoral system (Boston et. al., 2003).

Box 1.Germany—Bicameralism and Budget Powers of the Legislature

The house of representatives (Bundestag) has strong budgetary powers and a Bundestag budget committee actively examines the draft annual budget law. However, new legislative proposals that affect the budget have to be approved by the government. This allows parties of coalition governments to have their budgetary policies adopted in the Bundestag. The second chamber (Bundesrat), which represents the interests of the regions (Land), possesses considerable budget powers. No tax legislation can be adopted without Bundesrat approval. In principle, only the approval of the first chamber is required for the adoption of expenditure proposals of the annual federal budget. In practice, since the government often does not always hold a majority in the second chamber, the Bundesrat may block the entire budget that has already been approved by both the government and the Bundestag. This problem is acute during the latter half of the Bundestag’s electoral cycles, as the minority parties of the Bundestag typically gain seats in the Bundesrat during Land elections (which have a different electoral cycle from the federation). This results in a loss by the government of its majority in the Bundesrat. Fluctuations in the size of the consolidated deficit of all levels of government appears to be correlated with the government’s majority (or lack thereof) in the Bundesrat.

Sources: Decressin and Braumann, 2004; OECD, 2003; Wehner, 2001.

B. Constitutional and Legal Limitations on the Executive’s Budgetary Powers

The index of budgetary powers of the legislature used in this study is based on five constraints that have generally been included in formal law. This subsection discusses why constitutional or legal constraints may have been imposed on executives.

  • The legislature perceives that it must “win back” powers lost in earlier periods. This is more likely to happen when there is divided government. In presidential systems, when consensus is reached in the legislature that the executive has too much budgetary power, by initiating and adopting a new law, the executive’s budgetary powers can be curbed. This is what happened in the United States in 1974 (Box 2). To a lesser extent, France’s legislature regained some budgetary authority when it adopted a new Organic Budget Law in 2001.

  • In parliamentary systems, governments generally initiate laws, which enable them to strengthen their own budgetary powers. Majoritarian governments with unicameral legislatures are particularly likely to approve the government’s budget proposals, as such governments control the chairs of parliamentary committees that examine draft new laws.

  • Constitutional constraints on the budget are imposed for specific reasons. Political instability and lack of budget control may be linked. To end such instability a constitutional change may be required. This happened in France in 1958, when both new political arrangements and a strong curbing of the legislature’s budget amendment powers were incorporated in the new constitution. In Germany, during the 1960s, the constitution was amended to incorporate principles of Keynesian economics, including exceptions to the Constitution’s prohibition of deficits on current transactions.

  • The legislature may approve the medium-term macro-fiscal strategy. In most OECD countries, a medium fiscal outlook, covering at least two years beyond the new budget year, accompanies the proposed annual budget law. Although legislatures debate the medium-term fiscal strategy, seldom does the legislature take an alternative viewpoint to the executive. In particular, it is rare for the legislature to adopt a law approving binding medium-term expenditure limits. Italy’s parliament approves a medium-term strategy prepared by the government; however, the only binding limits relate to the year following the budget year. Sweden’s parliament also approves such a medium-term strategy—the difference with Italy being that the expenditures limits may be binding for a number of years. The United States is the only country whose legislature adopts a budget resolution that specifies a medium-term macro-fiscal strategy.

  • The legislature may first approve binding aggregates for the total expenditures and the deficit/surplus, and later approve the detailed estimates expenditure. Sweden is one of the few countries which have introduced a two-stage approval process of the annual budget by the legislature (Box 3).

  • Fiscal rules may be adopted to prevent economic crises. Many countries have adopted laws that include fiscal rules—quantitative or qualitative constraints on budget aggregates such as deficits, expenditure or debt (for a review, see Banca d’Italia, 2001; Kopits et al, 1998, Dában et al, 2003). Although it is difficult to reach strong conclusions, fiscal rules generally become more rigid according to the type of government. The legislatures of countries with presidential systems are more likely to adopt a law with quantitative fiscal rules than countries with parliamentary systems, since in the latter the government can propose a deficit-reducing budget each year whose chances of non-adoption are lower. This is especially the case for parliamentary systems with majoritarian governments, as they only have a few veto players. As with legislatures, governments with few veto players can ensure the enactment of a budget of its choice.

