Distribution of household income, consumption and saving in line with national accounts
By Jorrit Zwijnenburg (Jorrit.ZWIJNENBURG@oecd.org), Statistics and Data Directorate, (OECD)
Economic inequality continues to be a matter of concern for policy makers and citizens. The COVID-19 pandemic has re-emphasised the need for more detailed information on how specific household groups are faring economically, particularly during times of crisis. Whereas distributional information is available from micro statistics, it is traditionally missing from macroeconomic statistics. However, as explained by the Stiglitz-Sen-Fitoussi commission (2009), distributional measures in line with national accounts’ totals may provide important, additional insights to the inequality debate.
In 2011, the OECD and Eurostat started developing methodology to compile disparities in line with national accounts totals (DNA) for income, consumption and saving (see OECD, 2020). These estimates complement existing inequality measures by providing more comprehensive measures of inequality, extending the analysis from income to consumption and saving, and by yielding results that are fully consistent with macroeconomic aggregates, ensuring a high degree of international comparability.
Over the past few years, member states have begun to compile experimental results according to the DNA methodology and, as of the end of 2020, these results have been included in the public databases of the OECD and Eurostat for a first time. 1 Results are available broken down by income quintile, but for several countries also on the basis of household composition and main source of income. The databases also include information on the socio-demographic characteristics of persons and households included in the income quintiles. Furthermore, a working paper has been released (Zwijnenburg et al (2021)), providing further insight into the methodology and highlighting the main results, focusing on thirteen countries, i.e. Australia, Canada, the Czech Republic, France, Ireland, Israel, Italy, Mexico, the Netherlands, New Zealand, Slovenia, Sweden, the United Kingdom and the United States.
Income inequality
The DNA results provide information on the level of income inequality within and across countries. Figure 1 displays the relative adjusted disposable income of five income quintiles for six member states. Mexico reports the highest ratio to the average of the household sector as a whole, followed by the United States, with households in the highest income group earning 2.99 and 2.44 times the average, respectively. Sweden reports the lowest ratios for this group, at 1.54 times the average. For the first quintile, the United States and Mexico report the lowest ratios (with only 33% and 35% of the average), whereas the United Kingdom records the highest, at 61% of the average. The trend across quintiles is more or less the same for all countries, with a fairly smooth trend upwards from the first quintile to the fourth quintile, and a steeper increase from the fourth to the fifth quintile. Furthermore, for all countries, the ratio for the third quintile is below 1. As this can be used as an approximation for the median, this implies that the median equivalised adjusted disposable income is below average in all countries. For Mexico and the United States even the equivalised adjusted disposable income of the fourth quintile is below average. To provide additional insight into the relative income inequality across countries, Figure 2 presents the relative difference in income between the highest and the lowest income group for all twelve countries that provided this information. Mexico is the country recording the highest ratio (8.56), followed by the United States (7.41). The other countries are relatively closer together, reporting much lower ratios. Ireland records the lowest (at 2.70), followed by Sweden (2.78), the United Kingdom (2.82), Slovenia (2.84), the Netherlands (2.90) and the Czech Republic (2.91), all recording ratios below 3.
The working paper and databases also include information on the underlying income items. This shows that rising self-employment income and property income (such as dividends and interest income) have been the main drivers of income inequality. Conversely, compensation of employees has tended to be more stable across countries. Furthermore, taxes, social benefits and social contributions have a strong mitigating effect on income inequality in all countries, though to differing extents. For example, for Australia, the ratio of the highest quintile to the lowest drops from 11.6 for primary income to 3.1 for adjusted disposable income, whereas it only drops from 11.7 to 7.4 in the United States. This detailed information provides very relevant evidence for policymakers.
