Tools for Legislative Oversight: An Empirical Investigation By Riccardo Pelizzo1 and Rick Stapenhurst2 Abstract Parliaments are the institutions through which governments are held accountable to the electorate. They have a wide range of tools with which to carry out this oversight function, but until recently little analysis had been undertaken on the characteristics or use of such tools. This paper uses data for 83 countries that was collected in 2001 to investigate whether the oversight potential relates to three variables, namely the form of government (presidential, semi-presidential, or parliamentary), per capita income levels, and the level of democracy. The paper finds that oversight potential is greatly affected by the form of government, per capita income levels, and levels of democracy. Countries with parliamentary forms of government, higher income levels, and which are more democratic have a greater number of oversight tools and greater oversight potential. While the oversight potential follows this general trend, the use of committees of enquiry, interpellations and ombudsman offices follows a different pattern. The use of interpellations as an oversight tool is most common in high income countries, less common in low income countries and least common in middle income countries while the presence of committees of enquiry and of the ombudsman offices is most common in middle income countries, less common in high income countries and least common in low income countries. World Bank Policy Research Working Paper 3388, September 2004 The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the view of the World Bank, its Executive Directors, or the countries they represent. Policy Research Working Papers are available online at http://econ.worldbank.org. 1Assistant Professor, Singapore Management University 2Senior Public Sector Management Specialist, World Bank Institute An earlier abridged version of this paper was presented at the Southern Political Science Association conference in New Orleans, January 2004. A modified version of that earlier draft can be found in Quaderni di Scienza Politica, vol. 11, n. 1, (gennaio) 2004, pp. 175-188. Contents 1. Introduction..................................................................................1 2. The Oversight Tools........................................................................3 3. The Distribution of Oversight Tools and the Potential for Oversight................6 4. Oversight Tools and the Forms of Government........................................ 7 5. Oversight Tools and Income Level......................................................13 6. Oversight Tools and the Level of Democracy....................................... 17 7. Conclusions.............................................................................. 18 Tables 1. The Tools of Parliamentary Oversight............................................. 8/9 2. How Common are these Oversight Tools?.......................................... 10 3. Countries and Number of Parliamentary Oversight Tools........................ 10 4. How Common are the Tools of Parliamentary Oversight by Form of Government............................................................ 11 5a Form of Government and Number of Parliamentary Oversight Tools.............12 5b Number of Oversight Tools by Form of Government............................... 12 6. Number of Oversight Tools by Income Level.........................................14 7. Committee Hearings by Income Level.................................................14 8. Hearings in Plenary Session by Income Level........................................15 9. Use of Questions by Income Level..................................................... 15 10. Use of Question Time by Income Level............................................... 15 11. Use of Interpellation by Income Level.................................................16 12. Committees of Enquiry by Income Level.......................................... ...16 13. Presence of an Ombudsman as Oversight Tool by Income Level............... 17 14. Number of Oversight Tools by Level of Democracy.............................. 