Does consensus democracy improve the quality of government?
The first matter I will address are the variables in the question. This will allow for a more thorough discussion about whether consensus democracy increases the quality of government, because I will refer to the elements in each variable, in my causal analysis.
Save your time!
We can take care of your essay
- Proper editing and formatting
- Free revision, title page, and bibliography
- Flexible prices and money-back guarantee
Lijphart distinguishes between two forms of democratic government systems: consensus democracies and majoritarian democracies. These differ in two ways: the executive-parties dimension and the federal-unitary dimension. The executive-parties dimension is “the arrangement of executive power, the party and electoral systems, and interest groups”. The federal-unitary dimension are the features of democracies which are “commonly associated with the contrast between federalism and unitary government”. The former definition differs between unicameralism and bicameralism. However, Lijphart notes that “the federalist institutions of consensus democracy have little effect on the performance” of a country. I will restrict the scope of this essay to the executive-parties dimension, as a result.
I will define consensus democracy using the criteria outlined by Lijphart. Consensus democracy, according to Lijphart, seeks to maximise the size of ‘decision-making majorities’. The ‘majority rule’ touted by majoritarian democracy is only a ‘minimum requirement’ for consensus democracy, which aims to represent ‘as many people as possible’.
Lijphart says that government ‘for the people’ is government ‘in accordance with the people’s preferences’. Therefore, the ‘quality’ of government can be understood as how effectively a government can represent the ‘preferences’ of its citizens. I have split my analysis of ‘quality of government’ into three sub-sections in order to better represent the broad spread of interests that may relate to the effectiveness of government. These three sub-sections are as follows: a) the relative quality of democracy, b) relative inequality and representation and c) macroeconomic performance of any given democracy.
I will first explain why the quality of democracy is a variable in the quality of government. Dahl’s (1971, pp. 4) definition of democracy includes elections with high levels of ‘public contestation’. Dahl is alluding to the ability to hold competitive elections with several competing parties. In addition, the second feature Dahl alludes to is ‘inclusiveness’. This is the proportion of the population with suffrage. In application to real-world contexts, we can use the EIU Democracy Index. It includes variables which assesses the proportion with which the population politically participates in a democracy, the sufficiency of electoral process and civil liberties. Secondly, the other major measure used in this subsection is the World Bank’s Worldwide Governance measure. It measures the ‘governance’ of nations. This is defined by the World Bank as the ‘traditions and institutions by which authority in a country is exercised’. Both measures can yield more reliable results to determine whether our correlation analysis results in there being a positive, causal relationship between a consensus democracy and the quality of government.
I will now explain why relative inequality and a lack of representation in a democracy is a key component of the quality of government. If a government is institutionally equipped to represent the interests of a section of its citizens, but the percentage of this section is relatively lower compared to percentage size of the rest of the population, this results in a lower score on Dahl’s scale of inclusiveness. Consequently, the quality of government would be considered poor. I will look at a lack of representation from the perspectives of gender (an indicator of social inequality) and economic inequality. To gain some semblance of the effects of gender inequality, I am using the representation of women in parliament and cabinet, between 1990 and 2010, and between 1995 and 2008 respectively. Furthermore, I have included the United Nations Development Programme’s Gender Inequality Index, to further strengthen my claim. In order to summarise economic inequality, I will use the GINI index of inequality and the 90:10, and 80:20 ratios. This indicator takes into account the extremes ends of inequality. The latter indicators are an alternative to a summary statistic which formalises income inequality, as a parameter of central tendency.
I will now explain why I have used ‘macroeconomic variables’ in assessing the quality of government. Economics is the organisation of the resources in the economy. Much of the government’s role is to make decisions on the allocation of resources in the economy, in order to maximise social welfare. Since achieving macroeconomic objectives, such as sustainable economic growth, is in their interest, it should be concerned in the quality of government. When applying this conceptual analysis to data in real-world contexts, I will be using Lijphart’s macroeconomic variables, from the World Bank (2011) dataset, for GDP per capita growth, CPI inflation, the GDP deflator, unemployment and the budget balance.
Using multivariate regression analysis, I will show that there is a statistically significant positive correlation between consensus democracy and the quality of government of the consensus democracy. I follow Lijphart’s method of analysis, using the executive-parties dimension between 1981 and 2010, as the dependent variable. I with control the variable for economic development, by using the HDI of 2010 and the population size. The general line of best fit reveals a statistically less significant correlation compared to Lijphart’s results. The process by which I have yielded these results are seen below. I will now explain the data in more depth.
Unlike Lijphart, I found no statistically significant correlation between consensus democracy and the Worldwide Governance indicator for the quality of government. For example, the p-value seen by the Rule of Law between 1996 and 2009, is 0.05679. The other results have a result that is not in the region to reject the null hypothesis, that is if we were to average the result of the p-values (which come from variables in the quality of government). To average the results, we would merely calculate the mean of the p-values. Even though we used the same method by calculating the correlation coefficients, I have yielded different results compared to Lijphart. There may be issues with the datasets we use, which I will now discuss.
