Special Report: The DDPC Program
by: Mitchell A. Seligson Department of Political Science University of Pittsburgh Pittsburgh, PA 15260 (412) 648-7268 [email protected] December, 2000
Prepared for U.S. Agency for International Development La Paz, Bolivia
Version: June 25, 2007 Special Report: The DDPC Program 2 Contents
Contents
Introduction: Methodological Considerations ...... 1 Chapter I: Sample Characteristics ...... 1 Chapter II: Support for Democracy ...... 3 Chapter III: System Support and Tolerance ...... 7 Chapter IV: Corruption and System Support ...... 19 Chapter V: Ethnicity and Democracy ...... 24 Chapter VI: Support for Anti-Democratic Measures ...... 26 Chapter VII: The Rule of Law and Democracy ...... 29 Chapter VIII: Local Government ...... 37 Chapter IX. Women and Democracy ...... 37 Conclusions...... 40 Special Report: The DDPC Program 3 Contents List of Figures Figure 1 Years of Education, National vs. DDPC Samples: Confidence Intervals by Sample ...... 2 Figure 2 Income, National vs. DDPC Samples: Confidence Intervals by Sample ...... 3 Figure 3 Percent Preferring Democracy...... 4 Figure 4 Preference for Participation and Not a Strong-Fisted Government ...... 5 Figure 5 Is Warrant Required When Criminal Activity is Suspected? ...... 6 Figure 6 Preference for Liberty vs. Order...... 7 Figure 7 System Support: 5-Item Index ...... 8 Figure 8 System Support: Main National Institutions ...... 9 Figure 9 System Support: Main National Institutions, All Samples ...... 10 Figure 10 System Support: Judicial Institutions ...... 11 Figure 11 System Support: Judicial Institutions, All Samples ...... 12 Figure 12 System Support: Other Institutions, All Samples ...... 13 Figure 13 System Support: Local Government, All Samples ...... 14 Figure 14 System Support: Local Government, Confidence Intervals by Sample ...... 15 Figure 15 Political Tolerance...... 16 Figure 16 Political Tolerance: Confidence Intervals by Sample ...... 17 Figure 17 Attitudes Favoring Stable Democracy ...... 18 Figure 18 Attitudes Favoring Stable Democracy: Confidence Intervals by Sample ...... 19 Figure 19 Corruption: DDPC, 1988, 1999, 2000 ...... 20 Figure 20 Corruption: DDPC vs. National 2000, Selected Measures ...... 21 Figure 21 Corruption Experience, DDPC Samples: Confidence Intervals by Sample, Combination of 1998, 1999 and 2000 Samples ...... 24 Figure 22 Ethnic Self-Identification by Sample: All Years ...... 25 Figure 23 Indian vs. Western Dress: National vs. DDPC Samples ...... 26 Figure 24 Justification for a Coup: Composite Index, Confidence Intervals by Sample ...... 27 Figure 25 Coup Justified Because of April 2000 Protests: Confidence Intervals by Sample ...... 28 Figure 26 Approval of Civil Disobedience: Sample Comparison ...... 29 Figure 27 Support for the Rule of Law: Confidence Intervals by Sample ...... 30 Figure 28 Crime Victimization: Confidence Intervals by Sample ...... 31 Special Report: The DDPC Program 4 Contents Figure 29 Heard about the New Criminal Code: Confidence Intervals by Sample ...... 32 Figure 30 Support for Citizen Judges: Confidence Intervals by Sample ...... 33 Figure 31 How Easy Is It to Report a Crime? Confidence Intervals by Sample ...... 34 Figure 32 Treatment in Courts: Confidence Intervals by Sample ...... 35 Figure 33 Reporting of Crime: Confidence Intervals by Sample ...... 36 Figure 34 Treatment by the Public Ministry: Confidence Intervals by Sample ...... 37 Figure 35 Believe That There Is Discrimination against Women: Confidence Intervals by Sample ...... 38 Figure 36 How Serious Is Discrimination against Women? Confidence Intervals by Sample ...... 39 Figure 37 Do Women Have Equal Employment Opportunities with Men? Confidence Intervals by Sample ...... 40 Special Report: The DDPC Program 5 Contents
Introduction: Methodological Considerations
This report is a follow-on to the study The Political Culture of Democracy in Bolivia: 2000 that was prepared for USAID. That report dealt with the national sample of 2000, and compared it to the national sample of 1998. In a few places, especially with regard to the chapter on municipal government, that report did analyze the special DDPC samples that were drawn in 1998 and 2000. In this special report, the results of all three DDPC samples are presented: 1998, 1999 and 2000.