  • Supranational organizations may constrain the shape and size of the budget prepared by the executive. This is the case for the EU countries, where scrutiny of national budgets is exercise by the European Commission, with a view to examining conformity with the Maastricht treaty, which has been ratified by all euro area countries, and which imposes limits on general government deficits (3 percent of GDP) and debt (60 percent of GDP).

  • Laws may be passed to create a Budget Office at the Legislature. This is most likely to happen in presidential systems, where a strong and independent legislature seeks to receive alternative budget analyses and policy scenarios to those proposed by the executive. Only a few countries have followed the United States’ example of creating a CBO. In 1998, Mexico (also a presidential system of government) established a Center for Public Finance, attached to the chamber of deputies. Korea, with a semipresidential system of government, also established a National Assembly Budget Office, by adopting a special law in 2003. As in Mexico, it is modeled largely on the United States’ CBO. Apart from these three countries, no other countries have established nonpartisan budget offices serving solely the legislature.

  • Laws may be adopted to increase the length of debate on the budget. It is no coincidence that in presidential systems the allowable debate time on the draft budget is particularly long. The most extreme example is the United States, where the President must, by law, present a draft budget 8 months before the new fiscal year begins. In Indonesia, law requires the budget to be presented to Congress in 4-5 months before the new fiscal year begins. In contrast, most parliamentary forms of government have only 2-4 months to debate the annual budget in parliament. In the Westminster countries, the annual budget law does not even have to be presented to parliament prior to the beginning of the new fiscal year (Lienert and Jung, 2005, Table II.4).

Box 2.The United States Legislature Regains Budgetary Powers

The United States budget process has been analyzed in terms of three cycles (Schick, 2002a):

Pre-1921. The legislature controlled the budget process—there was no central budget office in the executive

1921-1973. The executive dominated budgetary policies and processes.

Post-1973. The legislature regained its budgetary dominance.

In 1974, in reaction to the non-execution of appropriated expenditures by President Nixon, Congress adopted the Congressional Budget and Impoundment Control Act. The law:

-- prevented the executive from refusing to spend congressionally-approved budget programs.

-- limited deferrals of approved expenditures into the next fiscal year.

-- established a Budget Committee in each chamber of Congress (which supplemented, rather than replaced, the many pre-existing congressional committees), with powers to establish the annual budget framework (different from that proposed by the President).

-- created a new non-partisan Congressional Budget Office (CBO), with some 250 staff. The CBO provides Congress with budget-related analyses and other budget information, independently of the Office of Budget and Management under the President.

Box 3.Sweden—The Two-Stage Budget Approval Process

Sweden adopted its first-ever State Budget Law in 1996, in part to clarify the respective budget powers of Parliament (Riksdag) and the government, but more importantly, to formalise measures that taken to strengthen fiscal discipline, notably a two-stage budget approval process. Key provisions are:

  1. In April, “top-down” expenditure ceilings are proposed by the government to the Parliament in a Spring Fiscal Policy Bill. In May, the Riksdag examines this Bill and the Opposition presents alternatives. In June, the Riksdag takes a non legally-binding decision on the Bill.

  2. In September, the Government presents the Budget Bill to the Riksdag. Again the Opposition presents alternatives to the government’s budget proposal.

  3. In November, the Riksdag begins by deciding the expenditure ceiling for the next few years. The Parliamentary Committee on Finance proposes ceilings for each of the 27 separate expenditure areas covered by the Bill and takes a position on taxes and other revenues. Other committees state their opinions on the estimates. After this, each committee makes a proposal on the allocation of the money between appropriations in the expenditure areas for which it is responsible.

  4. In mid-December, the Riksdag takes a formal decision on the detailed annual budget appropriations.

Source: Sweden (2004)

C. Non-Legal Constraints on the Legislature’s Budgetary Authority

In countries with multiple political parties that do not easily form two blocks such as those of the “left” and those of the “right”, agreements between selected political parties play an important role in determining the legislature’s budgetary powers.

