1

Transparency at Work:

Monitoring Corruption with the Government Integrity Index System

Lung-Teng Hu, Ph.D.Assistant Professor

Department of Public Policy and ManagementShih Hsin University

Taipei, Taiwan

Director of Knowledge ManagementTI-Chinese Taipei

2

Government Integrity IndexG I I

4 Dimensions13 Constructs27 Indicators

4 Dimensions13 Constructs27 Indicators

Objective Indicators&

Subjective Indicators

Objective Indicators&

Subjective Indicators

23 municipalities & counties

3

ImpactImpact

Structure of Government Integrity Index (GII)

OutputOutputProcessProcessInputInput

4 Dimensions

4

Structure of Government Integrity Index (GII)

13 Constructs

Input Human Resources

Budget

Law and Regulations

Process Procurement

Anti-Corruption Audit

Public Education on Anti-Corruption

5

Structure of Government Integrity Index (GII)

Output Complaints

Disclosure

Misconduct

Law Breaking

Impact Media Report

Staff Perception

Public Opinion

6

Structure of Government Integrity Index (GII)

Government Integrity

Index

Input Index

Process Index

Output Index

Impact Index

Human Resources

Budget

Law and Regulations

Procurement

Anti-corruption Audit

Public education on

anti- corruption

Complaints

Disclosure

Misconduct

Media report

Staff perception

Public opinion

Law Breaking

Objective Indicators

Subjective Indicators

7

Structure of Government Integrity Index (GII)

Objective Indicators: come from official statistics

Subjective Indicators: come from two surveys• Public opinion telephone survey (hereafter

Public Opinion Survey)• Staff mailing survey (hereafter Staff Survey)

8

Operationalization of GII

Stage 1Standardization: from original statistics to

standardized Z scores.Normalization: multiply each standardized Z

score by -1, if necessary

• If the statistics look neutral, use their correlations with public opinion survey results to determine the directions

9

Operationalization of GII

Stage 2Combining normalized standardized scores into

sub-dimension scores.

• Weighting method:

(1) using consensus by Delphic method, or

(2) performing factor analysis for each sub-dimension

extract only the first factor then using regression

method to get weights

10

Operationalization of GII

Stage 3

Combining sub-dimension scores into dimension scores.

Weighting method:

(1) using consensus by Delphic method, or

(2) performing factor analysis for each

dimension

Dimension score adjustment using linear transformation,

• SAx = 70 + (10*Sx)

11

Operationalization of GII

Stage 4Combining dimension scores into final index.

Weighting method:

(1) using consensus by Delphic method, or

(2) performing factor analysis on six

dimension scores

Final index adjustment

using linear transformation

GII = 70 + (10*FI)

12

Features of GII Results

We have finished our Beta Version of GII with data from 23 municipalities/counties

• We are working on the second round data collection

Grouping rather than ranking by multiple comparison technique

13

Citizens’ Assessment on Governmental Integrity in General

新竹市

嘉義市

宜蘭縣

臺南市新竹縣

臺北縣

高雄市

基隆市

澎湖縣高雄縣嘉義縣臺中市

屏東縣桃園縣 彰化縣花蓮縣 苗栗縣南投縣 臺東縣

臺中縣雲林縣臺南縣

臺北市

-40%

-30%

-20%

-10%

0%

10%

20%

30%

40%

0 1 2

14

Citizens’ Assessment on Magistrates/Mayors’ Integrity

高雄縣新竹市

南投縣臺東縣

新竹縣

基隆市

臺中市桃園縣嘉義市 屏東縣

臺北市高雄市 花蓮縣澎湖縣 雲林縣宜蘭縣

彰化縣臺中縣 苗栗縣

嘉義縣

臺南市臺北縣臺南縣

-45%

-25%

-5%

15%

35%

55%

75%

0 1 2

15

Citizens’ Assessment on Department Chiefs’ Integrity

臺中市

新竹市

嘉義市

臺北市

高雄市

花蓮縣

澎湖縣

雲林縣

臺東縣

臺南市

臺北縣臺南縣

基隆市

高雄縣桃園縣

屏東縣

宜蘭縣

彰化縣

臺中縣苗栗縣

嘉義縣

南投縣

新竹縣

5%

10%

15%

20%

25%

30%

35%

40%

0 1 2

16

Citizens’ Assessment on Public Employees’ Integrity

高雄縣

嘉義市

高雄市

澎湖縣

雲林縣

苗栗縣南投縣

臺東縣

臺南市

臺中市新竹市

桃園縣 屏東縣

臺北市

花蓮縣宜蘭縣

彰化縣臺中縣嘉義縣

臺北縣

臺南縣

新竹縣

基隆市

35%

40%

45%

50%

55%

60%

0 1 2

17

, 57.2高雄市

, 60臺北縣, 61.1臺南市

, 62.4南投縣

, 63.9雲林縣

, 65.4臺南縣

, 66.8新竹縣

, 69.6宜蘭縣, 70.3臺東縣

, 71.9苗栗縣, 72.7彰化縣, 73.6高雄縣, 74.3桃園縣

, 77.2臺中市

, 79.8花蓮縣

, 81.2屏東縣

, 84.5新竹市

, 85.7嘉義市

, 67.1臺中縣

, 70.1臺北市

, 73.2嘉義縣

, 81.4澎湖縣

55

65

75

85

0 1 2

Final Scores in GII Beta Version

18

Why grouping?

Think about this…

If the score difference between the Last No.4 city/country and the Last No.3 is 50, while the difference between the Last No.2 and the Last No.1 is 0.5…

Can we say this ranking is fair??

19

Conclusions

We believe that• Using grouping technique rather than ranking

method has some advantages:

• taking the concept of “variation” into account,

• making the assessment results are fairer and more acceptable,

• minimizing the emotional overreaction or critique from the evaluated objects.

20

Conclusions

Who has been involved in the GII measurement?Directly involved:

• Citizens

• Public employees

Indirectly involved:

• The media (by news reports/coverage)

• Governments themselves (by official statistics input)

21

Conclusions

Impacts:

• Educating public officials that corruption/integrity can be measured.

• Requesting agencies to collect needed data regularly.

• Promoting the idea of “indicator management” to government-wide Department of Government Ethics.

Challenges:

• Responding rate of staff survey is quite low, probably due to the sensitivity of the issue.

• Need to prevent from the systematic bias occurring from specific departments/local governments.

22