The only country missing from my analysis is Barbados. It is not included in the EIU Democracy Index and the Bahamas. Barbados is not measured in either the Corruption Perception Index or the EIU Democracy Index. There is no statistically significant correlation between consensus democracy and the Corruption Perception Index or Civil Liberties. However, Lijphart claimed that there was a statistically significant positive relationship between consensus democracies and civil liberties, for example.
Furthermore, Lijphart considers the Product Moment Corellation Coefficient from the EIU Democracy Index variables as significant at the 1% level, whereas I viewed them as significantly significant within a 5% level. All the same, my analysis does indicate a statistically significant correlation between consensus democracy and the EIU Democracy Index, which may suggest a correlation between consensus democracy and quality of democracy. In this situation, our results match. I increased the region of rejection of the null hypothesis, because I could account for other democracies which were close to being considered as having a sufficiently strong relationship between consensus democracy and the quality of government, but could not, within a 1% significance level.
The only country not included in the Gender Inequality Index is the Bahamas. For the 90/10 and 80/20 ratios, Mauritius is not included along with Barbados, Iceland, Luxembourg, Malta and the Bahamas. These democracies have populations below 500,000. They are considered to be outliers in the data set, because small changes in the population are more likely to create biases in the results. Furthermore, I have removed Botswana as an extreme outlier for its ‘extremely high inequality’. When analysing the data, using the Gini measure of income inequality, the democracies, referred to as outliers, are not included, except Botswana. This is to ensure that my results are the same as Lijphart’s.
Another point of contention in my analysis is that Lijphart records a statistically significant correlation at the 5% level for the variable: Women’s Cabinet Representation in both 1995 and 2008, whereas my analysis does not find a statistically significant correlation. Lijphart finds a statistically significant correlation at the 1% level for the 90/10 and 80/20 ratios, whereas my analysis only records it at the 5% level. Yet, as synonymous to Lijphart, my results largely map a strong statistically significant correlation for consensus democracy with my representation and inequality indicators.
For the data which analysis the quality of government based on their macroeconomic performance, I will now evince the outliers they contain. In the GDP, inflation and unemployment measures, the countries considered to be outliers are excluded. Additionally, Argentina, South Korea and Uruguay are excluded from the 1981-2009 performance variables because they only democratised in the 1980s. Having experienced inflation above 100%, Israel and Uruguay have also been removed from the CPI Inflation and GDP Deflator as extreme outliers for the dates between 1981 and 2009, 1991 and 2009. Unemployment between 1981 and 2009 suffers severely from missing data, whereas for the unemployment variable, between 1991 and 2009, we have data for all countries except for the outlier countries (from previous paragraphs only), Botswana and India. The budget control includes the outlier countries, because the data produced were not affected by biases. Norway is excluded as an outlier for both measurements, because they had large budget surpluses, in the given period.
The inferences one can draw from Lijphart’s analysis match mostly mine, except for two results. Firstly, Lijphart finds a statistically significant correlation at the 5% level for unemployment, which is contrary to my regression analysis. Secondly, for both the CPI and GDP Deflator between 1991 and 2009, there is a statistically significant correlation at the 5% level, compared to Lijphart’s analysis, at the 1% level. As such, I would argue that my analysis proves a weak correlation between consensus democracy and macroeconomic performance.
A further issue with my analysis is the use of HDI as a control for the level of economic development. The HDI is a measure of life expectancy, education and per capita income in the economy. There are other, more holistic, measures of economic development and well-being in the economy. HDI, for example, does not include environmental sustainability. Since the government’s role may be to maximise social welfare, ignoring environmental sustainability could greatly disable them from their purpose. If the environment is in ruins, then people’s quality of lives will be affected in several ways. More people will even die, as a result. It may even be better to use other macroeconomic variables, which reveal a different aspect of the economy, such as the unemployment levels in the economy. However, I chose to use HDI, because it synthesis different economic variables, which contain variables that influences people’s quality of lives.
In conclusion, I have argued that consensus democracy does improve the quality of government. While the correlation may not be as statistically significant, when comparing our results to Lijphart’s, there is a statistically significant correlation between consensus democracy and quality of government in the three subsections of quality of government. Due to limitations in the practicalities of writing this essay, I have not considered the much wider debate of what else may be included in the quality of government. Even of the indicators I had chosen, I could have analysed, more closely, which aspects of social inequality was related to the quality of government, or which macroeconomic variables should have been used (there is economic debate on whether inflation matters over unemployment as an indicator of good macroeconomic performance).
Stuck on your essay?
