Two-Century Trend Analysis Finds that Wealth Falls Short in Safeguarding Democracy

Education is Critical

Abstract

Wealth (measured by economic growth) is considered a key driver of democratization in the context of modernization theory. Anthropologically, however, wealth is a by-product of early population density (city-states) and specialized industry, but one that created socio-economic stratification. Today, such divisions have led to gross inequality resulting from disenfranchisement from society. Education, also associated with democracy, is a stronger candidate than wealth as a key driver and supported of democracy. The project includes summary graphics by global regions and interactive trend analysis plots.

Summary of Research Questions and Findings

Lipset (1959) asserts that modernization is influenced by four key factors: wealth, industrialization, education, and urbanization. In his 1959 article, he presents indices from various regions to test this hypothesis. From a Marxist perspective, wealth is seen as a driver of inequality. Conversely, an anthropological lens suggests that inequality originates from factors such as population density (urbanization) and craft specialization (a precursor to industrialization). Egalitarian societies, on the other hand, are characterized by minimal social structure, often limited to a gendered division of labor linked to biological differences between males and females. Thus, arguments emphasizing wealth, industrialization, and urbanization may not strongly support the promotion of democracy. Education, however, has been shown to correlate with greater participation in democracy and warrants further exploration in this context.

My analysis suggests that economic growth, a marker of wealth, while seen as fueling modernization likely leads to social stratification and inequality, leaving some citizens marginalized. This analysis highlights that better access to education correlates with increased democratic participation. When exploring the link between educational equity and participatory democracy, it becomes clear that wealth alone doesn’t ensure stability in these factors. This suggests that while wealth plays a role, it’s just one piece of the puzzle driving modernization and democratization.

The Data

The vDem dataset was utilized to investigate the potential influence of education within modernization theory.

The participatory democracy index (v2x_partipdem) serves as the chosen metric for this analysis, as it evaluates the active involvement of citizens in all facets of political processes, encompassing both electoral and non-electoral spheres. This index distinguishes between direct and electoral democracy, where the latter entails the delegation of authority to representatives rather than direct citizen voting. While direct governance by citizens epitomizes the purest form of democracy, its realization is not always feasible. Therefore, the inclusion of electoral democracy provides valuable insights into the varying degrees of democratic practices across countries and regions. The index operates on a scale ranging from 0 to 1, where 0 signifies minimal participation and 1 represents maximal citizen involvement in political processes.

The education variables analyzed were the average years of education among citizens 15 or older (e_peaveduc) and inequality in education among citizens 15 or older (e_peedgini). Education 15+ uses interpolation for missing data.

Context variables for analysis include region, country, year, and population.

Code

library(vdemdata); library(tidyverse); library(janitor)
library(forcats); library(colorBlindness); library(plotly); library(grateful)

#get data from vdem
dem_data <- vdem |>
  select(
    country = country_name, 
    vDemCtryId = country_id,
    year,
    region = e_regionpol_6C,
    parDem = v2x_partipdem, #To what extent is the ideal of participatory democracy achieved
    eduPlus15 = e_peaveduc, #What is the average years of education among citizens 15 and older
    eduUnequal = e_peedgini, #how unequal is education level achieved by population aged 15 and older
    pop = e_pop #size of population
    ) |>
  mutate(region = case_match(region, 
     1 ~ "Eastern Europe", 
     2 ~ "Latin America",  
     3 ~ "Middle East",   
     4 ~ "Africa", 
     5 ~ "The West", 
     6 ~ "Asia")
  )

Inverse Relationship between Participatory Democracy and Educational Inequality

This section presents a comparative analysis of data spanning the 19th and 20th centuries, showcased through two separate interactive plots, each focusing on a 10-year date range around the turn of each millennium. In both plots, the variable of interest concerning modernization (education) is plotted along the x-axis, while the measure of democracy (participation) is depicted on the y-axis. Employing the loess (Locally Weighted Scatterplot Smoothing) method, a trend line is applied to visualize the relationship between variables, which is particularly beneficial when dealing with noisy or highly variable data. Unlike linear regression, loess accommodates curvilinear relationships and facilitates the identification of trends over time. Additionally, the plots are enhanced with annotations indicating the mean for the respective decades under scrutiny.

