Paraguay User s Manual to The Government Expenditure Database U The World Bank April 2013
Content 1. Introduction... 3 2. Paraguayan Budget Structure... 4 2.1 Particularities of the Data... 5 3. Database Organization... 6 4. How to use the Paraguay BOOST database... 10 The World Bank
1. Introduction A core function of any government involves collecting and disbursing public funds and maintaining records of such expenditures. Budget execution reporting systems vary greatly in their organization, transparency, accuracy and level of detail. For researchers, the ability to easily access and analyze public expenditure data is essential to providing thorough and timely advice to policymakers and relevant stakeholders. For government officials, rapid access to budget execution data in an easyto-use format is essential for better decision-making and increasing analytical capacity within ministries. The purpose of the BOOST project (started by the World Bank in early 2010) is to enhance the transparency and efficiency of public spending across the globe by improving access to government expenditure data and linking spending to relevant results. The BOOST team has collected and processed detailed government expenditure data from more than twenty countries across diffetenr regions, creating easy-to-use databases that have been employed by World Bank researchers on a variety of projects. In Paraguay, the BOOST government expenditure database has been constructed at the intitiative of World Bank country economist as part of the Public Expenditure Review. This manual describes the resulting database and provides guidance on its use. For help with issues that are not addressed here, please contact the BOOST team. This document is organized into the following section Part II outlines the structure of the data provided by the Paraguayan Government. Part III presents the organization of the database. Part IV explains how to use the database with an Excel pivot table. The BOOST team would like to thank staff at the Ministry of Finance, as well as the growing BOOST user community of researchers and policymakers who continues to provide useful feedback as this project strives to grow and improve. We hope that this tool is helpful in opening new avenues for analysis and providing answers to important questions regarding the efficiency, equity, and effectiveness of government spending. The core BOOST team for Paraguay consists of Friederike Fritzi Koehler-Geib (fkoehlera@worldbank.org), Massimo Mastruzzi (mmastruzzi@worldbank.org) and Francisco Vazquez-Ahued (fvazquezahued@worldbank.org)
2. Paraguayan Budget Structure Paraguay s national public budget can be divided into three main parts, as shown in Figure 1 below: (i) the Central Administration Ministries and Agencies; (ii) the Decentralized Agencies (tha includes the Regional/Departmental Governments); (iii) and the Municipal Governments. Figure 1: Structure of the Paraguayan National Budget Paraguayan National Budget Central Administration Decentralized Agencies Municipal Governments Figure 2: Structure of the Central Administration Ministries and Agencies Central Administration Ministries and Agencies National Legislature Executive Branch (including Treasury) Judicial Branch Comptroller Ombudsman The Paraguay BOOST originally only had information from the Country s IFMIS (SIAF in Spanish), this information is sometimes referred to as central government data which basically includes the Central Administration Ministries and Agencies and the Decentralized Agencies. Figure 2 and Figure 3 show a further breakdown of these two respective categories. This data was obtained with the help of the Vice-ministry of Economy of the Ministry of Finance of Paraguay.
Figure 3: Structure of the Decentralized Agencies Decentralized Agencies Central Bank Departmental Governments Autonomous and Autarchic Entities Social Security Public Entities State Owned Enterprises National Universities For the central government database, besides having an administrative classification, there are functional, economic, programmatic classifications. This data is at the most disaggregated level that was available. For the World Bank s Non-lending Technical assistance Paraguay project on Improving Quality of Expenditures at the Municipal Level a database was developed that included expenditure data for municipal governments. From this project we obtained the Municipal Government data. However, this data doesn t have as much detail (there are no programmatic or functional classifications) and only data for the years 2006-2010 was obtained, whereas we have central government data for the years 2003-2012. 2.1 Particularities of the Data The Paraguayan central government data presents the following particularities: The inclusion of a classification variable, Tipo Presupuesto, which is one of the main classification variables of the Paraguayan budget. It divides the Paraguayan budget into four main categories: Administration, Action, Investment, and Debt Service. There are two programmatic classifications: Programa and Sub-programa. There is another similar classification variable labeled Proyecto, however there is no clear and consistent relationship with the programmatic variables, therefore it is considered to be a different kind of variable. The budget classification in Paraguay has an economic classification with three variables for each label: Econ1, Econ2, and Econ3. There is also another set of variables that could also be classified as an economic or quai-economic classification; these are the Objeto del Gasto variables, which include Grupo, Sub-grupo and Objeto Gasto. There is a variable to identify the intra-governmental transfers. There are also variables that identify what government unit is the recipient and which one is the contributor of the transfer; there is also a variable that identifies the type of transfer. The inclusion of the transfer variables is important since some official budget reports show figues data including transfers and some show figures excluding transfers. All of these transfer variables were created for the BOOST database. There is a geographic variable, Departamento, which identifies the department/region where the specific budget item is implemented. For items where it includes two or more departments it is considered to be countrywide (Alcance Nacional in Spanish). Since there are some changes in the administration structure of the Paraguayan government a variable was created to reflect these changes. For instance the Treasury at the beginning of the period was considered a different agency, but later it is considered to be part of the