  • Coalition agreements—majoritarian government. In parliamentary systems with proportional representation, inter-party agreements takes on special importance. Some European countries with coalition governments have found that clear ex ante agreements on social and economic policies are necessary. In the Netherlands, such agreements have, at times, specified detailed budgetary rules (Blondál et al, 2002). In Finland, the Constitution requires parties that will compose the government coalition to specify their jointly-agreed policies for the parliamentary term. Coalition agreements in such countries have exerted a powerful influential in reducing fiscal deficits and debt levels (Hallerberg, 2004; Schick, 2002b).

  • Annual budget agreements—minority governments. If the government is formed with parties with a minority in parliament, an agreement with opposition parties is necessary to pass the annual budget. This is the case, for example, for Denmark, Norway and Sweden whose legislatures are elected by a proportional representation system (Table 1). These three countries have had mostly minority governments over the past 50 years. Annual budget agreements between the parties comprising the minority government and selected opposition parties are generally reached prior to the presentation of the draft budget to parliament. Although parliamentary committees are active, the most crucial decisions affecting annual budgetary policies and deficits/surpluses occur prior to these discussions (for Sweden and Denmark, see Blondál et al 2001 and 2004, respectively).

Table 1.Interparty Agreements in Non-Westminster Parliamentary Monarchies
CountryCut-off thresh- holdTypical governmentType of agreement between parties for the budgetDuration of agreement
Denmark2%Minority, with more than 1 partyAnnual budget agreement between government and selected opposition partiesOne year (maximum)
Netherlands0.66%Majority, with several parties in coalitionCoalition agreement, with conservative revenue estimates and expenditure ceilingsLife of parliament (4 years maximum)
Norway4%5Minority, of 1-2 main partiesAnnual expenditure ceilings for 21 spending areas, formally endorsed by Parliament in budget discussionOngoing
Sweden4%Minority, of 1 main partyAnnual multi-year expenditure ceilings for 27 spending areas, formally endorsed by Parliament in budget discussionOngoing (based on 1996 State Budget Law)

Coalition government agreements, or budget agreements of minority governments, are voluntary, rather than legal, constraints on participating political parties. For this reason, in non-Westminster monarchies—including Belgium, Denmark, Netherlands, Norway and Sweden— the actual amendments to draft budgets by legislatures are voluntarily constrained, i.e., the legislatures do not take full advantage of their unrestricted legal powers to amend draft budgets (Figure 3).

Figure 3.Legal Amendment Powers and Actual Amendments6

Apart from the political constraint imposed on coalition governments with non-legal agreements to maintain fiscal discipline, in some countries legal constraints are dominant. In particular, if there are constitutional fiscal rules (e.g., France, Germany), a constitutional council (court) is in place to ensure that the rules are not breached. This is not the case in, say, the Scandinavian countries.

D. Other country-Specific Factors

There are several other country-specific factors that determine the budgetary powers of the legislature. Only two are examined in this study:

  • Inheritance from former ruling powers. The budget systems of several countries are largely modeled on those of their former foreign-controlling powers. This is particularly the case in anglophone and francophone Africa, where colonial influences continue to be pervasive (Bouley et. al, 2002; Lienert, 2003). Unless there are strong pressures to change the balance of budgetary powers between the legislature and the executive, it is possible for the balance of these powers to remain unchanged for long periods of time. For example, Ireland has retained most of the main features of the Westminster budget system (Box 4).

  • Specific constitutional provisions may prevent a strong legislature. Mexico’s constitution prohibits consecutive legislative terms by members of the chamber of deputies. Hence, they can never become experienced and influential on congressional budget-related committees. The Mexican legislature has not established a dedicated budget committee. In the past Mexico’s chamber of deputies served to legitimize the president’s budget proposals (Meyers, 2000). The anti-reelection heritage is a barrier to the development of legislative budgeting expertise.

  • The Royal Prerogative. In the United Kingdom, an accepted source of law is the “royal prerogative”—constitutional understandings derived from an earlier age when all executive power resided in the Monarch. These are the powers unique to the executive that the courts recognise it possesses for the purposes of carrying out the business of government. This is sometimes overlooked aspect of executive-parliamentary relations, as it is a source of considerable power exercised by ministers (Winetrobe, 2000). With respect to the budget, it is reason why the executive (not parliament) determines the structure of the annual appropriations and a source of strong Treasury control over expenditures (Daintinth and Page, 1999).

