Distributional national accounts for the Netherlands and a comparison with 12 other countries

• Published date: 01 December 2023
• Author: Arjan Bruil
• Date: 01 December 2023
• DOI: http://doi.org/10.1111/roiw.12609

bs_bs_banner

Review of Income and Wealth

Series 69, Number 4, December 2023

DOI: 10.1111/roiw.12609

DISTRIBUTIONAL NATIONAL ACCOUNTS FOR THE NETHERLANDS

AND A COMPARISON WITH 12 OTHER COUNTRIES

BY ARJAN BRUIL∗

Centraal Bureau voor de Statistiek

Most inequality studies rely on micro data that do not capture a substantial share of income identi-

ed in the national accounts. In the Netherlands, almost one fth of household disposable income is

missed by current inequality statistics. In this paper, we present inequality statistics for the Nether-

lands that capture all of household income, so-called distributional national accounts. Compared to

the current inequality statistics, the Gini coefcient fordisposable income increases substantially from

0.289 to 0.337. Cross-country comparisons show thatsuch a change between Gini coefcients based on

micro-dataversus Gini coefcients based on distributional national accounts does not apply to all coun-

tries.The difference between both Gini coefcients varies not only between countries in the size, butalso

in the sign of the difference.

JEL Codes: D31, E01

Keywords:households, income distribution, income inequality, micro–macro reconciliation

1. INTRODUCTION

Driven by the Report by the Commission on the Measurement of Economic

Performance and Social Progress(Stiglitz et al., 2009), more attention has been paid

to the inclusion of distributional measures in the System of National Accounts

(SNA). This is also in line with the Data Gaps Initiative (Financial Stability

Board, 2009), which stresses the need for indicatorsthat capture better households’

well-being, supplementing the macro aggregates of the SNA framework.

Recommendations in these reports were followed, among other initiatives, by

the Expert Group on Disparities in a National Accounts framework (EGDNA).

This group aims to produce measures of disparities in income across different

household groups, which are consistent with national accounting concepts and

totals, following a harmonized approach (Fesseau and Mattonetti, 2013). Even

though many countries participate in the Expert Group and the importance is

acknowledged by national statistical institutes (Vienna memorandum, 2016), only

a few of those, among which Statistics Netherlands (2019a), started publishing

results of these breakdowns (Ofce for National Statistics, 2015; Insee, 2017;

Australian Bureau of Statistics, 2018; Republic of Slovenia Statistical Ofce, 2018;

Stats NZ, 2018; Statistics Canada, 2019).

Adding distributions to the macrodisposable income contributes to the current

debate byincluding a substantial share of income identied in the national accounts

*Correspondence to: Arjan Bruil, Centraal Bureau voorde Statistiek, Postbus 24500, 2490 HA Den

Haag, The Netherlands (a.bruil@cbs.nl).

© 2022 International Association forResearch in Income and Wealth.

886

Review of Income and Wealth, Series 69, Number 4, December 2023

that is not included in the micro statistics. There are manydifferences in scope, pop-

ulation, or concepts between micro and macro disposable income, which also lead

to different measures of inequality. Much of the work in the expert group focuses

on data gaps, the difference between the macro total and the sum of micro values.

These are considered as quality checks on the resulting distributions. Data gaps

mainly occur because data used to compile the national accounts, and data used

for distributional measures do not come fromthe same data source. The difculties

faced when linking micro and macro dataare not new, however.In the 1980s, Rug-

gles and Ruggles (1986) for the USA, and Adler and Wolfson (1988) for Canada,

investigated the links and (in)consistencies between micro and macro statistics. In

both cases, adjustments to both the micro as well as the macro data were proposed

to harmonize the statistics. In recent years the focus shifted more towards the use

of micro data to arrive at a break-down of macro data by income groups, without

adjusting the macro data themselves.

A second initiative that boosted the compilation of Distributional National

Accounts (DINA) worldwide is the WID.world project by the World Inequality

Lab (Alvaredo et al., 2016). Like the EGDNA, this group aims at a breakdown

of the national accounts. There are many similarities between these projects, but

the results are not directly comparable. The main difference is that the scope of the

EGDNA is limited to the household sector,while the scope of the World Inequality

Lab refers to the total economy. Moreover, the World Inequality Lab denes new

income concepts, instead of using the SNA balancing items (Zwijnenburg, 2019).

In this article, we present our distributional national accounts for the Nether-

lands for 2016, following the EGDNA methodology. We follow a multi-source

approach, combining a large number of different data sources. Even though many

of these data sources are administrative data, we still encounter data gaps. Using a

naive method to calculate the data gap, 82.1 percent of macro disposable income

is covered by micro data. This means that for the remaining 17.9 percent assump-

tions or proxies are needed to allocate these income components to the correct

households.

In 2016 the Gini coefcient for micro disposable income was 0.289 (Statistics

Netherlands, 2019b). According to our calculations the Gini coefcient is 0.337

when the income concepts are made consistent with the nationalaccounts. For users

of the data it is important to understand why two similar statistics (in name) arrive

at completely different levels of inequality. Therefore, we break down this differ-

ence, showing how for each step in the process the degree of inequality is adjusted.

For the Netherlands the population scope, and imputations of concepts that are

not in micro data prove to have a large effect. The largest differences are caused

by gaps between the micro and macro aggregates. Especially the coverage of mixed

income improves, which adds to the change in inequality as well. However, when

placed in the comparative context of the other EGDNA countries, the increase in

the Gini coefcient is not directly conrmed. The difference between micro and

macro inequality is found to be very different across countries. This is important to

investigate because it could be the result of the techniques used and not because of

different characteristics of the economies of countries.

For the Netherlands, our results show that inequality is mainly driven by the

split between labor and capital. Labor and capital shares of primary income are

© 2022 International Association forResearch in Income and Wealth.

887