The mean values for the five years before and after the turn of the 19th century shows a clear linear trend, the higher the participatory democracy index value, the lower the educational inequality value. Barbados, Spain, and Japan have much lower than expected participatory democracy index score than expected given their low educational inequality values. Spain experienced tremendous turmoil across multiple wars during the 19th century that resulted in the restoration of the monarchy at the turn of the century. Likewise, this was also a period of war for the Japanese, contesting control over Korea. Barbados was undergoing radical change arising from a rebellion of enslaved peoples followed by a period of emancipation. Such turmoil in these countries would disrupt normal educational opportunities and may explain their outlying values.

Code

#filter and plot span 1
demDataSummary1 <- dem_data  |> 
  filter(year %in% 1895:1905) |>
  group_by(country, region)  |> 
  summarize( 
    parDem = mean(parDem, na.rm = TRUE), 
    eduUnequal = mean(eduUnequal, na.rm = TRUE)
  ) |> 
  arrange(desc(parDem))

#plot
nineteenth <- ggplot(demDataSummary1, aes(x = eduUnequal, y = parDem)) + 
  geom_point(aes(color = region)) + 
  geom_smooth(method = "loess", linewidth = 1, color="black") +
  labs(
    x= "Education Inquality", 
    y = "Particapatory Democracy (Percentage)",
    title = "Education and democracy, 1895 - 1905", 
    caption = "Source: V-Dem Institute", 
    color = "Region"
    ) +
  scale_y_continuous(labels = scales::percent) +
  scale_color_viridis_d() +
  theme(legend.title=element_blank()) +
  aes(label=country)

#annotate plot
nineteenth <- nineteenth +
  geom_hline(yintercept=.1607, linetype="dashed", color = "black", size = 1) +
  annotate("text", x = 73, y = .19, label = "Average Participatory Democracy Score", fontface=2)

#interactive
ggplotly(nineteenth, tooltip = 'all')

The mean values for the five years before and after the turn of the 20th century reveal a less stable relationship. Hovering over the values at the far right of the graph, three sub-Saharan African countries (Burkina Faso, Niger, and Mali) show higher than expected participatory democracy scores given their lower education inequality scores. Two outliers with the lowest participatory democracy scores and least educational equality are Afghanistan and Somalia, both globally notorious for infringing upon women’s rights. While both these countries have achieved limited women’s representation (~23%), the status of women is dire. For instance, in Somalia, twice as many males are literate as females ([https://data.unwomen.org/country/somalia]), and 28.4% of Afghan women are married before the age of 18 ([https://data.unwomen.org/country/afghanistan]). In contrast to expectations of education driving democracy, the literacy rate among Afghan women is 43% compared to 29.8% among men. Afghanistan has been in political flux for many years and recently suffered tremendous setbacks for women’s rights and access to education with the reversion to Taliban rule, rendering it a less stable data point. Many of the countries showing extreme values for low participatory democracy scores and low educational inequality scores are far right states (e.g., the Stans).

Code

#filter and plot span 2
demDataSummary2 <- dem_data  |> 
  filter(year %in% 1995:2005) |>
  group_by(country, region)  |> 
  summarize( 
    parDem = mean(parDem, na.rm = TRUE), 
    eduUnequal = mean(eduUnequal, na.rm = TRUE)
  ) |> 
  arrange(desc(parDem))
#plot
twentieth <- ggplot(demDataSummary2, aes(x = eduUnequal, y = parDem)) + 
  geom_point(aes(color = region)) + 
  geom_smooth(method = "loess", linewidth = 1, color="black") + 
  labs(
    x= "Education Inquality", 
    y = "Particapatory Democracy Score",
    title = "Education and democracy, 1995 - 2005", 
    caption = "Source: V-Dem Institute", 
    color = "Region"
    ) +
  scale_y_continuous(labels = scales::percent) +
  scale_color_viridis_d() +
  theme(legend.title=element_blank()) +
  aes(label=country)

#annotate plot
twentieth <- twentieth +
  geom_hline(yintercept=.1607, linetype="dashed", color = "black", size = 1) +
  annotate("text", x = 68, y = .29, label = "Average Participatory Democracy Score", fontface=2)

#interactive
ggplotly(twentieth, tooltip = 'all')

Time Series Visualization Reveals that Wealth is Insufficient in Buffering External Threats to Democracy

Summary data plots for the variables indicated that the West and Latin America had the strongest indicators of democracy (see Appendix). Several Latin American countries fall within the high democracy/low education inequality area, such as Uruguay and Chile (country level data not shown but area available via copying the code within this document). Chile is among the wealthiest of Latin American countries (and other research suggests affluence as a promoter of democracy), but Uruguay is not. Uruguay has legislated against discrimination, and women and men achieve similar literacy rates and access to pensions ([https://data.unwomen.org/country/uruguay]).

