Good Governance Index (GGI)

Technical Notes on the Good Governance Index1


I. Introduction
II. Conceptual  Framework
III. Indicators Used and Data Sources for GGI
IV. Definition of Terms
V. Computation of the GGI
VI. Limitations and Improvements Done on the GGI  Methodology
VII. Computation of the Voters’ Index

I. Introduction

Good governance promotes the collective responsibility of the government, civil society and private sector for  improving the lives of all citizens, particularly the poor. Good governance, therefore, is  the key to a successful development agenda. But this calls for good governance not only at the national and international levels but also at the subnational level. While several governance indicators  have been developed at the international and country level often based on perception surveys, there are not enough useful specific and more objective indicators that  can be used  to guide our policymakers and stakeholders at the national and subnational level.

To promote good governance in the country, the former NSCB developed the Good Governance Index (GGI). The  GGI aims to come up with objective, comprehensive and comparable measures of good governance to guide policy makers and stakeholders in the country in assessing local and national performance,  thereby promoting more evidence- based  policy making and decision-making towards  good   governance.  Specifically, the GGI aims to (1) draw attention to good governance outcomes particularly the achievements of the best  performing local government units (LGUs) in order to encourage  and sustain and promote  best practices on governance in the country; (2) determine  and address the specific areas for improvement in order to promote good governance; and (3) improve the generation, analysis and utilization of government statistics by addressing the most pressing  information needs of the society in order to  contribute to  people empowerment  through  statistics.

Based on the framework developed by the former NSCB, the GGI results had been presented in a series of governance papers by Virola, et al, during the the 9th, 10th and 11th  National Convention on Statistics (NCS) conducted in EDSA Shangrila Hotel, Mandaluyong City in 2004, 2007 and 2010, respectively, and  Statistically Speaking articles, other national and international fora.

II.  Conceptual  Framework

Governance, as defined  in the former NSCB framework, is the manner in which power is  exercised in the management of the country’s economic and social resources for  development. It also refers to the exercise of economic, political and administrative authority to manage the nation’s affairs at all levels. Thus, the framework covers  three  types  of governance, namely: Economic Governance, Political Governance and Administrative Governance. The framework  diagram is shown in the Appendix A.

To measure these  three major areas of governance, the GGI includes the   following three (3) major indices, namely, the Economic Governance Index (EGI), the Administrative Governance Index (AGI) and the  Political Governance Index (PGI). The GGI and its component indices attempt to measure the following important  dimensions of good governance: (a) sustainable management of resources through generation of adequate financial resources and  responsiveness to/ alleviation of poverty ( through  EGI) ; (b) rule of law through  improvement of internal and external security, law enforcement and administration of justice ( through PGI) ; (c) efficiency of the delivery of services on health, education, and power supply (through  AGI);  and (d)  people’s  empowerment and participation  ( through PGI).

Although included in the former NSCB framework on governance, the following dimensions of governance were not  represented in the computation of the GGI  due to unavailability of   data or incompleteness of  data  for all provinces: (a) elimination of  graft and corruption, (b) improved transparency and accountability (through EGI), (c) ICT readiness and improvement of technology, and (d) continuous building of  LGU capacities.  Since there are no complete updated provincial data at present on basic infrastructure such as construction of roads and bridges which are common projects of the local executives, the infrastructure index was not included in the latest computation of the 2005 and 2008 AGI and GGI.

III.  Indicators Used and  Data Sources

In the choice of the indicators in this study, aside from the availability of provincial level data and availability of more timely data, only those indicators which are more or less within the control of the executive are included since the purpose is to assess the performance of the LGU executive. The following indicators with their data sources were used in the computation of the GGI:

Governance Indicator

Sources of Data

Economic Governance

     Total Financial Resources Generated

Commission on Audit (COA)

     Total Revenue Collections(Tax and Non-Tax Revenue)

Bureau of Internal Revenue

     Total Deposits

Bangko Sentral ng Pilipinas

     Expenditure on Social Services


     Unemployment Rate

Philippine Statistics Authority (PSA)

     Underemployment Rate

     Poverty Incidence

     Poverty Gap

     Inflation Rate

Political Governance

     Crime Solution Efficiency Rate

Philippine National Police (PNP)

     Voters’ Turn-out Rate

Commission on Elections

Administrative Governance

     Elem. Teacher to Pupil Ratio

Department of Education (DepEd)

