The objective of this study was to understand the organizing principles of the social representations of democracy in higher education students in Colombia, and their anchoring in sociodemographic and territorial variables. In three periods of time (2010, 2012 and 2017) students at university and technical/technological careers (total number: 12,171) from the 32 capital cities of each department in Colombia responded to a Scale of Social Representations of Democracy (ERSD) and to a list of sociodemographic variables. In a macro-level analysis, departmental averages of Democracy Balance were crossed with macro indicators of crime, extreme poverty, life expectancy and indigenous population percentage. Results show a seven dimensions factorial structure without significant relationships with age, gender or social class. At a macro level, regions with a high proportion of indigenous population showed a more positive evaluation of democracy over time, while higher crime rates are associated with a more negative evaluation of democracy.
Democracy, forms of political participation, social mobilizations in the decade 2010-2020, somehow interrupted by the isolation measures due to Covid19, generates an important academic reflection in Latin America, expressed in a relatively high academic production from different perspectives in Mexico (
This topic, democracy, has also been approached from the theory of social representations (
In the case of Colombia, there is also a body of works based on the theory of social representations about democracy, most of them taking young people, high school or university students as samples or participants and using qualitative techniques such as open-ended questions, group interviews or free association of words. Thus,
Elsewhere (
In this framework, and taking into account the armed violence history that Colombian society has suffered for decades, violence that is still in force, and that most studies on social representations of democracy in Colombia have been carried out with relatively small samples, transversally and in a single municipality or department, the objective of this study aims to show a quantitative, broad territorial scope, and longitudinal overview of social representations of democracy in the young Colombian population.
The survey was applied to university and technical and technological career students in three different periods: 2010, 2012 and 2017, as shown in
Sociodemographic and Economic Status of samples per year:
Year:
2010
2012
2017
N
4073
3891
4278
Mean Age (SD)
21.7 (4.4)
22.3 (5.5)
22.06 (6.01)
% women
62.0
70.5
58.5
% singles
85.2
80.7
82.8
Low social class
60.4
62.5
61.6
% university students (versus technical careers)
47.2
51.4
60.4
Participants make up a non-random sample and responded to a survey composed of several sections about the perception of the social fabric. For the purposes of this survey, the Social Representations of Democracy Scale (ERSD,
On the other hand, the survey included questions on sociodemographic variables: age, gender, marital status, socioeconomic status, type of studies (university or technical), major and semester.
This study also presents macro-level analyses based on the average calculation of the Democracy Balance for the sample of each department (32 cities, including Bogotá) and its cross-checking with a series of macro indicators of Colombian society:
Extreme Poverty Population Percentage: data for 23 departments of Colombia (
Life expectancy (
Crime rates based on data on arrests for crimes at the departmental level, from the Colombian National Police, weighted per 100,000 inhabitants (
Indigenous Population Percentage at departmental level. From the 2005 population census (
First, the ERSD dimensional analysis was carried out by means of confirmatory and exploratory factor analysis, in addition to internal reliability calculation according to factors found. Next, the difference in Democracy Balance was analyzed, according to the participants sociodemographic level, age and gender. Following, Democracy Balance is projected in a georeferenced manner at the departmental level, and Spearman correlations are established between the Democracy Balance departmental average with macro indicators.
Implementation of each of the survey waves (2010, 2012 and 2017) followed specific instructions requested to the
The authors declare that they have no conflict of interest in this research.
From the initial study by
Expected ERSD dimensional model (Ruiz-Pérez et al., 2016).
Given these results, which indicate a low fit to the model, an exploratory factor analysis of the scale was carried out, with parallel analysis and maximum likelihood extraction method and promax rotation. Coefficients that support the exploratory analysis were: KMO: 0.925 and Bartlett's test of sphericity χ² (190): 23.15, p <.001. Regarding coefficients of the resulting factorial model, a TLI value was obtained: 0.974, BIC: 159, RMSEA: 0.032 (90% CI 0.030 – 0.034), with χ²/df: 816/79=11.49, p <.001.
The factor solution obtained is seven factors, which explain 50.9% of the variance. Regarding its composition, it largely corresponds to the structure found by
SRDS Exploratory Factor Analysis and Internal Reliability Note: DV: Democracy values, EL: Elections and Laws, FS: Freedom of Speech, BD: Benefits of Democracy, PP: Political Power, DD: Distrust in Democracy, CD: Corruption of Democracy.