18 Introduction In a recent issue of the Quaderni di Scienza Politica, Roberta Maffio published a comparative analysis of parliamentary oversight3in which she elaborates a conceptual mapping of the oversight tools, discusses the characteristics of oversight tools, and investigates whether the adoption of oversight tools by parliaments is related to the model of democracy (i.e. majoritarian, consensual, or mixed)4. The analysis between type of democracy and oversight potential was conducted in a sample of 24 countries. Twenty-one of these countries are the democracies investigated by Lijphart (1984),5 while the other three are Spain, Portugal, and Greece. Maffio divided the 24 countries into three groups depending on whether their government was majoritarian, consensual, or mixed. She then estimated the oversight potential for each country, using the 1986 data of the Inter-Parliamentary Union (IPU)6. Maffio concluded that there is no correlation between model of democracy and oversight potential. "There are some majoritarian democracies with strong oversight potential (such as Greece) and others with weak oversight potential (such as Ireland, New Zealand, United Kingdom)." Similarly, among consensual democracies, she found cases 3Roberta Maffio, "Quis custodiet ipsos custodes? Il controllo parlamentare dell' attivitá di governo in prospettiva comparata", Quaderni di Scienza Politica, vol. IX, n. 2, (Agosto) 2002, pp. 333-383. 4Roberta Maffio, "Quis custodiet ipsos custodes? Il controllo parlamentare dell' attivitá di governo in prospettiva comparata", op. cit., p. 359. 5Arend Lijphart, Democracies. Patterns of Majoritarian and Consensus Government in Twenty-One Countries, New Haven, Yale University Press, 1984. 6Inter-Parliamentary Union, Parliaments in the World: a Comparative Reference Compendium, Aldeshot, Gower, 1986. 1 characterized by both high oversight potential (such as Belgium, Germany, the Netherlands) and by low oversight potential (for example, Japan)7. The purpose of the present paper is to investigate whether the oversight potential of parliaments is related to other variables; however, we present two major differences from Maffio's paper. First, instead of using the 1986 IPU data, we use the data that the IPU collected in collaboration with the World Bank Institute (WBI) in 2001. Our analyses are performed on the 83 cases for which the IPU-WBI data are available8. And second, in contrast to Maffio, we investigate whether there is any relationship between oversight potential and types of constitution (i.e. presidential, semi-presidential, or parliamentary), national income levels, and levels of democracy. In the first section of this paper, after providing a working definition of oversight, we review some of the tools that parliaments can use to oversee the activities of the executive. We underline that oversight tools can be grouped on the basis of two criteria : the first concerns whether a specific oversight tool is employed before (ex ante) or after (ex post) the enactment of a government-sponsored policy, and the second concerns whether oversight is exercised internally or externally to parliament. Though the data collected by the IPU in collaboration with the WBI do not provide evidence as to the effectiveness of the oversight tools, they do provide some measure regarding the potential for oversight. Our line of reasoning is straightforward, as we suggest that a parliament's potential to oversee the executive increases as the number 7Roberta Maffio, "Quis custodiet ipsos custodes? Il controllo parlamentare dell' attivitá di governo in prospettiva comparata", op. cit., p. 361. 8See Table 1 in this paper. 2 of oversight tools available to that parliament increases. The second section of this paper investigates the adoption of oversight tools. The subsequent three sections investigate how the adoption of oversight tools and, hence the potential for oversight, are related to and, possibly, affected by other variables. The third, the fourth and the fifth sections analyze the relationship between the number of oversight tools and the type of constitution, the relationship between the number of oversight tools available and the countries' income level, and, finally, the relationship between the number of oversight tools available and the level of democracy. In the sixth and final section of the paper we draw some conclusions as to the relevance of our findings. The Oversight Tools Before we discuss what the various tools that parliaments can use to oversee the executive branch are, we need to define what is meant by `legislative oversight'. In a Research Paper of the National Democratic Institute, legislative oversight was defined "the obvious follow-on activity linked to lawmaking. After participating in law-making, the legislature's main role is to see whether laws are effectively implemented and whether, in fact, they address and correct the problems as intended by their drafters"9. This definition captures the role that parliaments play in overseeing government policies and activities after they have been enacted but it overlooks that parliaments may be engaged in oversight activities well before a policy is enacted. Parliaments oversee the 9National Democratic Institute, "Strengthening Legislative Capacity in Legislative-Executive Relations", Legislative Research Series, Paper # 6, Washington, DC, 2000, p. 19. 