Distributional National Accounts in New Zealand

In this study we investigated how are national accounts distributed among income percentiles. Our series display the evolution of distribution of National Accounts (either Factor Income or National Income) using two separate data bases (IR data and HES data) so our results have shown how results are different using different database as well. Findings determine that the gap between poor and wealth has changed neither over period 2006-2015, nor in a wider period of time (2000-2018) since share of bottom50 percentiles from National Income has been between 9 and 13 percent, share of middle class has been between 34 and 39 percent, and share of top 10 percentiles hasbeen between 50 and 55 percent. share top 1 percentile from National Income has increased drastically in years 2005, 2011, and 2016 while share of percentiles 90th-99thhas dropped in those years. We should stress again that our methods and results should be viewed not as a final product, but rather as a prototype and part of an ongoing attempt to provide more and more complete and transparent inequality statistics. As better sources and methods become available, the results always can be improved accordingly.


Motivation
Even a quick consideration of income inequality studies reveals the fact that the adverse effects of this phenomenon are too serious to be ignored. For instance, Hsieh and Pugh (1993) show that more inequal societies face the higher rates of homicide and crime and Fajnzylber et al. (2002) provide evidence that inequality affects health, longevity and quality of social relations in the society. Therefore, policy makers should have special attention to this matter as one of their most important obligations. But, the primary concern is having reliable measurements of income inequality. The best decisions or policies for any economic problem are made when a comprehensive and precise vision as well as accurate indicators to measure or compare that matter are available. For this aim either accessing to reliable sources of data or corroborated method for processing data are critical. Also, comparability of indicators is another point that makes choosing proper data set and methodology more substantial.

Problem Statement
Considering the importance of measuring inequality, establishing a time series of different group income shares to reflect the distribution of income inequality and its changes over time has been attracting the attention of many researchers, economists and policy makers across the world, specifically in the Organisation for Economic Cooperation and Development (OECD) countries. Due to this fact, there has been several efforts for measuring income inequality, finding the share of different income groups from economic growth, and monitoring the change of income inequality specially across the top 1.2 Problem Statement income percentiles. Particularly, it became even more attractive to focus on during and after the period 1984-96, that New Zealand experienced a wide range of economic and social policy reform (Evans et al., 1996) . To obtain this aim, researchers mostly have used tax income data and information from surveys such as Household Economic Survey to determine Gini coefficient or top income shares. For example, Podder and Chatterjee (2002) examines the trends of household income inequality in New Zealand using Household Economic Survey (HES) data in unit record form which indicates a steady upward trend in income inequality in New Zealand. Although these endeavours shed to a light over this matter in New Zealand, some limitations with their data bases and methodologies has left this picture still vague in some aspects. Relying merely on micro-data (tax and survey data) causes a gap between National Accounts and these studies, that leads the result to be inconsistent with economic total aggregations and growth. Therefore, there is no satisfactory answer for estimating the share of different income groups (bottom 50 percentiles, middle class, top 10 percentiles and top 1 percentile) from economic growth. Moreover, we are interested in knowing if changes in income inequality is because of changing in capital-labour ratio in national income or changes in distribution of income sources (wage and salary or capital income) and returns to capital. 'How does the participation of female as labour force shape the trends in income distribution' and 'how are the distributions of income trends reshaped by interfering the government with using tax and transfers tools' are both noticeable questions that there are no clear responses for them with mentioned previous studies. New studies came by new methods to overcome all these blind points.
Distributional National Accounts (DINA) is an idea suggested by Alvaredo et al. (2016) that provides decomposition of changes in income by different income groups using a scale factor that is constructed by both micro-data and national accounts sides, also by gender factor to cover the inconsistency problem as well as female contribution problem. The World Inequality Database (WID.world) Database (2020) aims to provide open and convenient access to the most extensive available database of the world distribution of income and wealth, both within countries and between countries, focuses on this methodology to make comparable and consistent series for the countries across Introduction the world.