The reader needs to keep in mind the point made in the 1999 report on the DDPC samples, namely that the 1998 sample covered only six municipalities, while the 1999 (and now the 2000) covered nine. That is because at the time the 1998 study was being designed, the DDPC program was not yet working in more than six municipalities. The 1999 study went to considerable lengths to dissect the sample, comparing the original six municipalities with their match in 1999. Furthermore, an added complexity was that the 1998 national study contained 100 interviews in municipalities that were also part of the DDPC sample. That was done in 1998 for cost considerations. The present study makes direct comparisons of the three DDPC samples to the two national samples. In order not to make these comparisons impossibly complex, they treat the three DDPC samples as if they were all the same, and also overlook the fact that 100 of the DDPC interviews in 1998 (but not in 2000) were also countable as being part of the national sample. The advantage of following this procedure here is that comparisons can be easily made between the three DDPC samples (1998, 1999 and 2000) and with the two national samples (1998 and 2000).
The plan of analysis follows directly the main report on the national sample. That is, this special report replicates many of the analyses shown in the larger report, but the focus here is on the comparisons of the DDPC samples to each other and to the nation. The discussion of measurement and theory will not be repeated here. Interested readers can refer directly to the larger report to find that material, which is presented in the same order that it is presented in this report, with the chapter numbers repeated in each section.
Chapter I: Sample Characteristics
We first present here the basic socio-economic information on the samples, leaving ethnic comparisons to chapter V. In terms of education, Figure 1 shows the comparisons. As can be seen, the national samples (including both the 1998 and 2000 samples) average significantly higher years of formal schooling than the DDPC samples (including 1998, 1999 and 2000). This difference is a function of the fact that the DDPC samples did not include any of the major cities of Bolivia, where education levels are higher. Special Report: The DDPC Program 6 Contents
Years of Education, National vs. DDPC Samples
Confidence Intervals by Sample 10.5
9.5
8.5 95% CI, year of education
7.5 N = 6069 2293 National DDPC
Sig. < .001 Figure 1 Years of Education, National vs. DDPC Samples: Confidence Intervals by Sample
Income also varies between the national and DDPC samples. Figure 2 shows that the DDPC samples have a significantly lower income than the national samples. Special Report: The DDPC Program 7 Contents
Income, National vs. DDPC Samples
Confidence Intervals by Sample 3.0 501-1,000 Bs.
2.8
2.6
2.4
2.2 95% CI monthly income 2.0 251-500 Bs.
1.8 N = 5605 2130 Nat D DP io n C a l
Sig. < .001 Figure 2 Income, National vs. DDPC Samples: Confidence Intervals by Sample
Chapter II: Support for Democracy
The central question in this chapter was AOJ14, measuring a preference for democracy, authoritarian rule or not having a preference for either one. This item, as noted in the main report, was not asked in 1998, but was included for the first time in the 1999 survey. Figure 3 shows the comparisons for the 1999 survey with the 2000 survey, broken down by national vs. DDPC. As can be seen, over 70% of the respondents in each of these samples preferred democracy over authoritarianism or over not caring which regime is in power. There are, however, no significant differences among the samples. Special Report: The DDPC Program 8 Contents
Percent Preferring Democracy 100%
90%
80%
70% 72 72 73 60%
50%
40% % preferring 30% YEAR 20% 1999 10% 0% 2000 National DDPC
Sig. = NS Figure 3 Percent Preferring Democracy
The question on a preference for a strong-fisted (mano dura) government (AOJ15) is analyzed in Figure 4. This item was asked for the first time in 2000, so data from 1998 and 1999 are excluded. As can be seen, respondents in both the national and DDPC samples overwhelmingly oppose strong-fisted rule, with higher, but not significantly higher opposition in the DDPC sample. Special Report: The DDPC Program 9 Contents
Preference for Participation
and Not a Strong-Fisted Government 77.5% 77.3 77.0%
76.5%
76.0%
75.5%
75.0%
% preference for participation 74.5% 74.6 74.0% National DDPC
National vs. DDPC
Sig. = NS Figure 4 Preference for Participation and Not a Strong-Fisted Government
Another question in this series (AOJ11) asked about the need for a search warrant when criminal activity is suspected. This question was not asked in 1998, so comparisons are only possible for the 1999/2000 DDPC samples. Figure 5 shows the results. Both DDPC samples show stronger support for this democratic norm than does the population as a whole, but there was a decline between 1999 and 2000. Special Report: The DDPC Program 10 Contents
Is Warrant Required When Criminal
Activity is Suspected? 72%
71 70%
68%
66% 66 % requiring warrant 64% 1999 63 62% 2000 National DDPC
Nationnal vs. DDPc < .001; DDPC Sig. = .02 Figure 5 Is Warrant Required When Criminal Activity is Suspected? Special Report: The DDPC Program 11 Contents The final item in this series is AOJ12, which asks about the trade-off of order vs. liberty. Figure 6 shows the results. In the 1999 survey, the DDPC sample was significantly more likely to prefer liberty over order compared to the 2000 DDPC sample. By 2000, the DDPC and national samples were identical. Nonetheless, in all cases, majorities prefer liberty to order.