Ministry of Finance. A variable was created, admin2 Reciente (Entidad), to reclassify the government agencies and entities into their 2012 equivalent. This is done only for the admin2. This variable was created for the BOOSt database. The Municipal Government data is currently not included in Paraguay s IFMIS. The municipal governments are legally mandated to send written reports on municipal finances for every budget cycle to the Ministry of Finance; however the process of collecting and processing this municipal data is still no automatic. Another issue is the late submissions from some municipal governments. As mentioned above, the World Bank project on Improving Quality of Expenditures at the Municipal Level for Paraguay collected information on municipal finances. With the help of Accounting Directorate of the Vice-ministry of Financial Administration of the Ministry of Finance and the Departments and Municipalities Unit of the same Ministry the World Bank team from that project was able to collect expenditure data for municipalities. There is much less detail than the central government data, there is practically only the variable to identify the municipality and two of the Objeto del Gasto variables: Grupo and Sub-grupo (the more detailed Objeto Gasto is not included). Therefore the potential analysis with the municipal expenditure data is much more limited than with the central government data. 3. Database Organization Currently the database contains information Paraguay BOOST of public spending at the most dissagregated level for central government agencies and decentralized entities for the years 2003-2012; while the data for espenditure by municipal governments is limited to the years 2006-2010. The BOOST database contains variables that correspond to fields in the treasury data that cover different types and levels of Paraguay s budget classification. A complete list of BOOST variables and descriptions can be found in Table 1 below Table 1. List of BOOST Variables BOOST Variable Variable Description Tipo presupuesto Main classification type in Paraguay's budget. Divided into four types: Administration, Action, Investment, and Debt Service Administrative and Programmatic Classification Variables admin1 (Nivel) admin2 (Entidad) admin3 (Unidad Responsable) Programa Sub-programa Level 1: Level of Government Spending (e.g. Judicial Branch, Regional Governments, National Universities) Level 2: Ministry, Agency, Sub-national government, or entity Level 3: Vice-Ministry or equivalent sub-entity Program Sub-program Functional Classification Variables
Finalidad Función Sub-función Level 1: Top-level classification Level 2: Mid-level classification Level 3: Bottom-level classification Economic Classification Variables Econ1 Econ2 Econ3 Level 1: Top-level classification Level 2: Mid-level classification Level 3: Bottom-level classification Classification Variables by "Objeto del gasto" Grupo Sub-grupo Objeto Gasto Level 1: Top-level classification Level 2: Mid-level classification Level 3: Bottom-level classification Source of Funding Variables Fuente de financiamiento Origen de financiamiento Level 1: Level 2: Transferencia Consolidable Tipo de Transferencia Consolidable Entidad Madre (transferencia) Entidad Hija/Receptora (transferencia) Other Classification and Created Variables Filter variable used to identify the transfers within the government Variable that identifies the type of transfer Variable that identifies the parent entity, i.e. the entity from which the transfer originates. Variable that identifies the recipient entity of the transfer Other variables Año Proyecto Departamento admin2 Reciente (Entidad) Fiscal Year Project The department where the expenditure is implemented Admin variable that puts admin2 categories into their 2012 equivalent Budgetary Amounts Presupuesto inicial Initial Budget (Paraguayan Guaranies) Modificación en el Budget revisions (Paraguayan Guaranies) presupuesto Presupuesto vigente Revised Budget = Initial Budget + Budget Revisions (Paraguayan Guaranies) Obligado Released Budget (Paraguayan Guaranies), in practice it is treated as the Executed Pagado expenditure Actual Payments (Paraguayan Guaranies)
Each variable is labeled with their original name as used by the Government of Paraguay. For the administrative and economic classification variables a standard BOOST name is given that identifies the type of budget classification to which the variable corresponds e.g., Econ1 corresponds to the first (top) level of the country s economic budget classification. For the administrative classification there is both the BOOST name and the original name in parentheses. The database contains three variables that correspond to different administrative classifications within the Paraguay budget. The first level of administrative classification, admin1, identifies the Level of Government Spending (e.g. Judicial Branch, Regional Governments, and National Universities). The second level of administrative classification, admin2, identifies the Ministries, agencies, Sub-national governments, or entities. The third level administrative classification, admin3, specifies the spending units under which the funds were spent. The admin2 reciente variable converts the admin2 classification used for that year into the equivalent for the year 2012. There are also two levels of program classification (Programa and Sub-programa). Additionally, the database contains variables that correspond to the three levels of economic classification (Econ1, Econ2 and Econ3); also a similar set of variables classified by the Objeto del Gasto (Grupo, Sub-grupo and Objeto Gasto) are included; a visual representation of these Objeto del Gasto variables can be found in Figure 4 below. Figure 4. Hierarchy of Objeto del Gasto Classification Variables Grupo Subgrupo Objeto Gasto The BOOST database includes three variables that correspond to the three levels of functional classification (Finalidad, Función and Sub-función), A visual representation of the functional hierarchy can be found in Figure 5 below.