Summary of Contributions
The contribution of the present study is estimating Distributional National Accounts for New Zealand to investigate income inequality changes in an specific period of time (2000 to 2018). For this purpose, Inland Revenue tax data are combined with National Income data. A second version is created using the Household Economic Survey data in place of the Inland Revenue tax data. Our series display the evolution of distribution of National Accounts (either Factor Income or National Income) using two separate data bases (IR data and HES data) so our results show how results are different using different database as well. We find that the gap between high and low incomes has changed neither over period 2006-2015, nor in a wider period of time (2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018). The income share of the bottom 50 percentiles from National Income has been between 9 and 13 percent, share of middle class (50 to 90th percentile) has been between 34 and 39 percent, and share of top 10 percent has been between 50 and 55 percent. The share of the top 1 percent from National Income has increased drastically in years 2005, 2011, and 2016 while share of percentiles 90th-99th has dropped in those years.

Thesis Outline
The rest of this paper is organised as follows. Section 2 describes some previous studies which focus on measuring inequality with different approaches for either New Zealand or other countries. In section 3, data sources and methodology are described. Section 4 includes results and main findings. Section 5 presents concluding comments and research perspectives.

Introduction
Growing interest in either measuring income and wealth inequality or developing efficient methodologies, for this measurement specially for developed countries, gave rise to a flourishing literature in this regard in recent years. In particular, studies for estimating the share of income for top income groups has increased significantly in last two decades. We review some efforts which have been made around the world using same approach to measure inequality. They all used the same method we used to estimate distributional national accounts. Results of all the efforts below are available on WID.world database for their methodology, data and other aspects of their measurements are consistent with this database's approach. Piketty et al. (2018) combined tax, survey, and national accounts data to measure the distribution of national income in the United States since 1913. Their work comprises estimating not only the distribution of pretax national income, but also distribution of national income after taking into account the government intervention. Their calculations display that average pretax real national income per adult increased 60 percent from 1980 to 2014, but they show that it decreased for the bottom 50 percent of the distribution around 16,000 dollar a year. The pretax income of the middle class has grown 40 percent since 1980. In this attempt, they also estimate the share of either capital or labour income for different income groups. According to their findings, upsurge of top incomes was first a labour income phenomenon but it has changed to capital 2.2 A General Literature on DINA income phenomenon since 2000. This study reveals that the government's role is not substantial in decreasing the income inequality in mentioned years. Their study has another sector which make it even more valuable. They broke down DINA regarding to the gender of individuals and their results cover the share of female income from total income in each percentile. According to that share of females has increased in last 20 years that has caused reduction in inequality among adults.

A General Literature on DINA
In France, in DINA fields there are two separate studies for pretax distributional national accounts and post tax distributional national accounts. Garbinti et al. (2018) focused on DINA before government intervention with capital vs. labour income, age and gender breakdown. As other cooperation in WID, they combined national accounts, tax and survey data in a consistent way to build homogeneous annual series on the distribution of national income by percentiles for years 1900 to 2014. They present one substantial result that displays taking advantage of DINA methodology.
They compare their series with previous tax-base findings in long-run inequality series in France. Their series show higher inequality levels for the recent decades. They indicate that although a sharp drop in the concentration of wealth and capital income seemed to cause a decline in inequality, a rising part of capital income increases it considerably. Their detailed breakdowns by age and gender show that gender inequality in labour income declined in recent decades, although it is not substantial among top labour incomes. They also take advantage of DINA results to compare inequality between the U.S and France. They point out that average pretax income among bottom 50 percent adults is 20 percent larger in France than in the U.S while national income per capita is 30 percent smaller in France. Bozio et al. (2018) worked on building DINA series in this effort considering government role in France. They discuss different ways of measuring tax progressivity then use their new series to understand the impact of taxes and transfers on redistribution of income. Their findings display the fact that taxes and transfers reduce total income inequality by 23 percent on average over the 1990-2018 period. For this, both upper-end and lower-end redistribution are applied.
They explain that increasing redistribution in France tax system in considered period of time, which is from reductions in non-contributive social security contributions for the bottom 50 percent of individuals and tax increases for the top 10 percent, has caused a relatively constant level of disposable income inequality.
WID.world provides DINA series for all countries which have built them and updates them constantly.