3 executive not only with regard to the execution and implementation of policies but also with regard to the preparation of policies. This is why, paraphrasing Lees' definition of legislative oversight, we refer to legislative oversight of the executive as to "the behavior of legislators and their staff which affects executive behavior"10. In overseeing the executive, Parliaments have several different oversight tools at their disposal. The most common oversight tools are committee hearings, hearing in plenary sessions of the parliament, the creation of commissions of inquiry, questions, question time, interpellations, the ombudsman, auditors general, and the public account committees11. These oversight tools can be grouped along two dimensions. The first dimension pertains to the timing of the oversight activity. If legislative oversight is performed before the government enacts a specific policy or becomes engaged in a specific activity, then the oversight tools are "instruments of control ex ante"12. Hearings in committees, hearings in the plenary sessions of the Parliament, along with the request of documentation are all tools that can be used ex ante. If the legislative oversight is performed after the government has enacted a policy to check whether the policy is properly implemented, then the oversight tools are instruments of control ex post. Questions, interpellations, the creation of committees of inquiry are the tools that are 10Lees defined oversight as "the behavior by legislators and their staffs, individually or collectively, which results in an impact, intended or not, on bureaucratic behavior", see John D. Lees, "Legislatures and Oversight: A Review Article on a Neglected Area of Research", Legislative Studies Quarterly, vol. 2, n. 2, (May) 1977, pp. 193-208. 11A description of some of these tools can be found in National Democratic Institute, "Strengthening Legislative Capacity in Legislative-Executive Relations", Legislative Research Series, Paper # 6, Washington, DC, 2000, especially pp. 19-32. 4 used ex post. The second dimension pertains whether the oversight tools are established inside or out side the Parliament, that is whether they are internal or external oversight tools. Questions, question time, interpellations, hearings, public account committees are internal tools, while ombudsmen and auditors general are external tools. The IPU, in collaboration with WBI, conducted a survey on Executive-Legislative relations; some 180 Parliaments were surveyed, and 83 responded (82 national parliaments plus the European Parliament). Respondents were asked questions concerning the accountability of the government to the parliament, the impeachment procedure, the dissolution of parliament, the oversight of government, budgetary oversight, oversight of implementation of the budget, oversight over foreign policy, oversight over national defense policy, the parliament and the state of emergency, the verification of the constitutionality of laws, and, finally, oversight over the application and evaluation of laws. In this note we will focus only on the oversight of the government. Respondents were asked several questions: whether the government in their country was considered an institution that must report to Parliament, how does the parliament exercise oversight, whether parliamentarians could question government officials, whether time for questions was allocated in Parliament, whether interpellations were foreseen and, finally, whether there was an ombudsman in the country13. 12Roberta Maffio, "Quis Custodiet Ispos Custodes? Il Controllo Parlamentare dell'Attivitá di Governo in Prospettiva Comparata", Quaderni di Scienza Politica, ..., pp.333-383. The quote is taken from p. 348. 13A discussion for why questions, question time, interpellations should be considered as instruments of parliamentary control can be found in Roberta Maffio, "Quis custodiet ipsos custodes? Il controllo parlamentare dell' attivitá di governo in prospettiva comparata", op. cit., pp. 333-383. See also David G. McGee, The Overseers. Public Accounts Committees and Public Spending, London, Commonwealth Parliamentary Association in Association with Pluto Press, 2002. 5 These data are relevant in two respects. First of all, they indicate parliaments' potential for effective oversight by showing which institutional arrangements have been adopted to enhance oversight. And second, they can be used to investigate whether the distribution of oversight tools is related to other variables such as a country's form of government, the national income level and a country's level of democracy as measured by the Gastil index. The Distribution of Oversight Tools and the Potential for Oversight The data presented in Table 1 provide interesting information in at least three respects. First of all, these data indicate that there is considerable variation in how common are these various tools of parliamentary oversight. For example, parliamentarians can put oral or written questions to the government in 79 of the 82 countries for which data are available or in 96.3 percent of the cases. Committees