High inequality lowers wealth

Transcription

1 High inequality lowers wealth This contribution has been elaborated in the frame of the collaboration between FEPS and ECLM... ( Chief Analyst Jonas Schytz Juul, September 2011 Any further information can be obtained through FEPS Secretary General, Dr Ernst Stetter, at or through FEPS Junior Policy Advisor, David Kitching at FEPS Rue Montoyer 40 B-1000 Brussels

2 A comparison of the level of wealth and inequality in different countries shows that countries with a high degree of inequality in general have lower levels of wealth. This is found whether one includes all countries, just the OECD-countries or only members of the European Union. September 2011 Contents: High inequality lowers wealth... 3 Correlation between inequality and wealth... 3 Low inequality causes many positive effects... 8 Other studies of inequality and wealth... 9 Appendix A. Regressions on wealth and inequality References

3 High inequality lowers wealth How the level of inequality affects growth and wealth is a subject of frequent discussion. A widespread point of view is that a high level of inequality is an advantage for the growth in so- ciety since it promotes incentives to work hard. The hypothesis is that if the gap between top and bottom is small, it leaves little incentives to make an extra effort, since the gain is too small. Conversely, if there is a big difference be- tween top and bottom there will be a big incentive to make an extra effort. In contrast there are a number of arguments leading to the conclusion that increased inequali- ty does not lead to higher growth. Firstly, it is more difficult to counter the social heritage with a large distance between rich and poor. Secondly, a country characterized by a broad middle class is more likely to invest in eg. educa- tion, infrastructure and health. These services will benefit the general level of prosperity in the country. Thirdly, it provides more political instability with big differences between top and bottom. A country with a small elite and a large lower class will typically be characterized by more insta- the entire vo- bility than a country with a higher degree of equality. Under these circumstances lume of talent in a country will not be used. Children who grow up in the lower class are at high risk of remaining in an uneducated lower class when they become adults regardless of qualifications. Fourthly, a number of studiess show that corruption is more prevalent in countries with a high degree of inequality. The conclusion is, that one can find theoretical arguments that inequality leads to higher growth but also arguments saying that low inequality leads to more prosperity. Though, empir- ical studies must be used to investigate the relationship between inequality and wealth. Correlation between inequality and wealth Observing the inequality and prosperity in countries around the world, there seems to be a re- lationship: The lower the inequality, the greater level of prosperity. There are several ways of illustrating that. One way is to compare the level of wealth between the countries with lowest inequality with the countries with the highest inequality. Whether you look at OECD countries or the EU countries the same result appear: On average, the coun- is narrowed tries with the lowest inequality have the highest level of wealth. If the analysis down to Western European countries only the result is still the same. This is illustrated in figwealth is shown ure 1. To show the robustness of the result, the link between inequality and for all high and very high developed countries (UN definition), OECD-countries and EU countries. Furthermore, the EU countries are divided into two groups: Western EU countries and 3

4 new EU countries. As shown from the graph, the level of wealth is in all cases greater for the most equal countries. For the OECD-countries the level of wealth is app. 18 pct. higher for the most equal countries, compared to the countries with the highest level of inequality. For the EU countries the differ- is a little lower, ence is app. 25 pct. Looking only at the Western EU countries the difference and for the new EU countries the difference is a higher. A different way to illustrate the correlation between wealth and inequality is by a statistical analysis. Figure 2 shows inequality and wealth level per person in 145 countries and the esti- that there is a mated linear relationship between the two variables. From the figure it appears clear tendency that the level of wealth is lower when inequality is high. This is also confirmed by a regression analysis thatt shows that the correlation is significant i.e. the lower the in- equality, the greater level of wealth. 4

5 The result in figure 2 is however questionable because of the large uncertainties in measuring inequality and wealth level of 145 very different countries. Especially in developing countries the uncertainties in measurements are very large. Futhermore, the 145 countries differ a lot in many other areas important to the level of wealth that are not included in the analysis. It is therefore appropriate to restrict the analysis by comparing a smaller group of countries. When the countries are more similar it causes less uncertainty in the measurements of inequacompare Western lity and wealth. Additionally it is more meaningful in a Danish context to countries, which are more similar to the Danish economy. Limiting the analysis in figure 2 to data only from the highly and very highly developed countries, one still finds a significant corlevel of inequality relation between wealth and equality: In general countries with the lowest has the highest level of wealth. The result of this regression on these 69 high and very highly developed countries in the UN data is shown in Appendix A. Instead of relying on the UN statistics of wealth and inequality, one can refine the analysis by looking only at OECD countries. Therefore figure 3 is based on OECD calculations of wealth and inequality in 30 OECD countries. Figure 3 shows that the significant correlation from figure 2 is also found by looking at the wealthier OECD countries and by using the OECD data as source. Again the countries as well as 5

6 the linear estimated relationship between wealth and inequality are shown. Figure 3 indicates that countries with low inequality in general have a higher level of prosperity. Appendix A shows a number of other estimates based on the data of OECD. The additional es- timates also show that countries with low inequality in general have a higher level of prosperity. A third source for comparison of inequality and wealth across countries is EUROSTAT. EUROS- TAT also tries to harmonize the statistics to make them more comparable. When using the da- the inequali- ta of EUROSTAT the same correlation as in the figures 2 and 3 is found: The lower ty, the higher the general level of prosperity. 6