Inequality in New Zealand
In this sector we summarise studies on inequality in New Zealand. This literature largely focuses on estimating share of income for top income percentiles and most of them using Gini coefficient as an index to talk about inequality. Some of the studies carry out regression approach in this regard, using the income shares of a specified percentile of the distribution as dependent variable or several variables have been used as the dependent variable. Podder and Chatterjee (2002) examined the trends of household income inequality in New Zealand over the period 1984-96. They investigate the manner in which the national income is divided up amongst different groups in society after implementing a wide range of economic and social policy reform in New Zealand. They also measure the contributions of the different sources of personal income to the overall inequality decomposing income inequality by income components. In this effort unit record data from the Household Expenditure and Income Surveys and the Household Economic Survey (HES) are used. Their results clearly show that inequality of household incomes in New Zealand has been on the rise over the mentioned period as indicated by the rising value of the Gini coefficient over the period (over 14 percent increase in 12 years from 1984 to 1996). They note there is a decline in the shares of the bottom eight deciles of the households, the share of the ninth decile has remained steady, the share of the top decile has increased significantly. To capture the changes in the relative income shares, they observe how the ratios of the shares of the top and the bottom percentiles for example have altered over time. The top 5 percent of income earners received over seven times as much as the bottom 10 percent in 1983/84. For the next part of the study they consider seven components of total income. Wages and Salaries, income from self-employment, income from investment, personal superannuation, na-2.2 A General Literature on DINA tional superannuation, government cash transfer, and other incomes. Then note that income from wages and salary also has a concentration coefficient higher than the Gini coefficient of total income which implies that this income is more unevenly distributed (in favour of the higher income groups) than total income. This uneven distribution affects the overall inequality strongly. Dixon(1996) used Household Economic Survey (HES) database to measure the distribution of individual earnings and investigate long-run changes in the earnings structure between 1984 and 1995. This study's results include aggregate earnings inequality, the gender earnings gap and shifts in relative earnings by level of educational attainment. This study take advantage of regressing variables. To explore the changes in the independent influence of each measured attribute on earnings, a simple linear earnings equation is estimated for each annual dataset. They indicate that the overall increase in inequality in mentioned decade was very small, and caused by a rise in the relative earnings of workers at the top of the distribution, rather than a decline in the relative earnings of low wage workers. In addition, there has been a substantial reduction in gender differentials over the decade, reflecting an upward shift of all levels of the female distribution and some down-ward movement of the male distribution. They also found an increase in earnings inequality among males with post-school qualifications. Winkelmann and Winkelmann (1998) examined inequality from a labour market perspective in an econometric framework. They examine the labour market outcomes, in terms of employment and incomes, for immigrants in New Zealand with the help of data primarily from the 5-or 10-yearly population censuses between 1981 and 1996.
This study looks at the income differentials within immigrant groups, and between specified categories of immigrants and native-born New Zealanders. They conduct a cohort analysis of immigrants' relative incomes and obtained results from regressions using all employed individuals aged 15-64 for whom income data are available. Their analysis answers the question of how much of the difference in incomes between immigrants and natives remains after we control for hours of work, gender, and productive characteristics. They display evidence for a substantial income disadvantage of arriving immigrants relative to natives after we account for differences in qualification levels and other personal characteristics. Bakker and Creedy (1999) analysed the effects of macroeconomic variables on the personal distribution of income over time. They use a conditional lognormal-exponential mixture maximum likelihood estimates to model the complete distribution of income in each year. This method has been used to make New Zealand income distribution for wage and salary earners over the period 1985 to 1994. They came to the conclusion that either increasing the unemployment rate and reductions in the rate of GDP growth have driven the increase in inequality in that period. Ball and Creedy (2016) used HES data and survey calibration method, first, for analysing annual income and expenditure inequality in New Zealand over 1983 and second, for comparing the inequality of market incomes with that of disposable incomes to investigate the extent of redistribution through the tax and benefit system. They aimed to describe components of inequality as well and to reach this purpose a decomposition method is used involving five sets of variables (age/gender structure, labour force participation, household type, housing tenure type and occupancy rate) along with the sample itself. They found an increase in the inequality of market and disposable income per adult equivalent person from the late 1980s to the early 1990s. They also point out that inequality changes are influenced by a range of factors associated with the structure of the population, which are expected to change over the relevant period. In addition, there has been an increase in female labour force participation as well as the increase in participation among older males over the period. Therefore, with a constant demographic and labour force structure, the inequality of expenditure displayed a 'flatter' profile over the period. Creedy et al. (2017) estimated the Gini index between 1935 and 2014, using tabulated data on personal taxable incomes from Inland Revenue for estimates between 1981 and 2014. They find that the Gini index is relatively constant. In addition, the authors investigate differences in the Gini index of males and females. They find that while overall income inequality is unchanged, income inequality among females has declined since 2000 and is lower than that of males over the entire period.