of inquiry and committee hearings are also common instruments of parliamentary control, utilized in more than 95 percent of the countries for which data are available. By contrast, interpellations and the ombudsman are substantially less common, with interpellations to the government used in about 75 percent of the countries while an ombuds office is instituted in less than 73 percent of the countries. Data are presented in Table 2. Second, the analysis of the data indicates that all countries adopt some parliamentary oversight tools and that most of them use more than one tool. Complete information is 6 available for only 49 of the 83 countries that were surveyed; in the remaining 34 countries information was provided incompletely or was not provided at all (as in the case of Lesotho). In any case, more than 12 percent of the countries for which complete information is available use 4 tools of parliamentary oversight, more than 14 percent of them used 5 tools, almost 33 percent of them used 6 tools while the remaining 40 percent used 7. Data are presented in Table 3. Oversight Tools and the Forms of Government In addition to providing information concerning the distribution of oversight tools and the potential for parliamentary oversight, the IPU-WBI data shed some light on how tools and types of oversight are related to another institutional feature, that is the form of government14. Here again, an analysis of the data is informative. First of all, the analysis makes clear that the number of oversight tools available varies from one constitutional form of government to another. As shown in Table 4, tools such as committee hearings, 14Respondents were asked to indicate what form of government was in their countries. Responses were coded : Presidential form of government was given value 1, Parliamentary form of government was given value 2, Semi-Presidential form of government was given 3, Parliamentary form of government in countries with a Constitutional Monarchy was given 4, Parliamentary form of government in countries with a Hereditary Monarchy received 5, and other forms of government were all given value 6. In our analysis, we adopted a different coding scheme. Countries that were given value 2, 4 or 5 by the IPU dataset were all considered to have a parliamentrary system and were hence collapsed into a single category. 7 Table 1. The Tools of Parliamentary Oversight Country Committee Hearing in Commissi Questions Question Interpellati Ombudsman Hearing plenary on of time ons sitting enquiry Andorra yes yes Yes yes yes n.a. yes Angola yes no Yes yes no yes no Armenia yes yes No yes yes no no Austria yes yes Yes yes yes yes yes Australia yes no Yes yes yes no yes Azerbaijan n.a. n.a. n.a. no n.a. yes yes Belarus yes yes Yes yes yes yes n.a. Belgium yes yes Yes yes yes yes yes Benin yes yes Yes yes yes yes no Brazil yes yes Yes yes no yes yes Bulgaria yes yes Yes yes yes n.a. no Cameroon yes yes Yes yes yes no no Canada yes yes Yes yes yes yes no China yes yes Yes yes n.a. no n.a. Congo n.a. n.a. n.a. yes no yes yes Costa Rica yes yes Yes yes yes yes yes Croatia yes yes Yes yes yes yes yes Czech yes yes Yes yes yes yes yes Republic Cyprus yes yes Yes yes yes no yes Estonia yes yes Yes yes yes yes yes EU yes yes Yes yes n.a. yes yes France yes yes Yes yes yes yes yes Gabon yes yes Yes yes yes yes yes Germany yes yes Yes yes no yes yes Greece yes yes Yes yes yes yes yes Guatemala yes yes Yes yes yes yes n.a. Guinea yes yes Yes yes yes yes no Bissau Guinea yes yes Yes yes yes yes no Hungary yes yes Yes yes yes yes yes Iceland yes yes Yes yes yes n.a. yes Indonesia yes yes Yes yes yes yes yes Iran yes yes Yes yes no yes n.a. Ireland yes yes Yes yes yes n.a. yes Ivory Coast no no Yes yes yes no yes Kazakhstan no yes Yes yes yes no no Korea yes yes Yes yes yes no yes Jamaica yes yes Yes yes yes no yes Japan yes yes Yes yes yes yes yes Jordan yes yes Yes yes no n.a. yes Latvia yes yes Yes yes n.a. yes yes Lesotho n.a. n.a. n.a. n.a. n.a. n.a. n.a. Liechtenstein yes yes No yes no yes no Lithuania yes yes Yes yes yes yes yes Luxembourg yes yes Yes yes yes yes no Macedonia n.a. n.a. n.a. yes no yes yes 8 Country Committee Hearing in Commissi Questions Question Interpellati Ombudsman Hearing plenary on of time ons sitting enquiry Madagascar yes yes Yes yes yes yes yes Mali yes yes Yes yes yes yes yes Mexico yes yes Yes no yes n.a. yes Mongolia yes yes Yes yes yes n.a. no Namibia yes yes Yes yes yes n.a. yes Netherlands yes yes Yes yes yes n.a. yes Nicaragua yes yes Yes no no yes yes Niger yes yes Yes yes yes yes no Palau yes no Yes yes no yes yes Philippines n.a. yes Yes yes yes n.a. yes Poland yes yes Yes yes yes n.a. yes Romania yes yes Yes yes yes yes yes Russia n.a. n.a. n.a. yes n.a. n.a. yes Rwanda no no No yes yes yes yes Samoa n.a. yes Yes yes yes no yes Senegal yes yes Yes yes no no yes Singapore yes yes Yes yes yes n.a. no Spain yes yes