     High School Teacher to Student Ratio


    Number of Public Elem. Schools per
    1000 School-Age Population


    Number of Public High Schools per
    1000 School-Age Population


    Enrolment in Government Elem. School


    Enrolment in Government High School


    Elementary Cohort Survival Rate


    High School Cohort Survival Rate


    Elem. Pupil-Classroom Ratio


    High School Student-Classroom Ratio


   Total Health Personnel per 1000 Population

Department of Health (DOH)

    % Birth less than 2500g


    Length of National and Local Roads

DPWH-for national

DILG- for local

    Percent of Energized Barangays

National Electrification Authority

    Telephone Density

National Telecommunications Commission

Examples of indicators excluded due to data constraints are court case disposition, cases resolved on graft and corruption and infant mortality rate. Examples of indicators that were excluded because they are not clearly within the control of the LGU executives are population growth rate, number of banks and number of pawnshops.

IV.   Definition of Terms

The following terms were used in the study:

Cohort survival rate - The percentage of enrollees at the beginning grade or year in a given school year who reached the final grade or year of the elementary/secondary level. (NSCB Resolution No. 15 - Series of 2006)

Pupil/Student-classroom ratio - The average number of pupils/students per classroom in elementary/secondary education in a given school year. (NSCB Resolution No. 14 - Series of 2006)

Pupil/Student-teacher ratio - The average number of pupils/students per teacher in elementary/secondary education in a given school year. (NSCB Resolution No. 14 - Series of 2006)

Skilled health personnel - Accredited health professionals, such as midwives, doctors, and nurses, who have been educated and trained to proficiency in the skills needed to manage normal (uncomplicated) pregnancies, childbirth, and the immediate postnatal period, and in the identification, management, and referral of complications in women and newborns. (NSCB Resolution No. 19 - Series of 2009)

Inflation rate - the annual rate of change or the year-on-year change in the Consumer Price Index. (NSCB Resolution No. 11 - Series of 2003)

Poverty gap  -  the total income/ expenditure shortfall (expressed in proportion to the poverty threshold) of families/ individuals with income/ expenditure below the poverty threshold, divided by the total number of families/ individuals. (NSCB Resolution No. 2 - Series of 2007)

Poverty incidence  -  the proportion of families/individuals with per capita income/expenditure less than the per capita poverty threshold to the total number of families/individuals. (NSCB Resolution No. 2 - Series of 2007)

Telephone density - Number of fixed telephone lines per 100 population. (NSCB Resolution No. 2 - Series of 2008)

Unemployment rate - Percentage of the total number of unemployed persons to the total number of persons in the labor force. (NSCB Resolution No. 14 - Series of 2007)

Underemployment rate - Percentage of the total number of underemployed persons to the total number of employed persons. (NSCB Resolution No. 14 - Series of 2007)

V.  Computation of  the  GGI

The GGI for each province is computed as the weighted arithmetic average of the Economic Governance Index (EGI), the Political Governance Index (PGI) and the Administrative Governance Index (AGI). Since the PGI has only two available indicators, more weights were given to the EGI and AGI. Thus,

GGI formula

In the course of assessing the provincial GGI, it was realized that the data constraints also gave undue weights to some variables. In the original formulation based on the framework, equal weights are assigned to the different variables. When some of the variables have no data support, they were dropped thereby increasing the weight of the other variables comprising the subindex which covers the dropped variables.

Thus, the provincial GGI formula  used in computing the 2005 and 2008 GGI  is  now weighted to rectify the bias. Moreover, in computing the  2008 GGI in 2010, the infrastructure index was not included due to unavailability of   updated complete  data on length of national and local roads. The original methodology of computing the GGI and component indices are found in the previous papers listed in the Appendix B.

On the other hand, the AGI, EGI and the PGI are computed as unweighted averages of the indexes corresponding to their subthemes. Thus,

            EGI      =         (SMRI + EGRPI)/2,
            PGI      =         (IIESI + LEAJI + EGCI)/3, and
            AGI     =          (EI + HI + ((PI +TDI)/2))/3

Computation of the Detailed Component   Indices:

To introduce some benchmarking of the index, the index for the base year which is Philippines 2000 is set at 100.