DINA in New Zealand
NZ (2018) followed the methodology recommended by the OECD Expert Group on Disparities in a National Accounts framework (EGDNA) to distribute the national accounts values across different household groups. To reach this purpose they use Household Economic Survey (HES) data as well as the National Accounts (more specifically the Household Sector Accounts). The results plot the ratio to the average of disposable income and adjusted disposable income by quintile for 2016. They show that the disposable income of households classified to the highest quintile is 2 times the average, while the lowest quintile is 30 percent of the average. The similar ratios for adjusted disposable income are 1.8 times and 50 percent respectively, illustrating how the addition of social transfers in kind reduces the income disparities. It also points out that the consumption disparity measure is much lower than the income measure.
While the disposable income of households in the highest quintile is 2 times the average, the level of final consumption expenditure for the same group of households is only 1.4 times the average. For the lowest quintile, similar figures are 30 percent and 70 percent respectively. Using a consumption ratio significantly lowers the measured level of inequality.
There are some valuable attempts to investigate the evolution of the income distribution focusing on top income shares applying Atkinson-Leigh method 1 . These studies measure top income shares using tabulated data on taxable income published on Inland Revenue's website(www.ird.govt.nz) Revenue (2020).These type of estimates have been repeatedly updated in the World Inequality Database by Leigh (2007, 2008) Atkinson and Leigh (2008) The closest effort to our study in terms of methodology to measure distributional national accounts in New Zealand is Kergozou (2017). Apart from using Atkinson-Leigh method to measure the evolution of top income shares, she also combines tabu-

Data and Methodology
In this section we describe the concepts, data sources and main steps of the methodology that we use to construct income distribution series. We use two main types of data, national accounts and fiscal data (income tax returns). A third type (survey data) is used to construct an alternative income distribution series. We describe our data sources for the period we can use micro-files of income tax returns in New Zealand, explain the methodology, and present our income concepts.

Micro Data Household Economic Survey(HES) Tables
One of the databases we used for distributing national accounts is one of the survey based databases of New Zealand's Integrated Data Infrastructure (IDI). The Household Economic Survey (HES) collects information on household income, savings, and expenditure, as well as demographic information on individuals and households. The  Inland Revenue is the New Zealand government's revenue collection agency. This database includes tax information of all taxpayers in New Zealand consists of more than 16 million rows and 100 million observations. The raw IR data sets in the IDI contain information about income from four main IR income tax return forms: 1. Employer Monthly Schedule (EMS) provides gross earnings where PAYE (Pay As You Earn) 1 is deducted at source. The EMS consists of all wage and salary earners, withholding payments, government transfer payments, and payments from ACC (Accident Compensation Corporation) 2 . It includes categories for government benefits, student allowances, paid parental leave, and New Zealand Superannuation payments. The EMS is filed monthly by the employers and provides pay details of employees who work for them.
2. IR3 for self-employment (filed annually by sole traders) which includes non-zero partnership, self-employment, or shareholder salary income, as well as rental income.
3. IR4S filed by companies includes remuneration income paid to shareholders, directors, and relatives of shareholders (filed annually).
In March 2014, Statistics NZ introduced derived tables. Table data.income-tax-yr has been used in this study. This table is comprised of all records in the Employer Monthly Schedule (EMS), plus additional records from the IR3, IR4S and IR20 tax forms. It also orders the monthly data into tax years,as the first records begin in April 1999.
Thus, when referring to month 1 of this tax year, this means the month of April(e.g.