Box 4.Ireland—Inherited Features of Budget System

Ireland’s Parliament (the Dáil) has weak budgetary powers. Many of the following features of the role of parliament and the head of state are due to the inheritance of the Westminster system. The strong budgetary powers of the Irish executive are mirrored in the British budget system. In particular:

-- The President of Ireland has very few independent powers and always follows the advice of his/her ministers.

-- There is no separation of political power. The President is an integral part of Parliament. All Cabinet ministers must also be members of the Irish parliament.

-- The Cabinet of Ministers decides all essential budget issues. The Cabinet’s proposed budget cannot be amended.

-- The Dáil may not vote money unless it is so requested by the government.

-- The second chamber of the Dáil has no authority to amend the annual budget.

-- Parliamentary committees are weak. The most influential committee is the Public Accounts Committee, which focuses on ex post budget execution, not on the ex ante budget.

-- There is no legal requirement for the budget to be presented by a certain date.

Source: Ireland (2005)

VI. Conclusions

This paper first establishes a strong link between the political powers of legislatures and the degree of separation of the legislative and executive branches of government. The separation of political powers in parliamentary monarchies based on the Westminster model is extremely weak. Since cabinet ministers are obliged to be members of parliament, the executive essentially controls the legislature. The political powers enjoyed by the executive are accompanied by strong budgetary powers. Governments in Westminster countries have not only initiated laws to strengthen their own budget powers but, by virtue of their control over parliamentary committees, have also inserted into parliamentary regulations limitations on parliament’s capacity to change the draft annual budget proposed by the government. Mainly for these reasons, the development of strong parliamentary committees to consider the draft budget has been thwarted.

This is in sharp contrast with presidential systems, which have strongly separated powers and powerful legislatures. Budget-related committees at the legislature can flourish. The United States is an extreme example among presidential systems—there are multiple committees and subcommittees in both chambers of Congress (which result from separate and independent electoral systems). These committees are very influential on budget processes, the size of the annual budget, and the allocation of spending. An independent nonpartisan Congressional Budget Office provides a powerful counterweight to the President’s Office of Budget Management, which prepares the initial draft budget. This budget must be submitted to the legislature by the President eight months before the beginning of the new fiscal year—much further in advance than any other country.

Whereas one can be categorical that the authority to shape the size of the annual budget is strong in a presidential form of government and particularly weak in countries with Westminster parliamentary monarchies, it is difficult to reach firm conclusions regarding intermediate forms of government.

Differences in the budget powers of legislatures between semipresidential forms of government, parliamentary republics and non-Westminster parliamentary monarchies pale in comparison with the sharp contrast between presidential and Westminster forms of government. Although these intermediate forms of government have a similar degree of separation of political powers of the executive and legislature, the budgetary powers of legislatures in these countries varies considerably. Differences between countries with the same form of government can be at least as great as the differences across different forms of governments. This suggests that country-specific factors, rather than form of government, are predominant in determining the budgetary powers of the legislature.

The impact of the electoral and voting systems that give rise to two-party, coalition, or minority governments, the budget powers of a second chamber of the legislature, and political party decision-making processes are all important determinants of the relative strength of the executive and the legislature in determining the size and shape of annual budgets. In multi-party political systems, fragile governments have to rely on consensus between parties of the majority (for coalition governments) or of the opposition (for minority governments). In these countries, although strong budgetary powers for the legislature are provided in law, parliaments are constrained by voluntary party agreements. Legislatures do not use their budget amendment powers to the fullest extent. Each political party realizes that it is in their interest to avoid a downfall of the government because of a disagreement over the annual budget.

Constitutional and other formal legal instruments also constrain the budgetary powers of legislatures. Constitutions in some countries include specific limitations on the legislature’s powers to amend draft budgets. Laws with legally binding quantitative limits on deficits or debt have been adopted in some countries—with varying degrees of success. Other countries have preferred the inclusion of qualitative fiscal rules in budget systems laws.

This paper has not fully explored all country-specific factors that influence the relative budgetary powers of the legislature and the executive. In some cases, these are rooted in history, sometimes dating back over centuries. The “royal prerogative” in the United Kingdom is an example. The manner in which budgetary consensus is attained in Japan is partly due to cultural factors inherited over long time periods. As with political and legal factors, it could be expected that the weight of tradition varies across countries and over time.