Item
DV
EL
FS
BD
PP
DD
CD
Uniqueness
DEMO13
0.938
0.293
DEMO12
0.866
0.318
DEMO11
0.667
0.423
DEMO15
0.347
0.306
0.524
DEMO02
0.937
0.337
DEMO01
0.694
0.466
DEMO03
0.606
0.484
DEMO04
0.689
DEMO20
0.785
0.405
DEMO18
0.634
0.479
DEMO17
0.395
0.639
DEMO08
0.831
0.306
DEMO09
0.719
0.347
DEMO05
0.736
0.622
DEMO06
0.667
0.432
DEMO07
0.387
0.395
0.524
DEMO16
0.657
0.616
DEMO14
0.641
0.604
DEMO19
0.344
0.390
0.621
DEMO10
0.450
0.690
Eighenvalue
2.59
1.73
1.61
1.59
1.14
1.06
0.47
% Variance
12.95
8.62
8.07
7.96
5.69
5.28
2.34
Cronbach α
0.837
0.751
0.678
0.757
0.624
0.55
ω
0.84
0.761
0.689
0.77
0.632
0.563
Mean
2.81
2.85
2.73
2.57
2.63
2.44
3.21
SD
0.74
0.75
0.70
0.75
0.71
0.70
0.98
Thus, the first factor called Democracy Values, covers references to values such as equity, loyalty, equality, justice and freedom. The second factor, Elections and Laws, refers to democracy as a system of elections, rights and duties translated into laws based on the Constitution, and values to organize the future of society. For its part, the third dimension highlights expressive aspects of democracy, such as the expression of freedom and diversity that translates into representations in parliament. The fourth factor refers to Benefits of Democracy, such as social harmony, well-being, tranquility, autonomy, while the Political Power factor includes items that refer to politics and the results of the work of politicians. Finally, the last two factors bring together negative aspects of democracy, such as that democracy does not really exist or is a form of control to maintain inequality (Distrust in Democracy factor), or is seriously affected by corruption (last factor).
One-way ANOVAs were calculated to cross factorial scores saved from the preceding analysis, with gender and age grouped into intervals. In relation to gender, the Kolmorov-Smirnov test was significant for all factorial scores indicating with this, that such scores are not normally distributed: KS, between 0.014 for the seventh factor, with p <.05, and 0.044, with p <.001 for the other six factors. On the other hand, Levene's statistic to compare variances homogeneity was not significant for any of the factors. And as for the means comparison,
In relation to Age, there is a tendency for older participants to score higher in the Democracy Values dimension, participants between 31 and 40 years old, in “Political Power”, and young people between 21 and 30 years old, as well as people over 41 years old in Distrust in Democracy. However, at the effect size level, differences between age groups on democracy are marginal (see
At the social class level, a low effect size ( η²p: 0.007) is found in the corruption factor, with higher scores in people from a high social class (Mean: 0.1668, SD: 0.958, n: 531), followed by the middle class (Mean: 0.1013, SD: 0.967, n: 3557), and the lower class (Mean: -0.0524, SD: 1.006, n: 5913), for an F(2, 1464): 34.04, p <.001. In other words, lower social class participants have a more positive social representation of democracy, considering that it is less affected by corruption than other social classes.
Democracy Dimensions and Gender: ANOVA Note: SD: Standard Deviation; EE: Error estimation; η ²p: Effect seize Squared partial Eta. + p <, 010; * p <.05 ** p <.01
Women
6484
0.01139
1,004
0.0125
1.44 ns
0.001
Men
3721
-0.01323
0.993
0.0163
Women
6484
0.02562
0.991
0.0123
7.94**
0.001
Men
3721
-0.03250
1,009
0.0165
Women
6484
0.00729
0.995
0.0124
0.72ns
0.001
Men
3721
-0.01014
1,005
0.0165
Women
6484
-0.01080
0.992
0.0123
1.47 ns
0.001
Men
3721
0.01426
1,014
0.0166
Women
6484
0.01640
1,001
0.0124
3.57+
0.001
Men
3721
-0.02249
1,001
0.0164
Women
6484
-0.00561
1,008
0.0125
0.55ns
0.001
Men
3721
0.00958
0.983
0.0161
Women
6484
0.02669
0.983
0.0122
5.77*
0.001
Men
3721
-0.02305
1,021
0.0167
Democracy Dimensions and Age. Note: SD: Standard Deviation; EE: Estimation Error, η ²p: Squared partial Eta. ** p <.01 *** p <.001
Democracy Values
14-20
4995
-0.01683
0.967
0.0137
4.13** Levene: 5.25***
0.001
21-30
4427
-0.00132
1,030
0.0155
31-40
470
0.11753
1,036
0.0478
41-60
140
0.18311
0.980
0.0828
Political Power
14-20
4995
-0.04652
0.982
0.0139
7.90*** Levene: 3.78**
0.002
21-30
4427
0.04752
1,010
0.0152
31-40
470
0.07660
1,047
0.0483
41-60
140
-0.04525
1,126
0.0952
Distrust in Democracy
14-20
4995
-0.03083
0.989
0.0140
6.37*** Levene: 1.45 n.s
0.002
21-30
4427
0.03026
1,007
0.0151
31-40
470
-0.05515
1,023
0.0472
41-60
140
0.24297
0.960
0.0811
In relation to territorial anchoring, that is, the projection of the samples average of the departments in Colombia in the Democracy Balance for each period covered in this research,
Democracy Balance Indicator Georeferencing at departmental level in Colombia. Note: darker areas indicate a more positive balance of Democracy SRs.
