Mobile Edition
Print
Get the Mag
Weekly UpdatesINTRODUCTION
Many studies on the determinants of happiness or subjective well-being (SWB) have explored the impact of particular demographic variables (sex, age, marital status, etc) and also of economic and political performance variables, such as unemployment status, inflation rates, efficiency of government services, civil liberties, and so on. Political and economic systems and the institutions composing them may be another source of SWB but, until recently, their impact has received relatively little attention. In recent years, however, Frey and his colleagues (eg, Frey and Stutzer, 2000, 2002; and Frey, 2008) have shown that SWB varies between Swiss cantons according to the degree of direct democracy; and Helliwell (2002) and Helliwell and Huang (2006) provide cross-country evidence that citizens are happier in nations with constitution establishing presidential systems and some form of proportional voting.
The relation between economic systems and happiness has received virtually no serious theoretical or empirical analysis in the scholarly literature in any discipline. In the popular media one sometimes finds two types of vague arguments about this relationship: (a) Some leftwing commentators assert that happiness is greatest in 'gentle' welfare states, such as the Nordic nations, while it is lowest in English-speaking nations with their alleged 'dog-eat-dog' capitalism. Some rightwing commentators assert the opposite, that Nordic economic systems stifle initiative and frustrate would-be entrepreneurs, while lassez faire capitalism allows people to reach their greatest potential and, thus, become happier. (b) Others seem to say that most people are unmoved by abstract economic arguments and view the economic system instrumentally, that is, they are more concerned with the performance of the economy, rather than the system per se. As the economic performance of the four capitalist economic systems studied in this note are roughly similar by most criteria (Pryor, 2008), this argument can be taken to imply that there is no significant difference in happiness in countries with these different systems.
This study is motivated by three concerns: (a) to explore empirically whether the relation found between the political system and happiness also holds for the economic system so that a stylised fact can be established, which can provide a focus on future theoretical economic analysis; (b) to determine which economic institutions are responsible for this relationship, assuming that such a correlation between system and happiness can be found; (c) and to test empirically the two ideological conjectures outlined in the previous paragraph.
The next section focuses on three methodological issues: the measurement of happiness, the selection of the sample, and the definition of the economic systems. The subsequent section surveys 53 possible determinants of happiness on a national level and explains why and how I have used principal component analysis to reduce these possible explanatory variables to a manageable number. The next following section presents the results of the regression analysis attempting to relate economic system to the level of happiness. The penultimate section explodes several objections to this type of analysis. The final section summarises the results and explores in greater detail why economic systems do not influence SWB, whereas political systems do.
This topic raises a number of intriguing side issues, but because this is a note with very few positive results, rather than a treatise, I leave these matters to others. And for similar reasons I also do not review the happiness literature in-depth.
THREE KEY METHODOLOGICAL ISSUES
How is happiness measured?
Hundreds of public opinion surveys pose questions of the following type to measure SWB or happiness (I use the terms interchangeably): All things considered, how happy would you say that you are on a scale from 0 through 10? Sometimes the question is asked in terms of life satisfaction, but the results are roughly the same (Veenhoven, 2000, esp. pp. 268-273). The results of these surveys have been standardised and summarised in the World Database of Happiness. (1)
Several serious questions can be immediately raised about the meaningfulness of these tabulated answers: Are people telling the truth? Do the SWB scales yield comparable scores across individuals and societies? Are people's answers consistent over time? Are international comparisons biased by cultural norms concerning the propriety of talking about one's individual happiness? Such questions have been explored at considerable length in the literature on happiness; the general consensus is that in industrial nations the answers to these survey questions about SWB do reflect an important reality independent of cultural differences and the answers can be meaningfully compared from country to country. (2) For this note, I accept the international comparability of the measures of SWB, refer sceptics to the literature cited in the footnote, and turn to other issues.
Data on happiness used in this essay come from Veenhoven and Kalmijn (2005) and represent arithmetic averages of the results from various surveys taken during the 1990s and reported in the World Databank of Happiness (Veenhoven et al., 1994). These data, which are used throughout this note, are presented in Table 1 along with per capita GDP and economic system.
As there are many possible determinants of a nation's average level of happiness, special statistical techniques must also be used to reduce the number of possible explanatory variables to be held constant while investigating the impacts of the economic system, a matter discussed below in considerable detail.
Selection of the sample
For reasons of data availability this study uses countries as my unit of analysis, which means that, like other cross-country studies of happiness, we cannot easily determine the impact, of certain key demographic variables such as gender or age. To explore the relationship between economic system and happiness, it would be desirable to include in the sample countries representing as many different economic systems as possible. Unfortunately, implementing this sentiment raises some serious problems.
As the available data on happiness do not allow a time-series analysis, we are limited to a cross-section approach in which unspecified variables may play a crucial role. It is, therefore, useful to select a sample with as many similar features (excepting the economic system) as possible. The industrial nations of the world have had considerable contact with each other for many centuries so that many cultural features are similar. For this reason, I have excluded the developing nations from consideration. It is also sometimes difficult to determine the separate influence of political and economic factors on happiness, for example, disentangling the separate impacts of political and the economic institutions in communist or formerly communist nations, so I have also excluded them from the sample.
This leaves us with a group of 21 nations, all members of the Organisation for Economic Co-operation and Development (OECD), which, as shown below, have different economic systems. All are democracies and their political systems also have many features in common. Most also have relatively similar levels of economic development. Their similarities in the cultural and political dimensions give us more confidence that any discovered differences in their average levels of happiness do not arise from these causes. Although the sample is small, the use of a regression analysis with careful attention to statistical significance of the estimated coefficient means that we can have some confidence in the derived conclusions.
Definition of an economic system
An economic system can be defined as a set of complementary economic institutions and organisations that directly influence the allocation of resources. It differs from the political system because these institutions are primarily concerned with property relations or with the production and distribution of goods and services, rather than the exercise of political power. Of course, many economic institutions have important political elements, for example, government regulation, but a definition of the economic system focuses on how and to what degree such regulation influences resource allocation, whereas a definition of a political system deals with how the regulators are appointed and how they respond to the people's will. Thus, definitions of economic and political systems may deal with the same phenomenon, but from different perspectives.
Although many designate economic systems in terms of one or two institutions (eg, a welfare state), the approach taken in this essay is more comprehensive and takes into account 40 institutions, measured by quantitative indicators representing various institutions in the market for goods and services (eg, product market regulation or barriers to starting new businesses), the labour market (eg, coverage of collective bargaining agreements or protection of jobs), the production and business sector (eg, concentration of corporate ownership or shareholder rights), the government sector (eg, government's share in total investment or share of employment in government owned enterprises), and the financial sector (eg, central bank independence or bank concentration). These 40 institutional indicators (listed in Pryor, website) do not include specific short-run policy variables or performance outcomes, or institutions that are more focused on political and social matters.
[FIGURE 1 OMITTED]
The 40 institutional variables were then subjected to a cluster analysis, a procedure that searches for complementarities between the defining characteristics (institutions) to determine which groups of countries feature roughly similar institutional configurations. Diagram 1 presents an example using just two, rather than 40, institutions. The axes A and B indicate the institutions used to define the economic system, for example, the ratio of government expenditures to the GDP and the degree to which wages are bargained at the national rather than the industrial level. The x's define a country's location in this two-dimensional institutional space and the spaces enclosed by the solid lines indicate the institutional clusters, which, in turn, define the three economic systems Q, R, and S (Figure 1).
FEED