Peference for Liberty vs. Order 65%
64% 64 63%
62%
61%
60% % preferring liberty 59%
58% 58 58 1999
57% 2000 National DDPC
DDPC sig. = .02 Figure 6 Preference for Liberty vs. Order
Chapter III: System Support and Tolerance
We now turn to system support, focusing on the five-item core index described in detail in the main report. Figure 7 shows the results. Two findings stand out. First, the DDPC samples as a group have significantly higher system support than does the nation as a whole. In the two years for which the information is directly comparable, system support was higher in 1998 and 2000 for the DDPC sample than for the nation. Second, there is minor variation among the DDPC sample years that are not statistically significant. Overall, in absolute terms, the variation in system support is quite small in these comparisons. Special Report: The DDPC Program 12 Contents
System Support 5-Item Index 100
90
80
70
60
50 47 46 40 44 43 44
30
Mean system support 1998 20 1999 10 0 2000 National DDPC
DDPC by year, Sig. = NS; DDPC vs. National, Sig. < .001 Figure 7 System Support: 5-Item Index
The survey included many other system support items. Figure 8 shows the result for the main national political institutions for the DDPC samples. As can be seen, there is only minor variation over the three years of the survey. Special Report: The DDPC Program 13 Contents
System Support: Main National Institutions 50
40 40 38 36 35 35 33 30
25 23 20 1998 19 1999
10 2000 P Ele Co art n ies c gress tion court
Figure 8 System Support: Main National Institutions
Figure 9 shows these same results, but in the context of the national samples. The low support for parties stands out here, but in 2000, the national and DDPC samples showed identical levels of support on this question. Special Report: The DDPC Program 14 Contents
System Support: Main National Institutions, All Samples 50
40 41 41 40 40 38 37 36 35 34 33 30 National 1998
28 28 National 2000
23 23 20 DDPC 1998
19 DDPC 1999
10 DDPC 2000 Parties Ele Co c n ti g on court re s s
Figure 9 System Support: Main National Institutions, All Samples
A broader range of institutions, focusing on the judiciary, is shown in Figure 10, which focuses on the DDPC samples alone. Special Report: The DDPC Program 15 Contents
System Support: Judicial Institutions 70%
65 60%
58
50%
46 46 45 44 44 40% 42 41 41
30% 1998 27 26 20% 20 1999
10% 2000
Pol Publi Pu O Ju Co mbu dic ic b n e c lic stit Pr d ia d s l Coun u o efe ma tional Court secutor n n der c i
Figure 10 System Support: Judicial Institutions
Figure 11 shows the same results, but this time for all of the samples in the data set. The only important difference appears in the evaluation of the police, which especially in 1999 is lower in the DDPC areas than in the nation as a whole. But, by 2000, this difference had largely disappeared. In the other areas of support for the judicial institutions, few differences emerge among the samples. Special Report: The DDPC Program 16 Contents
System Support: Judicial Institutions, All Samples 70%
65
60% 61
58
50%
46 46 45 46 45 44 44 National 1998 42 43 43 40% 42 41 40 39 National 2000 30% 31 31 28 DDPC 1998
26 20% 20 DDPC 1999
10% DDPC 2000 Pol Pu Publi Omb Judicial Co Co ic b n e lic c u s de d tit P s u rose ma ti fend o u nal c n nci u e Court to r l r
Figure 11 System Support: Judicial Institutions, All Samples
We also asked a series of support questions for other institutions. These are shown in Figure 12, including feelings about the President of the country, an individual rather than an institution. As in many other studies in the University of Pittsburgh series, the Catholic Church has the highest level of support. Journalist in Bolivia also do quite well, with little difference from year-to-year. Presidential popularity in the DDPC sample for 1998 is very close to that of the nation in that year and when presidential popularity dropped in the 2000 national sample, the same pattern is observed in the DDPC sample. Support for unions is somewhat higher in the DDPC samples than in the nation. Special Report: The DDPC Program 17 Contents
System Support: Other Institutions, All Samples 80
70 70
67 66
63 60 62 60 59 58 58 56 50 National 1998 50 49 47 47 46 44 National 2000 40 42 42 42 42 DDPC 1998
30 32 32 30 DDPC 1999
20 DDPC 2000 Catholic Ch J Presi Uni NGO o urnal o d ns s ent ist s u r ch Figure 12 System Support: Other Institutions, All Samples
For the purposes of the DDPC project, system support for the various municipal institutions are of particular interest. Figure 13 shows the results. Trust in local government is about the same in the DDPC samples as in the nation as a whole, but trust in the indigenous authority is far higher in the DDPC samples than in the nation. Moreover, trust in the vigilance committees is higher than in the nation in two of the three DDPC samples, including the 2000 samples, as it is for the OTBs. Special Report: The DDPC Program 18 Contents