Yes yes yes yes yes Slovak yes yes n.a. yes yes yes yes Republic Slovenia yes yes n.a. yes yes yes yes South Africa yes yes n.a. yes yes no yes Sudan yes yes Yes yes yes n.a. no Sweden yes yes Yes yes yes yes yes Switzerland yes yes Yes yes yes yes no Tajikistan n.a. n.a. Yes yes n.a. yes n.a. Tchad yes no Yes yes yes yes yes Thailand yes yes Yes yes yes n.a. yes Togo yes yes Yes yes yes yes no Tunisia yes yes Yes yes yes no yes Turkey yes yes Yes yes yes no no Uganda yes yes Yes yes yes n.a. yes Ukraine n.a. yes n.a. yes yes yes yes United yes yes Yes yes yes no yes Kingdom Uruguay yes no Yes yes n.a. yes no Zambia yes yes Yes yes no yes yes Zimbabwe n.a. n.a. n.a. yes yes n.a. yes Yemen yes yes Yes yes yes no no Yugoslavia yes yes Yes yes yes yes no 9 Table 2. How Common are these Oversight Tools? Committee Hearing in Committee Questions Question Interpellations Ombudsman hearing Plenary of Inquiry Time Sitting % of N % of N % of N % of N % of N % of N % of N 95.9 73 90.8 76 95.9 73 96.3 82 84 75 75.4 65 72.7 77 Table 3. Countries and the Number of Parliamentary Oversight Tools ______________________________________________________________________________________ 0 2 3 4 5 6 7 Lesotho Azerbaijan Congo Angola Australia Andorra Austria Russia Macedonia Armenia Bulgaria Belarus Belgium Tajikistan China Cameroon Benin Costa Rica Zimbabwe Ivory Coast Iran Brazil Croatia Kazakhstan Jordan Canada Czech Republic Liechtenstein Mexico Cyprus Estonia Rwanda Mongolia EU France Uruguay Nicaragua Germany Gabon Palau Guatemala Greece Philippines Guinea Bissau Hungary Samoa Guinea Indonesia Senegal Iceland Japan Singapore Ireland Lithuania South Africa Korea Madagascar Sudan Jamaica Mali Turkey Latvia Romania Ukraine Luxembourg Spain Yemen Namibia Sweden Netherlands Switzerland Niger Zambia Poland Slovakia Slovenia Tchad Thailand Togo Tunisia Uganda U.K. Note: The score of the countries in Italics indicates that information concerning the presence/absence of some tools of parliamentary information was not available as indicated in Table 1. 10 hearings in plenary sittings, question time and interpellations are generally more common in parliamentary forms of government, than in presidential and semi-presidential systems. As a result, legislatures in parliamentary systems have more oversight tools and oversight potential at their disposal than legislatures in either presidential or Table 4. How Common are the Tools of Parliamentary Oversight by Form of Government Form of Committee Hearings in Committee Questions Question Interpellations Ombudsman government hearings Plenary of Enquiry Time Sittings % of N % of N % of N % of N % of N % of N % of N Parliamentary 100 34 97.1 35 96.9 32 100 32 88.6 35 76.9 26 77.8 36 Presidential 88.2 17 83.3 18 100 19 85.7 21 78.9 19 72.2 18 77.8 18 Semi- 93.3 15 81.3 16 86.7 15 100 17 86.7 15 75 16 52.9 17 Presidential semi-presidential systems. See Tables 5a and b. As noted above, these results do not allow us to make any inference as to the effectiveness of the oversight tools, that is as to whether legislatures in parliamentary systems are more effective than legislatures in either presidential or semi-presidential systems in overseeing the executive branch of the government ­ only that they have more oversight tools at their disposal. 11 Table 5a. Form of Government and Number of Parliamentary Oversight Tools Form of 4 5 6 7 Government Ivory Coast Nicaragua Benin Costa Rica Kazakhstan Palau Brazil Indonesia Cyprus Presidential Guinea Korea Tchad Tunisia Liechtenstein Australia Canada Austria Turkey Germany Belgium Guinea Bissau Croatia Jamaica Czech Republic Luxembourg Estonia Parliamentary United Kingdom Greece Hungary Japan Lithuania Spain Sweden Angola Cameroon Niger France Armenia Senegal Togo Gabon Semi-Presidential Rwanda Yemen Yugoslavia Madagascar Mali Romania Other Swizterland Note: Zambia was not included in this table as it did not provide any answer as to what is its form of government. Table 5b. Number of Oversight Tools by Form of Government Number of Oversight Tools Form of Government 4 5 6 7 Tot Mean Presidential 2 2 7 2 13 5.69 Semi-Presidential 3 3 3 5 14 5.71 Parliamentary 1 2 6 11 20 6.35 Total 6 7 16 18 47 12 Oversight Tools and Income Level In the 2002 World Development Indicators published by the World Bank countries are divided in three groups : high-income economies, in which the gross national income (GNI) per capita was $ 9,266 or more; middle-income economies, which have a GNI per capita of between $ 755 to 9,265; and low-income economies, in which the GNI per capita is below $ 755. We transformed this information into a quantitative variable, by assigning value 1 to countries in the low-income group, value 2 to countries in the middle-income group and value 3 to countries that belong to the high-income group. Having created this variable, we then investigate whether there is a relationship (and