Figure 5. Hierarchy of functional Classification Variables Finalidad Función Subfunción There are consolidated transfers between public sector agencies and entities aimed at capital and current expenditure. These include transfers granted to government organizations and entities that come from revenue sharing agreements and other earmarked special resources. These are the transfers identified in the BOOST database. The Entidad Madre variable is the admin2 variable for the parent entity, i.e. the one that gives the transfer; Entidad Hija/Receptora is the admin2 variable for the corresponding recipient entity of the transfer. Finally, the database contains four variables that correspond to different budget amounts: Presupuesto inicial contains the nominal amount (Paraguayan Guaranies) that was approved by the Paraguayan national congress in each year s budget for a specific line item; Modificación en el presupuesto are the amends to the approved budget; Presupuesto Vigente is the modified budget (approved plus the amends) Obligado lists the corresponding amount the released budget; while Pagado contains actual payments associated with that line item.
4. How to use the Paraguay BOOST database For ease of use, the BOOST team has developed a standard user interface through which to access the BOOST government expenditure database with the help of Excel PivotTables. A PivotTable report is an interactive way to quickly summarize large amounts of data. Use a PivotTable report to analyze numerical data in detail and to answer unanticipated questions about your data. A PivotTable report is especially designed for: Querying large amounts of data in many user-friendly ways. Subtotaling and aggregating numeric data, summarizing data by categories and subcategories, and creating custom calculations and formulas. Expanding and collapsing levels of data to focus your results, and drilling down to details from the summary data for areas of interest to you. Moving rows to columns or columns to rows (or pivoting ) to see different summaries of the source data. Filtering, sorting, grouping, and conditionally formatting the most useful and interesting subset of data to enable you to focus on the information that you want. Presenting concise, attractive, and annotated online or printed reports. 1 PivotTables are straightforward and easy to use and allow for quick, customizable analyses of large amounts of data. This section presents several examples of using the BOOST PivotTable interface to general custom reports. With BOOST, as with many things in life, the best way to learn is by doing. 2 1 For more information on Excel PivotTables, please consult the help function in Microsoft Excel or this helpful overview of the Microsoft website: http://office.microsoft.com/en-us/excel-help/overview-ofpivottable-and- pivotchart-reports-hp010177384.aspx. 2 The minimum technical requirements for using the Paraguay BOOST government expenditure database are as follows: (i) a computer with at least 1 GB of RAM (2 GB preferred); (ii) Microsoft Excel version 2007 or later (or similar database software that allows loading of files with at least one million lines of data).
Example 1: Time Trend Analysis Figure 6 below presents a simple example of time trend analysis at the macro level. It reports (in nominal terms) the central government expenditures from Fiscal year 2003to 2012, broken down by the top- level administrative classification. To generate this PivotTable, admin1 is placed in the Row Labels box, Año in the Column Labels box. The values in the body of the table consist of the sum of the Presupuesto inicial variable. Figure 6. Examining Central Government Expenditures across Years
Example 2: Composition Breakdown Figure 7 below presents a breakdown of 2007 government expenditures by economic category and sub-category for each ministry in Paraguay s functional budget classification. To generate this PivotTable, Econ1 and Econ2 are placed in the Row Labels box, admin2 in the Column Labels box, and the Año ( 2007 ) variable is used as filter. The values in the body of the table consist of the sum of the Presupuesto inicial variable. Figure 7. Examining the Composition of Expenditures by Sector
Example 3: Geographic Variation Figure 8 below presents a breakdown of 2007 expenditures by department/region and the second level of functional classification. To generate this PivotTable, Función is placed in the Row Labels box, Departamento in the Column Labels box, and Año ( 2007 ) variables is used as filter. The values in the body of the table consist of the sum of the Presupuesto inicial variable. Figure 8. Examining Differences in Spending across Geographic Areas