System Support: Local Government, All Samples 60
55
53 50 49
47 47 National 1998 47 47 46 46 44 45 45 44 44 43 National 2000 43 43 42 40 42 DDPC 1998 38
DDPC 1999
30 DDPC 2000 Munic In V OT d igilanc igen B i p o a u e l g s Committee overnme a u th o r ity n t Figure 13 System Support: Local Government, All Samples
Given the importance of these comparisons, statistical significant information is helpful, even if the chart itself is quite complex. Figure 14 shows these confidence intervals. The major standout is the significantly higher support for the indigenous authority in the DDPC samples. The Vigilance Committee also is significantly more trusted in the DDPC samples than in the nation as a whole. Special Report: The DDPC Program 19 Contents
System Support: Local Government
Confidence Intervals by Sample 60
50 Municipal government
95% CI Indigenous authority 40
Vigilance Committee
30 OTB Na N DDPC DDPC 1999 DDP ation tio n C al 19 a 2000 l 199 2 000 9 8 8
Figure 14 System Support: Local Government, Confidence Intervals by Sample
Political tolerance results are shown in Figure 15. In the main report it was noted that tolerance increased significantly between 1998-2000. The same pattern is found among the DDPC respondents. Special Report: The DDPC Program 20 Contents
Political Tolerance 45
44 44
43
42 42
41 41 Mean Tolerance 40 40 1998
39 1999 39 38 2000 National DDPC
DDPC Sig. = < .001 Figure 15 Political Tolerance
Given the importance of this finding, it is useful to highlight its statistical significance. Figure 16 shows the results. The national increase in 2000 is significant when compared to the national data in 1998. The DDPC samples also show a significant increase with the 2000 sample being significantly higher than the 1998 sample, but insignificantly lower than the 2000 national sample. Special Report: The DDPC Program 21 Contents
Political Tolerance:
Confidence Intervals by Sample 46%
44%
42%
40% 95% CI Tolerance
38%
36% N = 2856 2915 575 847 850 National 1998 Nation DDPC D DD DPC PC a 199 1 2 l 9 0 2 9 0 000 8 9 0
Figure 16 Political Tolerance: Confidence Intervals by Sample
The combination of system support and tolerance provides a measure of support for stable democracy. In the main report, we compared Bolivia to other countries. Here, we compare the national sample to the DDPC samples. Figure 17 shows the results. It is important to observe that the DDPC samples, which in 1998 were lower than the national average, have now increased to above the national average. Special Report: The DDPC Program 22 Contents
Attitudes Favoring Stable Democracy 15%
15 14%
13 13%
12%
12 % in "stable democracy" cell 11% 11
10 10% Na National 2000 D DD DDP DPC tional 19 PC C 1 1 2000 9 9 9 9 9 8 9 8
Figure 17 Attitudes Favoring Stable Democracy Special Report: The DDPC Program 23 Contents These findings seem to show a clear trend, but the confidence interval data need to be consulted as well. Figure 18 shows the results. The 2000 DDPC is higher than the national average, but not significantly so.
Attitudes Favoring Stable Democracy:
Confidence Intervals by Sample 18%
16%
14%
12%
95% CI DEM 10%
8%
6% N = 2783 2850 553 806 816 National 1998 Nation DDPC D DD DPC PC a 199 1 2 l 9 0 2 9 0 000 8 9 0
Figure 18 Attitudes Favoring Stable Democracy: Confidence Intervals by Sample
Chapter IV: Corruption and System Support
The focus of chapter IV in the main report was on the linkages between corruption victimization and declines in system support. There is no reason to repeat that analysis for the DDPC subset. In this section we focus instead on corruption levels within the DDPC samples vs. the nation as a whole.
Figure 19 shows the three DDPC samples and the nine measures of corruption. Note that the terminology for judicial corruption (i.e., the “courts” item) changed in 2000, as explained in the main report, a factor that contributed strongly to the increase reported in the 2000 survey. The notable changes are a decline in public employees soliciting bribes in 2000 compared to prior years, but an increase in police bribe solicitation. Special Report: The DDPC Program 24 Contents
Corruption: DDPC, 1998, 1999, 2000 50%
44 40% 43 41
30% 29 28 28 27
23 22 20% 21 22 20 20 17 17 17 DDPC 1998
13 13 13 10% 11 12 11 % experiencnig corruption 10 10 9 DDPC 1999
0% DDPC 2000 B Un At Courts Munic Pu Pol S Sa a a nk fa w b ice sol w publicw employee p ir ork lic p ip emp o o a lic lic lit ic e e y lo ite ye d e
Figure 19 Corruption: DDPC, 1988, 1999, 2000