In summary, although the budgetary powers of the legislature augment with increasing separation of the legislature and the executive, this relationship is neither linear nor stable over time. Country-specific factors—emanating from electoral systems, political party decision-making arrangements, the legal framework, and customs formed over centuries—appear to be more important than a particular form of government.


Indices on the Separation of Powers and the Control by the Legislature

This annex describes the overall index on the separation of powers, based on eleven indicators of the separation of political powers and the control of the executive by the legislature. Based on differing forms of government, quantitative values are assigned, as follows:

Index of the Separation of Powers and the Legislature’s Political Control

Description of variableQuantification
Part A. Separation of Powers
1. Is the head State directly in elections separate from elections for the legislature?1 = Yes

0 = No
2. Is the head of the executive (or government) elected independently?2 = Yes, citizens elect him/her directly

1 = No, he/she is appointed by the head of State, upon recommendation of the legislature

0 = No, he/she is automatically the head of government, but virtue of his/her leadership of the party with the most votes in the legislature
3. Is the head of State also a constituent part of the legislature?1 = No

0 = Yes
4. Can the head of government be drawn from outside the legislature?2 = Yes, citizens elect him/her directly

1 = Yes, by appointment of the head of State

0 = No, a he/she must necessarily be an elected member of the legislature
5. Is it impossible to hold simultaneously positions in both the legislature and in the government (executive)1 = Yes

0 = No
6. Can the head of State call elections of the legislature?1 = No

0 = Yes
7. Can the legislature vote “no confidence” in the government (executive)1 = No

0 = Yes
Part B. Limitations on Executive Power by the Legislature
8. Can the government (executive) determine the timing of sessions of the legislature?1 = No

0 = Yes
9. Can the head of State or government veto bills modified (and possibly approved) by the legislature?0 = Yes

1 = Yes, provided the veto is not rejected by a supermajority in the legislature

2 = No, not under any circumstances
10. Can the executive adopt secondary laws with the same force as primary laws?1 = No

0 = Yes, some secondary laws are not reviewed by the legislature
11. Do nearly all draft laws originate in the executive?1 = No

0 = Yes
12. Can the legislature block the appointment of political appointees that are nominated by the executive?1 = Yes

0 = No

On the basis of the numeric values associated with each question, the following results were obtained:

Annex Table 1.Indices of Separation of Powers and Control of the Executive by the Legislature
PresidentialSemi-presidentialParliamentary RepublicParliamentary monarchy: non WestminsterParliamentary monarchy: Westminster
A. Separation of the Legislature from the Executive
1. Election of head of State11000
2. Election of head of the executive21110
3. Head of State separate from legislature11110
4. Selection of head of government21110
5. Non-compatibility of legislature and executive710.50.50.50
6. Election call by head of State10000
7. “No confidence” vote10000
B. Control of the Executive by the Legislature
8. Timing of legislature sessions11110
9. Veto power of head of State or government11000
10. Adoption of secondary laws by executive11110
11. Origin of draft laws100.5800
12. Legislative power over political appointees10000
Overall Index of Separation of Political Powers

(total possible score = 14)

Index of Legislature’s Budgetary Powers

The legislature’s power over the budget are mainly at the ex ante stage – before the annual budget law is approved. Four questions are particularly pertinent (see below). During budget execution, the main question concerns whether or not the executive branch can modify the budget without returning to the legislature and seek its approval. An index was constructed on the basis of the following five questions:

Legislature’s Budgetary Powers

Description of variableQuantification
1. Establishment of the medium-term budget strategy.

Does the legislature approve each year, an updated budget strategy covering at least 3 years (including the new budget year)?
2 = Yes, the legislature adopts a law each year

1 = Yes, the legislature adopts a resolution each year

0 = No, at best, the legislature is only informed of the government’s medium-term strategy
2. Power to amend the annual budget.