if so, what type of relationship) between income levels and oversight. By cross- tabulating our Income variable with the number of oversight tools that are available to a country's parliament, we find that there is a clear, and strong linear relationship between the income level and the number of oversight tools. The number of oversight tools in the countries that have provided information in this respect varies from a minimum of 4 to a maximum of 7. Parliaments in low-income countries have an average of 5.5 oversight tools at their disposal, parliaments in middle-income countries have an average of 6.25 oversight tools while parliaments in high-income countries have an average of 6.27 oversight tools. Further details are presented in Table 6. 13 Table 6. Number of Oversight Tools by Income Level Number of Oversight Tools Income Level 4 5 6 7 Total Mean Low 4 4 7 3 18 5.50 Middle 1 2 5 8 16 6.25 High 1 1 6 7 15 6.27 Total 49 Examining the distribution of each of the oversight tools by income level, the data presented in Table 7 suggest that committee hearings become an increasingly popular oversight tools as we move from low- to middle- and high-income countries. Such hearings are used in less than 91% of the low-income economies for which survey data are available, in almost 97% of the middle-income countries and in all the high-income countries. Table 7. Committee Hearings by Income Level Do countries have committee hearings? Income level No Yes Total % yes Low 2 19 21 90.5 Middle 1 29 30 96.7 High 21 21 100 Total 3 69 72 Hearings in plenary settings are not as common as committee hearings, as the data presented in Table 8 illustrated, but even these become an increasingly more common oversight tools as we move from low- to middle- and high-income countries : hearings in plenary settings are used in only about 82% of the low-income countries, in about 94% of the middle-income countries and in more than 95% of the high-income countries. 14 Table 8. Hearings in Plenary Settings by Income Level Do countries have hearings in plenary settings? Income level no Yes Total % yes Low 4 18 22 81.8 Middle 2 30 32 93.8 High 1 20 21 95.2 Total 7 68 75 The use of questions and question time as oversight tools follows the same pattern that we have observed with regard to both committee hearings and hearings in plenary settings. They all become increasingly more common oversight tools as we move from lower- to higher- income level. Questions are used in 92.3% of the low-income countries, 97% of the middle-income countries and in all the high-income countries. Similarly, question time is used instead in about 79% of the low-income countries, 83 % of the middle-income countries and in a little more than 90% of the high-income countries. Data concerning the use of questions and question Time as oversight tools are presented respectively in Table 9 and 10. Table 9. Use of Question by Income Level Do countries use questions as oversight tool? Income level no Yes Total % yes Low 2 24 26 92.3 Middle 1 33 34 97.1 High 21 21 100 Total 3 78 81 Table 10. Use of Question Time by Income Level Do countries Question Time as oversight tool? Income level no Yes Total % yes Low 5 19 21 79.2 Middle 5 25 30 83.3 High 2 19 21 90.5 Total 12 63 75 15 The use committees of enquiry, of interpellations and of the ombudsman offices follows a different pattern. The survey data suggest that the use of interpellations as an oversight tool is very common in high-income countries, less common in low-income countries and least common in middle-income countries : they are used in about 81 % of the high-income cases, in about 77% of the low-income cases, and in less than 70% in the middle-income cases. Data are presented in Table 11. Table 11. Use of Interpellation by Income Level Do the countries in these income level use interpellations? Income level no Yes Total % yes Low 5 17 22 77.3 Middle 8 18 26 69.2 High 3 13 16 81.3 Total 16 48 64 Finally, the presence of committees of enquiry and of Ombuds offices as oversight tools is most common in middle-income countries, less common in high- income countries and least common in low-income countries. Data concerning the presence of committees of enquiry and of the Ombuds offices by are presented in Table 12 and 13. Table 12. Committees of Enquiry by Income Level Do the countries in these income levels have Committees of Enquiry ? Income level no Yes Total % yes Low 2 20 22 90.9 Middle 30 30 100 High 1 19 20 95 Total 3 69 72 16 Table 13. Presence of Ombusman as oversight tool by Income Level Do countries have an Ombudsman? Income level no yes Total % yes Low 11 14 25 56 Middle 5 25 30 83.3 High 5 16 21 76.2 Total 21 55 76 Oversight Tools and Level of Democracy The purpose of this section is to investigate the relationship between the level of democracy and the number of oversight tools that are available to the legislature. Before we can do