Comparisons with the national data involve a complex chart because of the large number of corruption items and the five sets of samples. Examination of the data, however, revealed that most of the variables showed little difference between the national data and the DDPC data. A few variables are worth highlighting, however. These are shown in Figure 20. Municipal corruption is slightly higher in the nation as a whole than it is in the DDPC areas. Similarly, respondent observation of police and public employee corruption are higher in the nation than in the DDPC areas. Special Report: The DDPC Program 25 Contents
Corruption: DDPC vs. National 2000:
Selected Measures 50
47
44 40 43 41
36 30 DDPC 1998 29 28 28 DDPC 1999 20 22 % experiencnig corruption 20 20 DDPC 2000 17
10 National 2000 Mu Saw pu S a nic w polic ip a b lit lic e y e mplo
yee
Figure 20 Corruption: DDPC vs. National 2000, Selected Measures
The draft version of the main report contained a table giving breakdowns by municipality of corruption experience. That table was deleted from the public version of the report in order to avoid any difficulties with individual municipalities. It is reproduced here for the record. Readers should note very carefully the confidence intervals shown in the table. The very small sample sizes in most municipalities produce very wide confidence intervals. Special Report: The DDPC Program 26 Contents
Table I. Corruption by Municipality: National Sample, 2000 NMean 95% Confidence Interval of Corruption for Mean Victimization scale Department Lower Upper Bound Bound La Paz CALACOTO 20 17.9 8.7 27.0 CARANAVI 20 21.0 9.5 32.4 EL ALTO 99 24.0 18.7 29.2 LA PAZ 183 35.8 32.0 39.5 PUERTO CARABUCO 20 21.4 9.8 33.0 QUIME 20 40.7 24.6 56.8 TIPUANI 20 17.1 7.3 27.0 VIACHA 20 23.6 12.2 34.9 Total 402 29.2 26.6 31.8 Santa Cruz ABAPO 20 11.7 3.3 20.1 ASCENSION DE GUARAYOS 20 10.7 2.3 19.1 CAMIRI 20 17.1 6.4 27.9 MONTERO 41 16.0 9.0 23.0 PALOMETAS 20 4.3 -1.9 10.5 PUERTO SUAREZ 20 14.3 5.3 23.2 ROBORE 20 15.0 2.8 27.2 SAN IGNACIO 20 15.0 6.5 23.5 SAN JULIAN 20 11.4 3.7 19.1 SANTA CRUZ 199 31.1 27.5 34.7 Total 400 22.1 19.7 24.5 Cochabamba AIQUILE 20 16.5 7.3 25.8 ARBIETO 20 16.1 4.7 27.4 ARQUE 17 13.4 0.3 26.6 BOLIVAR 20 17.9 8.2 27.5 CAPINOTA 20 15.0 2.6 27.4 CHIMORE 20 16.3 3.2 29.4 COCHABAMBA 145 35.3 30.6 40.0 MOROCHATA 20 15.4 7.1 23.6 QUILLACOLLO 38 37.6 27.5 47.7 QUIROGA 20 12.1 0.8 23.5 SIPE SIPE 20 15.0 6.2 23.8 VACAS 20 9.3 0.8 17.8 VILLA TUNARI 20 18.6 8.4 28.8 Total 400 24.5 21.8 27.2 Oruro BELEN DE ANDAMARCA 20 21.4 10.5 32.4 CHIPAYA 20 13.6 3.5 23.6 HUANUNI 30 31.0 18.8 43.1 ORURO 188 35.2 31.4 39.0 SALINAS DE GARCIA MENDOZA 20 26.5 14.2 38.8 TOLEDO 20 27.1 13.8 40.5 Total 298 31.3 28.2 34.4 Chuquisaca ALCALA 20 5.0 -0.9 10.9 AZURDUY 20 11.4 3.7 19.1 LAS CARRERAS 20 22.9 9.1 36.6 MACHARETI 20 9.3 2.3 16.2 MENDOZA 20 10.7 3.9 17.5 MONTEAGUDO 20 10.0 2.8 17.2 POROMA 20 26.4 13.4 39.5 PRESTO 20 8.6 1.9 15.2 SAN LUCAS 20 28.6 13.7 43.4 SUCRE 100 23.6 19.0 28.2 ZUDAÑEZ 20 15.0 6.7 23.3 Total 300 17.7 15.2 20.3 Potosí ATOCHA 20 22.1 10.2 34.1 LLICA 20 21.4 11.1 31.7 OTAVI 20 22.1 11.2 33.1 PORCO 20 31.4 16.3 46.6 POTOSI 60 30.5 23.6 37.4 SAN AGUSTIN 20 19.3 7.4 31.2 SAN PABLO DE LIPEZ 20 25.7 13.1 38.3 THOLAPAMPA 20 17.9 6.6 29.1 Special Report: The DDPC Program 27 Contents
NMean 95% Confidence Interval of Corruption for Mean Victimization scale TINQUIPAYA 20 30.0 17.9 42.1 TUPIZA 20 12.9 3.9 21.8 UYUNI 20 37.9 24.8 50.9 VILLAZON 20 29.8 13.8 45.7 YURA 20 25.0 12.0 38.0 Total 300 25.8 22.8 28.8 Pando BELLA VISTA 20 23.5 10.3 36.7 COBIJA 101 18.0 14.0 22.0 CONQUISTA 20 35.5 18.1 52.8 FILADELFIA 18 32.8 17.9 47.7 GONZALO MORENO 20 5.0 -0.4 10.4 NACEBE 20 12.0 4.3 19.7 PORVENIR 19 27.3 14.2 40.4 PUERTO RICO 20 31.4 18.2 44.5 SAN IGNACIO 20 19.5 9.7 29.2 SANTA ROSA 20 23.0 12.5 33.5 SENA 19 20.6 7.7 33.4 Total 297 21.3 18.4 24.1 Tarija BERMEJO 20 29.3 14.0 44.5 PADCAYA 20 18.1 8.6 27.6 PALMAR GRANDE 20 15.7 7.3 24.1 PALOS BLANCOS 20 16.4 6.9 26.0 SACHAPERA 20 12.1 2.9 21.4 SELLA 20 17.9 6.6 29.1 TAIGUATI 20 12.4 4.3 20.4 TARIJA 98 28.3 23.1 33.5 VILLA MONTES 20 13.6 5.3 21.8 YACUIBA 40 22.9 14.3 31.4 Total 298 21.5 18.7 24.3 El Beni GUAYARAMERIN 40 18.6 11.0 26.2 LORETO 20 20.7 10.2 31.2 MAGDALENA 20 12.9 3.4 22.3 RIBERALTA 60 16.9 11.6 22.2 RURRENABAQUE 20 27.9 19.3 36.4 SAN BORJA 20 19.3 10.3 28.3 SAN IGNACIO 20 12.1 2.6 21.7 SANTA ANA DE YACUMA 20 12.1 4.5 19.7 SANTA ROSA 20 23.6 14.0 33.1 TRINIDAD 60 18.8 12.6 25.1 Total 300 18.2 15.8 20.6 Special Report: The DDPC Program 28 Contents We can also examine corruption within the DDPC samples at the level of the municipality. To minimize the width of the confidence intervals, in this table we have combined all three years of surveys. This provides a more reliable measure than any one year would provide. Figure 21 shows the results. Yocala is the standout for low levels of corruption. The next three municipalities in the figure are lower than the remaining five, but not significantly lower.