7 From the figures one can discuss whether certain countries should be omitted from the calcuan extreme obser- lation, as they could be seen as outliers. Especially Luxembourg appears as vation. Appendix A therefore has a number of additional regressions in which Luxembourg is removed and only the remaining EU countries are included. In these cases it is also significant that coun- tries with lower inequality generally have higher levels of prosperity. Furthermore when observing figure 4 one can consider whether the regression should be di- are placed vided into "new" and "old" EU countries. Many of the Eastern European countries below the estimated relationship, while Western European countries are placed above. In Fig- ure 5 the regression is therefore divided into "new" and "old" EU countries 1. Also in this scenario the relationship is significant. It is interesting to note that one finds the same relationship between wealth and inequality, for "new" as well as for the "old" EU coun- tries, but at different levels. The indication is therefore that the correlation between wealth and inequality is robust. 1 The estimation is performed with a "dummy" variables in new EU countries. I.e. all EU countries included in the same estimation, but the dummy variable provides two different levels of correlation between inequality and wealth. 7

8 Low inequality causes many positive effects The estimations above only show a correlation between inequality and wealth. What they do not tell, however, is which way the causality goes. Do the rich countries have a low level of in- rich because equality because they can afford to redistribute? Or do the countries become they have a low level of inequality? In a recent research article in the journal "Journal of Development Economics" William Easterly discusses this problem 2. He shows that higher inequality provides a lower level of wealth, and that the causality (cause-and-effect) is that way. This means that low inequality gives a higher level of prosperity in the long run. For this study he uses the Instrument Variable estimation, which is explained in Box 2. Using this method Easterly (2007) shows a high significant correlation between inequality and wealth. The result is therefore that high inequality in isolation affects the level of prosperity negatively. If the inequality rises, it results in a lower level of prosperity. In addition to the connectionn between inequality and wealth, Easterly (2007) shows that high- of the coun- er inequality also results in a lower educational level and in an inferior structure try. Altogether Easterly (2007) finds that higher inequality leads to: 2 William Easterly, "Inequality Does Cause Underdevelopment: Insights from a new instrument", Journal of Development Economics 84, Then: Easterly (2007). 8

9 1. A lower level of economic prosperity 2. Poor institutions in the form of more corruption, more political instability and lower levels of democracy 3. Lower level of education. The relationship between inequality on the one hand, and prosperity, quality of institutions and the educational level on the other is summarized in Table 1. Here the regression results from Easterly s preferred model is reproduced. Other studies of inequality and wealth Before Easterly (2007) there have been numerous studies that also examines the relationship between wealth and inequality. These studies, however, reached different conclusions as to whether inequality affects wealth positively or negatively. Among the studies finding that inequality is good for growth and prosperity one of the argubecause they have ments is that it is wise to concentrate wealth in a country among the richest a higher rate of savings. This leads to a higher level of investments which leads to a higher level of prosperity. Another argument is that increased inequality causes bigger incentives for hard work to climb the social ladder. Among the arguments against increasing inequality as growth enhancing is the risk that educaand poor are very tion will not reach the lower parts of society when the gap between rich large. This will lead to a larger group remaining in uneducated poverty and that affects the general level of wealth negatively. Additionally it is an argument that a high level of inequality 9

10 leads to a higher level of political instability and high inequality makes it harder to counter the social heritage. Therefore the potential of the population cannot be fully exploited in a country with a high level of inequality. Additionally it is shown in Easterly (2001) that a large middle class generally influences posithe democracy in tively on the level of education, the infrastructure, the political stability and the society. Through these channels a large middle class also leads to growth and greater prosperity for the country. Finally some studies argued that a certain level of inequality is an advantage, but too much incountry seen as a equality causes instability and is therefore harmful for the development of a whole. However, the majority of the previous studies investigating the relationship between wealth and inequality are limited both by the quality of the data and by the fact that they do not use the instrument variable method. That causes endogeneity problems in estimations. The Easter- and uses instru- ly (2007) on the other hand both uses a newer more quality-assured data set, ments for estimation. There is still, however, a need for further quality assurance of data in the future, since there are differences in each country's methods of calculating inequality and wealth causing uncertainties in estimations. 10

11 Appendix A. Regressions on wealth and inequality In the figures with UN as the source it is open for discussion how many countries should be indata from the clude in the regression. For example there are significant uncertainties in the least developed countries. In Figure A the regression is therefore limited so that the least developed countries are omitted. That leads to 106 countries in the regression. With this group of countries the relationship between wealth and inequality is still significant. In Figure B the group of countries is further limited to include only those countries which the UN defines as highly or very highly developed countries. With this total of 69 countries, the relationship is still significant. This is shown in Figure A and Figure B. In the estimations with OECD as a source an average of Gini-coefficients is used in the period In figure C the estimations are shown using only the latest data on Gini coefficient from the OECD. Again it is obvious that the correlation between wealth and inequality is signif- correlation is icant. Also if Luxembourg is omitted as an extreme observation the significant found. It is shown in Figure D below. In the estimations based on data from EUROSTAT it is obvious that Luxembourg is an outlier. Luxembourg is omitted in Figure E and Figure F and by doing so the relationship between 11

12 wealth and inequality remains significant as shown in Figure E. Another relevant regression may be looking only at the EU countries. This is also done in figure F, where the correlation is still significant. In Table A1 a summary of the obtained estimation results is shown. References William Easterly, The Middle Class Consensus and Economic Development, Journal of Economic Growth 6, William Easterly, Inequality does cause underdevelopment: Insights from a new instrument, Journal of Development Economics 84, OECD, Growing Unequal, OECD UNDP, Human Development Report 2010, FN