At the lowest level of indexing, for the positive indicators, the index for a province is obtained  by dividing the value of the indicator for the province by the value of the indicator for Philippines 2000. Examples of positive indicators include total financial resources generated, BIR tax and non-tax collections, total health personnel per 1000 population, cohort survival rate, etc. For negative indicators, the index for a province is obtained by dividing the value of the indicator for Philippines 2000 by the value of the indicator for the province.  Examples of negative indicators include inflation rate, poverty incidence, poverty gap, unemployment rate,  percentage of birth less than 2500g, etc.

For all the detailed component indices in the 2005 and 2008 GGI, the 1.96 SD limit truncation was used. 

1. Economic Governance Index  (EGI) :         

 EGI = Ave ( SMRI + EGRPI)


a. Sustainable Management of Resources Index( SMRI)= Ave( Management of Financial Resources Index + Management of Human Resources Index)

= Ave ( MFRI + MHRI),   where

            MFRI   = Ave (Generation of Adequate ResourcesIndex + Per Capita Expenditure on Social   Services Index)

=  Ave ( GARI +  PCESSI)


GARI =Ave( Per Capita Financial Resources Index + Per Capita Revenue  Index + Per Capita Total Deposits Index )


PCESSI=Per Capita Expenditure on Social Services Index

MHRI  = Ave(Unemployment Rate Index+ Underemployment Rate Index)                        

b. Enhanced Government Responsiveness to the Poor Index (EGRPI)

 (EGRPI) =  Ave(Poverty Incidence Index+Poverty Gap Index+ Inflation Rate Index)

2. Political Governance Index  ( PGI) 



a. Improvement of Internal and External Security Index ( IIESJI) =Crime Solution Efficiency Rate Index

b. Law Enforcement and Administration of Justice Index (LEAJI) = Voters’ Turn-out  Rate Index 

c. Elimination of Graft and Corruption Index  (no indicators used)         

3. Administrative Governance Index ( AGI)

AGI  = EI + HI + (( PI +TDI)/2)


EI = Education Index =Ave (Elementary and High School Teacher Pupil Ratio Index + Number of Public Elementary and High Schools Per 1000 Population Index +Total Enrolment in       Government Elementary and High Schools Per 1000 Population Index + Elementary and High School Cohort Survival Rate Index + Elementary and High School Pupil-Classroom Ratio Index)

HI = Health Index =Ave ( Health Personnel Per 10,000 Population Index + Percent of Households with Access to Safe Water Index + Live Births Less Than 250 Grams Per 1000 Births Index + Number of Barangay Health Stations per 100,000 Population Index)

PI = Power Index = Percent of Energized Barangays Index

TDI = Telephone Density Index 

There are no available  updated complete data for the following subthemes  so these were not represented  in the computation of  2005 and 2008 AGI:

  1. Improved Transparency and  Accountability Index (ITAI)
  2. Continuous Building of Capacities Index (CBCI)

The original  methodology  in computing the AGI is found in the 2004  and 2007 NCS papers  on governance listed in Appendix B.

VI.  Limitations and Improvements Done on the GGI  Methodology

The limitations of the GGI  have been presented in the previous papers of Virola et. al. The most crucial of these limitations are the following: (1) validity of the GGI as a measure of governance over a period of time, (2) data availability constraints including the desired level of disaggregation, (3) timeliness of the data, and (4) appropriateness of the indicators used.

While the weights have been changed, it was noticed that the imbalance in the availability of the indicators for the different components of the GGI has continued to give inordinately bigger weights to some indicators like the landline telephone density index and the unemployment/underemployment rates based on unupdated data.

It is not necessarily the case that a low/high GGI for a particular year is an indication of bad/good governance of the LGU officials at that time, even if the indicators used for the GGI are very timely. Some needed reforms and some good interventions take time to have an impact on the lives of the governed, meaning that the effect of these good governance efforts on the GGI will show up not necessarily during the year the reforms were implemented but a few years later, when possibly, a new set of LGU officials would have taken over. 

Based on the latest available information, the GGI is computed for 2005 and 2008. While the 2005 GGI may be considered as a good approximation of the state of the LGU at the start of the term of officials elected for the term 2007-2010, the 2008 GGI cannot measure all that happened during the 2007-2010 term. For instance, the 2008 GGI will underestimate the impact of interventions started during the second half of the term. It cannot be an accurate measure therefore of the progress of provinces like Albay where projects and programs were furiously being implemented until the end of the 2007-2010 term. Corollarily, it would overestimate the achievements of provinces where bad or no governance was in place. Also, given the different roles of the various LGU officials in running the LGU, the GGI cannot be interpreted as the sole responsibility/accountability of the top LGU official, say the governor or the mayor. Attribution should thus be very carefully managed.