Corruption Experience, DDPC Samples
Confidence Intervals by Sample
Combination of 1998, 1999 and 2000 Samples 40%
30%
20%
95% CI Corruption index 10% N = 200 200 200 295 299 198 298 300 297 Y C C C LLALLAGUAMIZQ P PUCAR P OC HAL O URA AT UN BIJA AL A AT L HU UE CAMA L AP A A A A AR NI TA A DE CA Y A
RA NGA
Figure 21 Corruption Experience, DDPC Samples: Confidence Intervals by Sample, Combination of 1998, 1999 and 2000 Samples
Chapter V: Ethnicity and Democracy
This chapter in the main report analyzes the impact of ethnicity on democratic beliefs. Patterns found in the national sample are reproduced in the DDPC samples. The task in this special report is to show in what ways the DDPC samples differ from the national samples. Once again, in order to maximize sample size, and thus minimize confidence interval size, the three years of results are used here. Figure 22 shows the results. The samples differ in that the national sample contains a much larger proportion of respondents who identify as white, whereas the DDPC samples have a larger percentage identifying as Indian. Special Report: The DDPC Program 29 Contents
Ethnic Self-Identification by Sample: All Years 100%
10 24 25 80%
13
60% 61 56
Negra 40% Chola
Indígena 20% Blanca
0% Mestiza National DDPC
Figure 22 Ethnic Self-Identification by Sample: All Years
The other major marker of ethnicity used in the study was dress. Figure 23 shows the comparison of the national samples vs. the DDPC samples on this variable. The proportion of the respondents in the DDPC samples who wear Indian dress far exceeds that of the nation as a whole. Special Report: The DDPC Program 30 Contents
Indian vs. Western Dress: National vs. DDPC Samples 100% 10 25
90 80%
75
60%
40%
20% Indian dress
0% Western dress National DDPC
Sig. < .001 Figure 23 Indian vs. Western Dress: National vs. DDPC Samples
Chapter VI: Support for Anti-Democratic Measures
We measured support for anti-democratic measures by first looking at conditions under which a coup would be justified. The individual items were summed to form an overall scale of justification for a coup. Comparisons among the samples are found in Figure 24. As can be seen, there is no significant difference among the samples. Special Report: The DDPC Program 31 Contents
Justification for a Coup: Composite Index
Confidence Intervals by Sample 34%
32%
30%
28%
26% 95% CI Mean index
24%
22% N = 2852 2912 572 845 854 National 1998 Nation DDPC D DD DPC PC a 199 1 2 l 9 0 2 9 0 000 8 9 0
Figure 24 Justification for a Coup: Composite Index, Confidence Intervals by Sample We also asked if the protests in April of 2000 justified a coup. As can be seen in Figure 25, there is no significant difference between the 2000 national sample and the DDPC 2000 sample. Special Report: The DDPC Program 32 Contents
Coup Justified Because of April 2000 Protests:
Confidence Intervals by Sample 38%
36%
34%
32% 95% CI Coup justified 30%
28% N = 2724 795 Na DDP ti o n C a 2000 l 20 0 0
Figure 25 Coup Justified Because of April 2000 Protests: Confidence Intervals by Sample Special Report: The DDPC Program 33 Contents The survey also included a series of items on support for civil disobedience. The scale ranged from 1-10, with 10 being the strongest approval score. Figure 26 shows the results.