Figure 1. Relative position of each household group compared to the average, by equivalised disposable income quintile
Figure 2. Relative position of the highest to the lowest income households, by equivalised disposable income quintile
Consumption inequality
For consumption, the disparities across quintiles are generally smaller and the results for the various income quintiles closer to the average than for adjusted disposable income. As for income, the United States and Mexico record the highest ratios for the fifth quintile, at 2.29 and 2.18 times the average, respectively. Furthermore, the United States records, by far, the lowest ratio for the first quintile, with consumption amounting to only 32% of the average, followed by Mexico with households in the first quintile recording consumption at 50% of the average. The Netherlands records the highest ratio for the first quintile, at 85% of the average. Turning to the ratio of the highest quintile to the lowest, Sweden records the lowest disparities between the highest and lowest income quintile (1.42), followed by the Netherlands (1.47), the Czech Republic (1.48) and Slovenia (1.53), whereas the United States (7.15) and Mexico (4.37) record the highest. For all countries, these ratios are lower than for income. Looking at underlying consumption items, the items related to basic needs (i.e. food and beverages, and housing) show relatively flat distributions across income quintiles for all countries, whereas items such as recreation, restaurants and hotels, and furnishings show relatively larger disparities. The same can be observed for education and health care, although largely depending on whether or not these services are (partially) provided by government.
Savings results
While distributional patterns are more or less similar for income and consumption, saving results show larger differences across countries, particularly for the lowest income quintile. Figure 3 shows saving as percentage of disposable income per income quintile, for six countries. New Zealand displays a very large negative saving ratio for the first quintile with a big jump to the second quintile, although still negative. Similar results can be observed for Canada, the Netherlands and Sweden. The other countries show smoother increases across quintiles, starting from less negative saving rates. France stands out with the most stable saving ratios across quintiles, with particularly small negative saving ratios for the first and second quintile, as compared to the other countries.
Figure 3. Saving as a percentage of disposable income by equivalised disposable income quintile.
Figure 4. Composition of the private household sector saving ratio
Delving further into the relative size of savings across quintiles, Figure 4 examines the contribution of the different quintiles to the saving ratio for the household sector as a whole. New Zealand records the lowest overall saving ratio (2.1%), with the lowest three quintiles contributing negatively. Canada and the Czech Republic also report low saving ratios for the household sector as a whole (4.4% and 4.7% respectively), but while in Canada this is due largely to particularly negative savings for the first quintile, in the Czech Republic this is due to the first four income quintiles recording negative savings. The latter is also the case for Mexico, but they still report the highest average saving ratio for the household sector as a whole (23.5%), due to very high savings ratio for the fifth quintile. Sweden and the Netherlands also record large positive saving ratios for the household sector as a whole, but in contrast to Mexico, this is not particularly related to substantial saving of the fifth quintile, but due to positive saving by households in the third and fourth quintile.
Socio-demographic information
The databases also include information on the sociodemographic characteristics of persons and households included in the various income quintiles, focusing on breakdowns by age group, labour market status, level of education and housing status. This provides additional insights into the background of the persons and households along the income distribution. For example, several countries record a higher concentration of younger people in the lower income quintiles, whereas the picture is more mixed for other age groups, with the United Kingdom recording a large concentration of people over 65 in the lower income quintiles, while this group tends to be more concentrated in the fourth and fifth quintile in the United States (see Figure 5). Additionally, the concentration of certain labour market status groups shows important differences across countries. For example, retired people tend to be more concentrated in the lower income quintiles in Israel, but the opposite is true for Mexico. Furthermore, whereas all countries show a relatively large representation of employers in the fifth quintile, Mexico and Slovenia also record substantial shares in the first quintile. More detailed information on the various socio-demographic breakdowns can be found in Zwijnenburg et al (2021).
Figure 5. Distribution of age groups across quintiles
Next steps
The DNA results have been made available in the public databases of the OECD and Eurostat as experimental statistics. This provides users with the opportunity to explore these results in more detail and to conduct their own analyses. In the meantime, the DNA work continues. Looking ahead, the main objective is to further improve the timeliness and granularity of the results, to broaden the country coverage, and to extend the scope to include the wealth dimension. This will further increase the relevance of the results for policy analysis.
References
- OECD (2020), “Distributional information on household income, consumption and saving in line with National Accounts – Guidelines”. https://www.oecd.org/sdd/na/OECD-EG-DNA-Guidelines.pdf
- Stiglitz, J.E., A. Sen and J.P. Fitoussi (2009), “Report by the commission on the measurement of economic performance and social progress”, http://library.bsl.org.au/jspui/bitstream/1/1267/1/Measurement_of_economic_performance_and_social_progress.pdf.
- Zwijnenburg, J., S. Bournot, D. Grahn and E. Guidetti (2021), “Distribution of household income, consumption and saving in line with national accounts – Methodology and results from the 2020 collection round”. OECD Statistics Working Paper forthcoming[SPS1]