so, however, we need to discuss how our `democracy' variable is operationalized. The Freedom House computes an annual index of freedom for all the countries in the world; this index is regarded by many social scientists as a proxy index of democracy; known as Gastil index, it is computed in the following way. The Freedom House assigns to each country a political rights score and a civil liberties score. Both scores are 7-point scales. The index of freedom is estimated by adding a country's political rights score to that country's civil liberties score and by dividing their sum by 2. This means that the Gastil index of freedom is also a seven point scale. So, for example, if the fictional country of Abbaba has a score of 3 for its political rights and of 4 for its civil liberties, then Abbaba's democratic score is (3+4)/2 that is 3.5. Countries that score from 1 to 2.5 points on this scale are considered democratic; countries scoring from 3 to 5.5 are quasi- 17 democratic, while countries with a score of 5.5 or higher belong to the group of non- democratic countries. Our `democracy' variable was constructed by re-coding the Gastil index of freedom by assigning value 1 to democratic countries, value 2 to quasi-democratic countries and value 3 to non-democratic countries. Having created this variable, we then investigated whether there is a relationship (and if so what type of relationship) between the level of democracy in a given country and the number of oversight tools available to that country's parliament. By cross-tabulating our level of democracy variable with the number of oversight tools that are available to a country's parliament, we find that there is a clear, and strong linear relationship between the level of democracy and the number of oversight tools. Non-democratic countries have on average 5 oversight tools, quasi- democratic countries have an average of 5.71 tools, while democratic countries have an average of 6.41 oversight tools. Data are presented in Table 14. Table 14. Number of Oversight Tools by Level of Democracy Number of Oversight Tools Level of Democracy 4 5 6 7 Total Mean Democracy 1 2 9 15 27 6.41 Quasi Democracy 2 3 6 3 14 5.71 Non Democracy 3 2 3 8 5.0 Total 6 7 18 18 49 Conclusions The major question that we addressed in this paper is whether the potential for parliamentary oversight, as reflected by the number of oversight tools available in a given 18 country, is related to/ affected by other variables. Specifically, we investigated whether there is a relationship between the number of oversight tools and form of government, national per capita income levels, and level of democracy. We found that the oversight potential is affected by these variables. Beginning with the relationship between oversight potential and form of government, we found that legislatures in parliamentary systems are better equipped ­ in terms of oversight tools - than legislatures in either presidential or semi-presidential systems, while we did not find a parliament's oversight potential is particularly affected by whether there is a presidential or semi-presidential form of government. Legislatures in parliamentary systems have an average of 6.35 oversight tools, in contrast to the average of 5.69 recorded in presidential systems and the average of 5.71 recorded in semi-presidential systems. Our second set of findings concerns the relationship between the oversight potential and national per capita income levels. In this respect we found that the oversight potential is greatly affected by the income level. In fact, parliaments in low-income countries have on average a much smaller number of oversight tools at their disposal than their counterparts in both middle- and high- income countries. Interestingly, the oversight potential is not particularly affected by whether the country is middle- or high- income. Our third set of findings concerns the relationship between the oversight potential and the level of democracy. In this regard we found that the average number of oversight tools is almost linearly related to the level of democracy, so that the more democratic a 19 country, the more oversight tools are at the disposal of that country's parliament. Non- democratic countries have an average of just 5 oversight tools, quasi-democratic countries have an average of 5.71 oversight tools, and democratic countries have an average of 6.41 oversight tools. What does this mean? Does this mean that the adoption of additional oversight tools makes countries more democratic or does it mean that because countries are already democratic that they adopt additional oversight tools? In other words, is it possible to detect a clear arrow of causality? The data at our disposal and the analyses that can be performed with these data do not allow us to formulate a positive answer for this question. A uni-directional, causal influence cannot be detected. In