To examine the association between wealth and democracy, the most affluent countries from each region were selected: Mexico and Brazil (Latin America) and Germany and the USA (The West). The black trend line shows a common trajectory across fall four countries, with a rapid increase in the participatory democracy index following the second World War. The participatory democracy indicators for all four countries increase over time, but recent years are characterized by declines that coincide with the global COVID-19 pandemic. These declines have coincided with the rise of populism and the increased influence exerted by far-right individuals within conservative political factions, as well as the emergence and consolidation of far-right political parties. Thus, even in when looking at the wealthiest countries, wealth being a putatively driver of democracy, we see that is not always the case. Even relatively and absolutely wealthy nations suffer declines in democracy when there is greater educational inequality.

Code

#filter data to limit to a few countries (3-4)
countries = c('United States of America', 'Germany', 'Mexico', 'Brazil')

#line chart with viridis color-blind friendly package
dem_data_viridis <- 
  dem_data |>
  filter(country %in% countries) |>
  ggplot(aes(x = year, y = parDem, color = country)) +
  geom_line(linewidth = 1) + 
  labs(
    x = "Year", 
    y = "Particapatory Democracy Index", 
    title = 'Participatory Democracy Index: Affluent Western and Latin American Countries', 
    caption = "Source: V-Dem Institute", 
    color = "Country"
  ) +
  scale_color_viridis_d() +
  theme_classic()
dem_data_viridis = dem_data_viridis + stat_smooth(color = "black", fill = "black",method = "loess")

#interactive
ggplotly(dem_data_viridis, tooltip = 'all')

Bibliography (grateful created the list of R packages used)

Lipset, SM. (1959). Some Social Requisites of Democracy: Economic Development and Political Legitimacy. Am Pol Sci Review 53:69-105.

Code

cite_packages(citation.style = "peerj", output = "table", out.dir = ".", include.RStudio = TRUE, passive.voice=FALSE)

         Package Version                                       Citation
1           base   4.3.1                                          @base
2 colorBlindness   0.1.9                                @colorBlindness
3        janitor   2.2.0                                       @janitor
4         plotly  4.10.2                                        @plotly
5      rmarkdown    2.24 @rmarkdown2018; @rmarkdown2020; @rmarkdown2023
6         scales   1.2.1                                        @scales
7      tidyverse   2.0.0                                     @tidyverse
8       vdemdata    13.0                                      @vdemdata

Appendix: Bar Charts for Democracy Indices

Bar charts offer a comprehensive visual overview of the data spanning from 2000 onwards. Notably, a striking relationship emerges between the education variables, revealing an inverse correlation between years in education and equality in access to education, as depicted vividly in the bar charts. Notably, the Western regions stand out with the highest number of years attained in education, coupled with minimal educational inequality, indicative of robust participatory democracy. Conversely, Africa exhibits the lowest number of education years alongside pronounced educational disparities. Furthermore, the Middle East portrays the lowest values in participatory democracy, underscoring regional disparities in political engagement and representation. Data since 2000 are grouped by region and summarized by mean.

Code

bar_chart_data <-
  dem_data |> 
  filter(year >= 2000) |>
  group_by(region)  |> #group by region
  summarize(           # summarize vars (by region, above)
    parDem = mean(parDem, na.rm = TRUE), 
    eduPlus15 = mean(eduPlus15, na.rm = TRUE), 
    eduUnequal = mean(eduUnequal, na.rm = TRUE)
  ) |> 
  arrange(parDem)

Code

#create factors
dem_data$region <- as.factor(dem_data$region)

#create bar chart for participatory democracy
ggplot(bar_chart_data, aes(x = fct_reorder(region,
                         parDem), y = parDem)) + 
  geom_col(fill = "#005699") + #NSF blue
  labs(
    x = "Region", 
    y = "Participatory Democracy", 
    title = "Participatory democracy by region, 2000-2022", 
    caption = "Source: V-Dem Institute"
    ) +
  theme_classic()

Code

#create bar chart
ggplot(bar_chart_data, aes(x = fct_reorder(region,
                         eduUnequal), y = eduUnequal)) + 
  geom_col(fill = "#005699") + #NSF blue
  labs(
    x = "Region", 
    y = "Equal access to education (population 15+)", 
    title = "Education Equality (population 15+): 2000-2022", 
    caption = "Source: V-Dem Institute"
    ) +
  theme_classic()