Macro data National Accounts
In this study all macroeconomic total amounts such as aggregate gross domestic products, compensation of employees, total capital income, national income, and all components to build factor Income and net national income are from Statistics New Zealand data bases available in Economic indicators tab at infoshare website 3 . National accounts available in this data base are calculated following SNA2008 4 -SNE

Data Issues
As there are various issues in terms of all three databases used in this study, we view our attempt an initial draft to construct distributional national accounts for New Zealand. It will be always possible to improve findings once more reliable databases, more knowledge to impute national account components, and improved methodologies cepts, definitions, classifications and accounting rules. In addition, the SNA provides an overview of economic processes, recording how production is distributed among consumers, businesses, government and foreign nations. It shows how income originating in production, modified by taxes and transfers, flows to these groups and how they allocate these flows to consumption, saving and investment. Consequently, the national accounts are one of the building blocks of macroeconomic statistics forming a basis for economic analysis and policy formulation. The SNA is intended for use by all countries, having been designed to accommodate the needs of countries at different stages of economic development. It also provides an overarching framework for standards in other domains of economic statistics, facilitating the integration of these statistical systems to achieve consistency with the national accounts. Furthermore, although the national accounts aggregate all the available information from survey, balance sheets, tax data and so on, they are still imperfect. Zucman (2013) Those findings that rely on tax data are biased as tax data excludes tax evasion and some forms of income, specifically some components of capital income, are not subject to tax and do not appear on income tax declaration that causes underestimating the inequality. In using survey data, we need to emphasis that there is a strong correlation between survey responses and administrative records that makes HES data imperfect. Ball and Ormsby (2017) 3.2 Methodology (How to Construct DINA)

.1 Overview of the approach
The methodology is applied to distribute the national accounts values across different income groups largely follows the step-by-step approach recommended by the Distributional National Accounts Guidelines: Methods and Concepts Used in WID.world Alvaredo et al. (2016). We present a brief overview of the method below: Step 1: Adjust the National Accounts and count the total aggregation of labour income and capital income in National Accounts.
Step 2: Designate each single individual income from micro data sources (HES or fiscal income) to capital, labour, or Pension source categories, then count sum of all income in each category.
Step 3: Scale the aggregate amounts from step 2 to the total aggregation of the relevant category of adjusted National Accounts from step 1.
Step 4: Calculate a scale factor and use it for scaling individual's income and build the considered series.

The Income Concepts
The income concepts that are used in DINA series are defined in the same manner in all countries and time periods, and aim to be independent from the fiscal legislation of the given country/year. Also, all national account concepts are codified as the official definition in SNA (System of National Accounts) version 2008. One of the central limitations of national account series, specifically GDP, is that they do not provide any information about the extent to which the different social groups benefit from growth.
Apart from coping with the gap between national accounts and tax data, DINA series overcome the mentioned lack of national account series as well. The four basic pretax and post-tax income concepts that are useful to measure income inequality are anchored upon the notion of "national income" that is defined as GDP, minus consumption of fixed capital, plus net foreign income. Including capital depreciation would artificially inflate the economic income of capital owners as it does not allow for consumption or the accumulation of wealth. Additionally, including foreign income is important as foreign investment income can be significant for top income earners. At the individual level, income differs whether it is observed before or after the operation of the pension system and government redistribution. We therefore define three income concepts that all add up to national income: pretax factor income, pretax national income, and post-tax national income. The key difference between pretax factor income and pretax national income is the treatment of pensions, which are counted on a contribution basis for pretax factor income and on a distribution basis for pretax national income. Post-tax national income deducts all taxes and adds back all public spending, including public goods consumption. It is worth mentioning that aggregate pretax national income, pretax factor income, and post-tax national income are all equal to aggregate national income, as defined by SNA 2008. As we explain below, we focused on Pretax concepts.
Calculating Post-Tax DINA series is out of this study's scope.