It must be remembered that the samples from each year are not random and therefore, results are difficult to compare. Despite this, a certain coincidence is found between each period in several regions with a more positive perception of democracy, particularly in the eastern departments of Vaupés, or Vichada. As for the departments with the most negative evaluation (lighter tones), for 2010, western Colombian departments stand out, such as Valle del Cauca and Cauca, areas of presence of various armed groups at the time, including drug trafficking organizations. For 2012, the department with the most negative balance is Tolima, towards the center of the country, with a strong presence in previous years of paramilitary groups. For that year, Colombia was experiencing many expectations with the beginning of peace talks between the government chaired by Juan Manuel Santos and the FARC guerrilla, which could explain that there were fewer areas of very negative perception of democracy. By 2017, these positive expectations of peace had partially transmuted into disappointment, because even though the peace process began, certain FARC fronts withdrew from the peace process, contrary to the agreement signed by the central committee of that organization, and armed violence intensified in several regions of the country, including the appearance and strengthening of the so-called Criminal Gangs (BACRIM), which in reality were paramilitary groups of a mafia nature that were nourished by former paramilitary combatants who had participated in processes demobilization in the years 2003 to 2006.
Correspondingly, relationships between Democracy Balance departmental averages and crime, poverty and ethnic composition macro indicators were analyzed using Spearman correlations (n: between 23 and 32). Results are shown in
Correlations between Democracy Balance and crime, poverty, life expectancy and indigenous communities’ presence indicators, in years 2010, 2012 and 2017. + p <.010; * p <.05 ** p <.01
2010
2012
2017
% Population in extreme poverty (n=23)
0.455*
Life expectancy (n: 32)
-0.341+
-0.494**
Security crime rate (n: 32)
-0.426*
Drug crime rate
-0.421*
Theft rate
-0.421*
Sexual crime rate
0.371*
Crimes against freedom rate
-0.408*
% Indigenous population
0.478**
0.383*
In general, although there are no stable correlations between the indicators over the years, the significant correlations are consistent with the processes that can socially affect democracy perception. Thus, for 2010, higher crime rates are related to a more negative democracy assessment. For 2012 and 2017, a higher percentage of indigenous population is related to a more positive democracy assessment. These regions, with a greater presence of indigenous peoples, have a lower level of socioeconomic development than other areas, including large Colombian cities such as Bogotá or Medellín. Also, these indigenous presence regions have suffered the action of all kinds of armed groups, in relation to attempts to forced recruiting for their youth, plunder of their natural resources, population displacement in relation to the presence of said armed groups, for the purpose of land dispossession, coca leaf crops planting and/or the establishment of drugs transportation and control of corridors in the territory. Finally, the positive relationship between rates of sexual crimes and balance of democracy in 2017 may be a spurious result of the highest recorded rate of sexual crimes in indigenous communities in Colombia, which may be related to the worldview of indigenous peoples regarding the relationship between women and territory (
Social representations of democracy seem to be organized around six dimensions, following the approach of
On the other hand, macro-level correlations confirm what other studies have shown, that is, the negative influence of crime on the valuation of democracy (
On the other hand, the positive relationship between indigenous population rate and a more positive assessment of democracy may reflect the possibility that minority groups in Colombia, such as indigenous peoples, see that democratic mechanisms allow them quotas for representation and political participation that allow them to overcome and change historical conditions of inequality, discrimination and violence suffered, caused by the different violent actors of the armed conflict in Colombia.
Finally, it is worth mentioning some limitations of this research. On the one hand, the Balance of Democracy averages georeferencing at the departmental level masks the fact that such averages generally come from non-random samples of the capitals, that is, they are not samples drawn from the entire department: capital, intermediate cities and/or rural areas. In this sense, georeferenced maps and their usefulness are based on data quality, of which may present biases (