addition to these methodological reasons, there are also some theoretical reasons why it may not be possible to answer the question above. It is well known, in the study of social sciences, that the relationship between variables instead of being uni- directional is often bi-directional. This means that one variable (Y) is determined by another variable (X) which, in turn, is determined by (Y). This could very well be the case of the relationship between level of democracy and oversight potential. If what distinguishes democratic regimes from non-democratic ones is that they entail representation, accountability and responsiveness, and if oversight tools are the institutional instruments that contribute to keeping governments accountable, then it is not surprising that democratic countries may want to adopt oversight tools. Yet as a country's oversight potential increases, so does the level of democracy, thus providing a virtuous circle. 20 Having established that the relationship between the level of democracy and the oversight potential is possibly bi-directional, let us focus on one of the two aspects of the relationship, that is on the fact that as oversight potential increases, so does the level of democracy. This finding has important practical consequences for the international community and those international organizations seeking to reduce global poverty and promote good governance. If democracy is "a condition without which development and poverty reduction strategies could not be properly implemented"15, and if oversight tools are the institutional devices that are associated with the transition of countries to full democracies, then it becomes of considerable importance for the international community to better understand the dynamics of parliamentary oversight, the role that oversight tools can play in the promotion of government accountability (and thus, democratic development and good governance) and to identify and disseminate examples of good practice practices. By doing so, the international community contributes to creating the conditions for sustainable long-term development. This paper contributes to such an understanding by explaining the incidence of parliamentary oversight tools and relating this to important economic and political variables. Further research needs to be undertaken on the effectiveness such tools ­ but given the difficulties of measuring parliamentary performance across countries16 such research should perhaps first be through the development of country case studies, attempting to measure changes in effectiveness over time. 15The quote is taken from Riccardo Pelizzo and Frederick Stapenhurst, "Legislatures and Oversight: A Note", Paper prepared for delivery at the Annual Meeting of the Southern Political Science Association, New Orleans, Louisiana, 7-10 January, 2004. 16Laurentian Seminar Proceedings, Parliamentary Centre, 1997 21 REFERENCES Stephen Frantzich, "Computerized Information Technology in the U.S. House of Representatives", Legislative Studies Quarterly, vol. 4, 1979, pp.255-ff. Keith E. Hamm and Roby D. Robertson, "Factors Influencing the Adoption of New Methods of Legislative Oversight in the US States", Legislative Studies Quarterly, vol. 6, 1981, pp. 233-ff. Inter-Parliamentary Union, Parliaments in the World: a Comparative Reference Compendium, Aldeshot, Gower, 1986. John R. Johannes, "Casework as a Technique of U.S. Congressional Oversight of the Executive", Legislative Studies Quarterly, vol. 4, 1979, pp. 325-ff. Laurentian Seminar Proceedings, Parliamentary Centre, 1997 John D. Lees, "Legislatures and Oversight: A Review Article on a Neglected Area of Research", Legislative Studies Quarterly, vol. 2, n. 2, (May) 1977, pp. 193-208. Roberta Maffio, "Quis custodiet ipsos custodes? Il controllo parlamentare dell' attivitá di governo in prospettiva comparata", Quaderni di Scienza Politica, vol. IX, n. 2, (Agosto) 2002, pp. 333-383. David G. McGee, The Overseers. Public Accounts Committees and Public Spending, London, Commonwealth Parliamentary Association in Association with Pluto Press, 2002. National Democratic Institute, "Strengthening Legislative Capacity in Legislative- Executive Relations", Legislative Research Series, Paper # 6, Washington, DC, 2000, p. 19. David Olson, Riccardo Pelizzo and Rick Stapenhurst (eds.), Trends in Parliamentary Oversight: Proceedings from a Panel at the 2004 Southern Political Science Association Conference (forthcoming). Riccardo Pelizzo and Frederick Stapenhurst, "Legislatures and Oversight: A Note", Paper prepared for delivery at the Annual Meeting of the Southern Political Science Association, New Orleans, Louisiana, 7-10 January, 2004, now avalilable in David Olson, Riccardo Pelizzo and Rick Stapenhurst (eds.), Trends in Parliamentary Oversight: Proceedings from a Panel at the 2004 Southern Political Science Association Conference (forthcoming). 22