Improvements made on the GGI methodology  in 2007

A  major improvement made in the original methodology in 2007 was the exclusion of the contribution of chartered cities in the computation of the provincial GGI.  Chartered cities act independently from their  provinces and  are largely self governing.  Aside from financial independence, they also have  their  own representatives in Congress. Additional indicators such as the cohort survival rate in education ,  inclusion of high school education variables instead of only elementary education, voters’ turnout rate, pupil-classroom ratio, prices as measured through the inflation rate, and proportion of barangays with health centers  which were not part of the original methodology were incorporated in the computation of  the GGI in the  revised or improved  methodology. However, it is to be noted that  a smaller average number of pupils per classroom may not necessarily be  more desirable than a bigger one.

Improvements made on the GGI methodology  in 2010

 Based on feedback about the provincial GGI, there does not seem to be any problem with the framework. The weakness is on the list of variables that make up the GGI and on the timeliness of the data used. But as already explained, this emanates from data availability constraints – there are not that many consistent and comparable datasets across provinces and municipalities that are regularly generated by the Philippine Statistical System  at present. Consequently, the following changes were made on the GGI methodology in 2010:

It was realized that the previous  truncation methodology was not very effective in toning down the effect of a single variable.  Hence, for all the indices in 2005 and 2008, the 1.96 SD limit truncation was used.In the old methodology, level 1 and level 2 indices were truncated so as not to exceed the 500 and 400 limit, respectively, while the level 3 index was truncated not to exceed 300. In the new methodology, values were truncated to 1.96 times the standard deviation, the 95 percent confidence interval for a normally distributed random variable.

      In the original methodology, it  was realized that the operational framework and its equal-weighting scheme did not give due importance to the education, health and poverty variables vis a vis the other variables like the landline telephone density index. It was thus decided to revise the weights, giving a weight of 3 to the Administrative Governance Index (AGI) where the health and education variables are included, 3 to the Economic Governance Index  where poverty variables are included, and 1 to the Political Governance Index, which has only two indicators. The AGI  was also revised to give greater weights to the education and health variables and less weight to the landline telephone density index.

Another major improvement was the development of  Municipal Good Governance Index.  For each municipality, a Good Governance Index (GGI) is computed as the unweighted arithmetic average of the Income Index (II), and Expenditure Index (EI). Thus,

GGI = formula1

The II, and the EI are computed as the unweighted averages of the indexes corresponding to their subthemes. Thus,

                        II   =     form2, and

                        EI  =     form3


            TIPCI = Total Income Per Capita Index,
            PCILSI = Per Capita Income from Local Sources Index,
            PCEEI = Per Capita Expenditure on Education Index,
            PCEHI = Per Capita Expenditure on Health Index, and
            PCEESI = Per Capita Expenditure on Economic Services Index

For the interim municipal GGI, the variables considered were those from the publication of the Bureau of Local Government Finance, entitled Statement of Income and Expenditures (SIE) of Local Government Units (LGUs). The SIE shows which municipalities/LGUs are able to acquire more financial resources and how they allocate these resources to various expenditure items. It gives us an indication of the amount of resources at the disposal of the municipal/LGU executives, how these resources are generated and spent, and whether they are spent for development.  Although the SIE has its limitations and the municipal/LGU executives cannot solely be held accountable for the indicators therein, these indicators tell us how well the municipalities/LGUs respond to the needs and promote the welfare of their constituents, how they govern.

Specifically, the variables used in the interim municipal GGI are per capita income, per capita income from local sources, per capita expenditure on education, sports and manpower development, per capita expenditure on health, nutrition and population control and per capita expenditure on economic services. These indicators provide indication on the  quality of management of resources thereby promoting more transparency and accountability which are important dimensions of good governance. However, unlike the provincial GGI which used 2000 as the base year, the interim municipal GGI has no base year.

The municipal GGI is computed as  the unweighted average of the two sub-indices as there is no reason to give greater weight to expenditures than income.