Approval of Civil Disobedience: Sample Comparison 5.0
4.8 4.5 4.5 4.4 4.0 4.0 4.0
3.5 National 1998
3.0 National 2000 3.0
2.7 2.7 2.5 2.6 DDPC 1998 2.6 2.5 2.6 2.5 2.6 2.5 2.5 2.5 2.4
Mean approval, scale 1-10 2.2 DDPC 1999 2.0 2.1
1.5 DDPC 2000 Block str Invade Ta Overthrow government k e o prope v e e e r bu ts rty ild ings
Figure 26 Approval of Civil Disobedience: Sample Comparison
Chapter VII: The Rule of Law and Democracy
Support for the rule of law was measured by a four-item series, AOJ10-13. In the main report, we compared support in Bolivia to other countries in Latin America on these items. The items were not included in the 1998 survey, but were included in the 2000 survey as well as the 1999 DDPC sample. Figure 27 shows the results. For the question on breaking or obeying the rules (AOJ10), there are no differences between the DDPC samples and the population as a whole. But for the other three items, some differences emerge. For example, in the DDPC 1999 sample the support for the need for a warrant (AOJ11) is significantly higher than for the national population. By 2000, however, the DDPC sample had declined on this item, placing it at essentially the same level as the national sample. There is also a decline between 1999 and 2000 on the liberty vs. order question (AOJ12), but the difference is not significant. Finally, the item measuring sense of security (AOJ13) is higher in the DDPC 1999 and 2000 surveys than the national 2000 survey, and also increases between 1999 and 2000 to a level that is just on the line of being a significant increase between 1999 and 2000. Special Report: The DDPC Program 34 Contents
Support for the Rule of Law:
Confidence Intervals by Sample 80%
70%
Rules 60%
Warrant 95% CI 50%
Liberty 40%
30% Security N = 3006 287229122980895 823 851 888 900 859 865 893 N DDP DDPC ation C a 1999 l 200 2 000 0
Figure 27 Support for the Rule of Law: Confidence Intervals by Sample
Crime victimization (AOJ3) was examined in the main report for the 2000 sample. Here we can compare all of the samples. Figure 28 shows the results. Victimization declined slightly between 1998 and 2000 in the national sample, but the drop is not significant. There is also some variation within the three DDPC samples, but none of that variation is significant, and variation is within the national parameters. Special Report: The DDPC Program 35 Contents
Crime Victimization:
Confidence Intervals by Sample 28%
26%
24%
22%
20% 95% CI Crime victim
18%
16% N = 2953 2990 599 894 897 National 1998 Nation DDPC D DD DPC PC a 199 1 2 l 9 0 2 9 0 000 8 9 0
Figure 28 Crime Victimization: Confidence Intervals by Sample Special Report: The DDPC Program 36 Contents In the main report, a graph is shown indicating the extent to which respondents had heard of the new criminal law code. The comparison there was among the DDPC samples alone, since this item was not asked in the 1998 national sample. Here we analyze the responses to AOJ8 for all three samples (national 2000, and DDPC 1999 and 2000). Figure 29 shows the results. The initial findings of a very sharp increase in knowledge between 1999 and 2000 is confirmed in this chart, a difference that is clearly statistically significant. In addition, we note that the DDPC samples, which are more rural and less well educated than the national population, have a lower level of awareness than the nation as a whole. In the main report, we showed that the urban/rural distinction and education had a major impact on awareness of the new penal code. Finally, the national level of awareness is 49%.
Heard About the New Criminal Code:
Confidence Intervals by Sample 60%
50%
40%
30% 95% CI AOJ8R
20%
10% N = 2950 872 867 National 2000 DDPC DD PC 199 2 0 0 9 0
Figure 29 Heard about the New Criminal Code: Confidence Intervals by Sample
We also asked about support for the citizen judge provision of the code (AOJ9). Figure 30 shows the results. As can be seen, support is very far into the positive end of the 0-100 continuum. Support in 1999 in the DDPC was lower than in 2000. The national sample of 2000 is higher than the 19998 DDPC, but not significantly more so than the 2000 DDPC sample. Special Report: The DDPC Program 37 Contents
Support for Citizen Judges:
Confidence Intervals by Sample 72
70
68
66
64
95% CI support for citizen judges (0-100) 62 N = 2727 796 780 Na D DDPC 2000 DPC tio n al 20 1 9 9 0 9 0
Figure 30 Support for Citizen Judges: Confidence Intervals by Sample We also carried out an evaluation of citizen perception of the police. Question AOJ1 asks about how difficult or easy it is to report a crime. It is here that we see some major progress in the DDPC areas, as is shown in Figure 31. In 1998, the DDPC sample was indistinguishable from the national sample, and in 2000, although the national sample increased in ease of reporting, the difference was not significant. The DDPC samples, however, have shown steady and significant increases in ease of reporting a crime. Special Report: The DDPC Program 38 Contents
How Easy Is it to Report a Crime?