Does the legislature have unlimited powers to amend the draft budget proposed by the executive? If there are any restrictions, how severe are these?
3 = no restrictions to change balance and composition

2 = budget deficit may only be increased if offsetting measures

1 = budget deficit may not be amended, but composition of expenditures may change

0 = neither budget balance nor composition may change except in very minor ways. Legislature must, in effect, accept or reject entire budget
3. Time allowed for discussion of annual budget.

How many months does the legislature receive the draft budget from the executive?
2 = More than four months

1 = From one to four months

0 = Less than one month or after the beginning of the fiscal year
4. Technical support in the legislature.

Does the legislature have a specialized budget advisory/research organization attached to provide budgetary analyses independent of the executive?
2 = Yes, with a substantial number of staff (more than 25)

1 = Yes, but with a limited number of staff

0 = No
5. Restrictions during budget execution.

Does the legislature oblige the government to implement its expenditure programs exactly as adopted? If not, what restrictions are there on the governments powers to modify the budget during implementation?
2 = Executive may neither withhold nor reallocate funds without legislative approval.

1 = Executive may withhold or reallocate funds, subject to certain limits.

0 = Executive has unlimited powers to limit expenditures in order to achieve overall budget objectives.
Table 2.Indices for the Legislature’s Budget Authority in 28 Countries
Type of government1. Medium-term framework2. Amendment powers3. Time for scrutiny of budget4. Technical support to legislature5. Restrictions during executionTotal index
BelgiumParlia. Monarchy031004
DenmarkParlia. Monarchy031015
GermanyParliam. Republic031004
IcelandParliam. Republic031026
ItalyParliam. Republic231017
JapanParlia. Monarchy031127
NetherlandsParlia. Monarchy031116
New ZealandWestminister000000
NorwayParlia. Monarchy031026
SpainParlia. Monarchy011013
SwedenParlia. Monarchy131129
TurkeyParlia. Monarchy011013
United KingdomWestminister000011
United StatesPresidential1322210
Sources: OECD (2002)9; Wehner (2005)
Sources: OECD (2002)9; Wehner (2005)

An earlier version of this paper was presented at the 14th annual conference of the Review “Policies and Public Management” (« Politiques et Management Public ») held at the University of Bordeaux, France, March 17–18, 2005. The author is grateful to Professor Hoffman-Martinot (University of Bordeaux), Professor Emeritus Montagnier (University of Lyon), Professor Wehner (London School of Economics), and Céline Allard, Ana Corbacho, Ketil Hviding, Bogdan Lissovolik, James Morsink, Brian Olden, Marc Robinson and Silvana Tendelli (all IMF) for comments on the earlier version. Nataly Sabharwal provided excellent research assistance. Any remaining errors are the author’s.

It is recognized that, within a particular form of government, important differences exist. In particular, the powers of the president vary considerably within “presidential”, “semipresidential” and “parliamentary republican” systems.

Wehner (2004) used a total of 15 criteria in an alternative index of the legislature’s budget authority. Perhaps the most important variable excluded from the parsimonious index used in this paper is that relating to supplementary budgets. In a few countries (e.g., Japan), a supplementary budget is a regular feature of the annual budget cycle— and supplementary budgets can be used to alter substantially the initial budget adopted by parliament.

In 2005, the Indonesian vice-President was not only from a different political party from the President, but he was also appointed as the leader of the country’s largest political party.

There is no lower limit on representation. However, at below 4 percent of all votes cast, a party does not get a chance to compete for an additional pool of seats, thus reducing their overall representation.

The legal amendment powers are based on column 2 of Annex Table 2. The actual amendments are based on question 2.7.i of OECD (2003).

The values of 0.5 reflect that fact that in some countries, it is not stated whether the functions are incompatible. Examples: Finland and Greece (semipresidential); Germany (parliamentary republic); Belgium, Denmark, Japan, Spain (non-Westminster monarchies). See Laundy, (1989), p. 32

The value of 0.25 reflects the fact that in a few parliamentary republics (e.g., Italy), most draft laws originate in the legislature, not the government (as in say, Germany). See Rose, 1984, Table 3.1.

OECD published the surveys as returned by member countries. Efforts to ensure quality control were not stringent. Moreover, in some cases, the survey results are debateable. As one example, Italy’s parliament’s “unlimited” budget amendment powers are in fact constrained by Article 81(4) of the Constitution.

Other Resources Citing This Publication