However, the methodology/framework for the interim municipal GGI, needs to be revisited. The computation in real terms also needs to be considered. It was also observed that there are fluctuations in the indicators comprising the municipal GGI both on income and especially on expenditures, such as when a school building is constructed or when grants are provided. Thus, some kind of a moving average of the GGI might have to be considered

VII.  Computation of the Voters’ Index

Good governance is not only the sole responsibility of the rulers but also the people of the land. Democracy and people empowerment is  best  exercised  when the people go to the polls to vote. But how do voters assess their candidates?

In order to assess the performance of voters, a Voters’ Index has been added since the 2004 NCS paper on the GGI. The Voters’ Index aims to measure the “wisdom” of the voters in selecting their candidates. A rate or grade of “0” or “1” is assigned to a province, depending on the results of the election and their GGI: 1 for best or most improved provinces whose governor ( or when governor’s  the three year term already ended it was his/her relative) won and 0 otherwise; 1 for worst or least improved provinces whose governor  lost and 0 otherwise.

If the governor(or his/her  relative): Best or Most Improved Worst or Least Improved
Won 1 0
Lost 0 1

The resulting scores are then added and  divided by the total number of a list of provinces wherein the governors or their relatives ran for reelection or election to another office to arrive at the Voters’ Index. The list can include the top or bottom ten, twenty and thirty provinces. Those provinces whose governors had ended their third term  and therefore are not allowed by law  to seek reelection for the fourth term and where a relative did not run as well as those provinces whose governor or a relative did not run during the elections are excluded in the computation.

1These technical notes are based on the list of governance papers cited in Appendix B.

Prepared by:

Severa B. De Costo
Noel S. Nepomuceno
Gerald Junne  L. Clariño
Priscille C. Villanueva

19  January 2011


Appendix A



Source: Report on the Development of Indicators and Design of a Database and Information Network of Governance Statistics Project, NSCB.


Appendix B

A.  Papers on the Good Governance Index (GGI)

  • Virola, Romulo A.,Severa B. de Costo, Noel S.Nepomuceno, Gerald Junne L. Clariño, Priscille C. Villanueva and  Mai Lin C. Villaruel. “Was Governance the Real Loser in the 2010 Elections?. Convention Paper, 11th National Convention on Statistics, EDSA Shangri-la Plaza Hotel, Mandaluyong City, 3-4 October 2010. PDF

  • Virola, Romulo A., Severa B. de Costo & Mai Lin C. Villaruel, Measuring Democratic Governance: An Emerging Challenge to Official Statisticians. Paper presented during the 3rd OECD World Forum “Statistics, Knowledge and Policy” on Charting Progress, Building Visions, Improving Life, Busan, South Korea, 27-30 October 2009.

  • Virola, Romulo A., Severa B. de Costo, Noel S. Nepomuceno , Ma. Kristine Faith S. Agtarap, Ma. Ivy T. Querubin,   Mai Lin C. Villaruel. Governance Statistics: Did Performance Matter in the 2007 Elections? Convention Paper, 10th National Convention on Statistics, EDSA Shangri-la Plaza Hotel, Mandaluyong City, 1-2 October 2007. PDF

  • Virola, Romulo A., Severa B. De Costo & Kristine Faith S. Agtarap, “Measuring Democracy, Human Rights and Governance: Should Official Statisticians Be Involved?”, presented during the 56th Session of the International Statistical Institute in Lisboa, Portugal on 22-29 August, 2007.

  • Virola, Romulo, “Empowering And Challenging Voters Through Governance Indicators: The Philippine Experience”,  presented  in the OECD’s 2nd World Forum on Statistics, Knowledge, and Policy “Measuring and Fostering the Progress of Societies – The Construction and Use of Indicator Sets: Lessons to Build Modern Democracies” in Istanbul, Turkey from 27-30 June 2007.

  • Virola, Romulo A., “Indicators of Democratic Governance as Guide to Voters in Philippine Elections: Inspiration from METAGORA", Regional Conference for Asia on Measuring and Fostering the Progress of Societies, Measuring Democracy and Human Rights, Seoul, Korea, 8-9 February 2007.

  • Virola, Romulo A., Severa B. De Costo, Joseph M. Addawe,  Leonor G. Reyes. The Best and Worst Provinces in the Philippines: What Happened to Their Leaders During the 2004 Elections? Convention Paper, 9th National Convention on Statistics,  3-4  October  2004.PDF


B. Statistically Speaking articles on the GGI