Confidence Intervals by Sample 66
64
62
60
58
56
54
95% CI How Easy Is it to Report a Crime? 52 N = 2885 2917 585 873 872 Na Na D DDPC 1999 DDP DPC t ti io o n n C al 19 a 1 2000 l 20 9 9 9 0 8 8 0
Figure 31 How Easy Is It to Report a Crime? Confidence Intervals by Sample
We also asked about treatment in the court system (AOJ2). Figure 32 shows the result. The higher the result, the more likely the respondent was to feel justly treated. As can be seen, the average for all samples is well below 50. The DDPC samples have been consistently higher than the national sample data since 1999, but the difference is not significant. Special Report: The DDPC Program 39 Contents
Treatment in Courts:
Confidence Intervals by Sample 42
40
38
36
34
95% CI Treatment in Courts 32
30 N = 2397 2780 561 790 851 Na Na D DDPC 1999 DDP DPC t ti io o n n C al 19 a 1 2000 l 20 9 9 9 0 8 8 0
Figure 32 Treatment in Courts: Confidence Intervals by Sample Reporting of crime was found not to be common in the national samples. But, as we can see from Figure 33, it is far more common in the DDPC areas. Special Report: The DDPC Program 40 Contents
Reporting of Crime:
Confidence Intervals by Sample 70%
60%
50%
40% 95% CI Reported Crime
30% N = 665 915 119 224 245 Na Na DDP D DD DP tional 19 tion P C C C a l 199 1 2 2 9 0 0 9 0 9 0 8 9 0 8 0
Reporting to police and community authority summed as one. Figure 33 Reporting of Crime: Confidence Intervals by Sample Special Report: The DDPC Program 41 Contents Treatment by the public prosecutor (AOJ7) was also measured. Figure 34 shows the results. The DDPC samples have a more favorable view than the nation as a whole, although this difference has been shrinking in the more recent samples.
Treatment by the Public Ministry:
Confidence Intervals by Sample 70
60
50 95% CI Treatment by Public Ministry 40 N = 1147 1231 210 301 267 Na Na D DDPC 1999 DDP DPC t ti io o n n C al 19 a 1 2000 l 20 9 9 9 0 8 8 0
Figure 34 Treatment by the Public Ministry: Confidence Intervals by Sample
Chapter VIII: Local Government
The main report covers this ground, and therefore there are no additional comparisons to be reported here.
Chapter IX. Women and Democracy
The main report contains a number of comparisons of males versus females. Chapter IX in that report examines a number of issues related to discrimination against women. In this section we examine these differences at the level of the DDPC samples.
We first analyze perception of discrimination against women (DM1). Figure 35 shows the comparison of the various samples. As can be seen, the 1998 and 1999 DDPC samples were somewhat higher in terms of perception of discrimination against women Special Report: The DDPC Program 42 Contents than the national population, but by 2000, the national and DDPC samples did not diverge significantly.
Believe That There is Discrimination Against Women
Confidence Intervals by Sample 92%
90%
88%
86%
95% CI 84%
82%
80%
78% N = 2827 2947 593 878 888 National 1998 Nation DDPC D DD DPC PC a 199 1 2 l 9 0 2 9 0 000 8 9 0
Figure 35 Believe That There Is Discrimination against Women: Confidence Intervals by Sample
The next item in this series (DM2) asks how serious is the discrimination. This item was not asked to those who said that there was no discrimination against women, so the results focus exclusively on those who think that there is discrimination. Figure 36 shows the results. The national samples and the first two DDPC samples do not vary significantly on this score. However, the 1999 DDPC scores show an increase, and the 2000 DDPC is significantly higher than the other samples. Special Report: The DDPC Program 43 Contents
How Serious is Discriminaton Against Women?
Confidence Intervals by Sample 66
64
62
60
58 95% CI 56
54
52 50 N = 2304 2409 514 767 731 Na Na D DDPC 1999 DDP DPC t ti io o n n C al 19 a 1 2000 l 20 9 9 9 0 8 8 0
Figure 36 How Serious Is Discrimination against Women? Confidence Intervals by Sample
The third item in the series (DM3) asks if women have equal opportunities to obtain employment as do men. Figure 37 shows the results. Note that the national results are slightly different from those reported in the main report (here there is no significant difference between 1998 and 2000) because in the analysis being presented in this chapter a portion of the 1998 sample (in Pando) is counted as part of the DDPC sample. Although there is some fluctuation among the DDPC samples, they do not differ significantly from the national samples. Only the 1999 DDPC sample is significantly lower than the other two DDPC samples, but even that sample is within the national sample parameters. Special Report: The DDPC Program 44 Contents
Do Women Have Equal Employment Opportunities With Men?
Confidence Intervals by Sample 70%
60%
50%
95% CI Women have equal employment 40% N = 2839 2951 592 885 885 National 1998 Nation DDPC D DD DPC PC a 199 1 2 l 9 0 2 9 0 000 8 9 0
Figure 37 Do Women Have Equal Employment Opportunities with Men? Confidence Intervals by Sample
Conclusions
This special report has reviewed many of the variables analyzed in the main report, but the focus here is on the DDPC samples. In the main report, some of the DDPC results were covered, and those findings are not replicated here. For the most part, the DDPC samples do not appear to be substantially different from the nation as a whole on the key democracy measures under study.