Factor Income
Pretax factor income, which for simplicity we sometime refer to as "factor income", is equal to the sum of all pretax income flows accruing directly or indirectly to the owners of the production factors, labour and capital, before taking into account the operation of the tax/transfer system (including indirect taxes), and before taking into account the operation of the pension system. One problem with this concept of income is that retirees typically have little factor income, so that the inequality of factor income tends to rise mechanically with the fraction of old-age individuals in the population, potentially biasing comparisons over time and across countries. However, we draw useful insights from this concept as well.

National Income
Pretax national income is equal to the sum of all pretax income flows accruing to the individual owners of the production factors, labour and capital, before taking into account the operation of the tax/transfer system, but after taking into account the operation of the pension system. That is, the difference with factor income is that pretax income includes Social Security (old-age, survivor, and disability insurance) benefits, unemployment insurance benefits, and private pension benefits, while it excludes the contributions to Social Security, private pensions, and unemployment insurance.

DINA
The outstanding advantage of our study is using granular income data include a breakdown of income by source. We go through our calculations using two separate frameworks since income source classifications table in HES data are different from those used by IR tax data. This gives rise to separate graphs. In appendix A, income source classification tables for either databases are attached. In both calculations we capture 100 percent of national account by construction.

From Taxable Income to Pretax Factor Income
The starting point of our distributional national accounts is the individual micro data reported by New Zealand Inland Revenue which is the largest sample of taxpayers information containing the income source categories. Tax data contains information about most of the components of Factor and National Income. However, they miss some parts of these components as they are untaxed. The first step is to designate each single income from each individual to one of the three following categories: 'labour income', 'capital income', or 'pension, Benefit, and transfer'. For that, we imputed wage and salary 6 , sole trader PAYE deducted income 7 , sole trader Withholding income 8 , partner income PAYE deducted 9 , partner income Withholding 10 , company director/shareholder PAYE deducted 11 ,and company director/shareholder WHT 12 to the labour income category. And partnership income 13 , director/shareholder income 14 , sole trader Income 15 , and rental income 16 to the capital income category 17 . Then, for each category sum up total income of each individual in that category. To distribute Factor Income based on its definition we need to find the total for labour income and capital income. So, sum up all individual incomes in each category to reach the total labour income and total capital income. The next step is comparing these results with their correspondent data in national accounts. As we already mentioned there are always a gap between these results. It is mostly because of the fact that some of the incomes which appear in national accounts are not reported in the tax returns.
For instance, rental incomes in national accounts include rent of the houses occupied by their owners, while tax data excludes information about this type of income. To For items which are known as self employment such as Sole Trader, we consider sole trader PAYE deducted and sole trader Withholding as labour income and sole trader as capital income. That is same for the partner income, the company director/shareholder income, and the partnership income.
3.2 Methodology (How to Construct DINA) cope with the mentioned problem, DINA method scales up/down the income of each individual using a scale factor which is built based on the process below. In this case, we need to find aggregate Labour and Capital income from factor national income (Factor Income). As we mentioned before, according to SNA, aggregate income from two factors of production('labour income' and 'capital income') add up to the Factor Income. Using relevant table in national accounts database 18 containing GNI 19 , CFC 20 and Total Compensation of employees, as well as following the equations below, we obtained the total amount of capital income and Factor Income.

DINA Scale Factor:
Scale factor for labour income: 'c' is total income from capital sources and 'w' is total income from labour income sources for each individual.
'c' is total income from capital sources and 'w' is total income from labour income sources for each individual.
We constructed a series from IDINA for all individuals and for each year.
The last step is to find 50th, 90th, 99th percentiles as well as share of bottom 50 percentiles, middle class (50th to 90th percentiles), top 10 percentiles and top 1 percentile using the above series.

From Taxable Income to Pretax National Income
For calculating Pretax National Income, the only difference is to take into account the incomes belong to the 'Pensions, benefits, and transfers' category with a contribution approach. We imputed income from ACC (Accident Compensation corporation) 21 ,  (SNZ)) where all pensioners receive the same amount, NZS is recorded as 'BEN' in tax data. 27 Computer code is also in https://github.com/zarisoleimani/read-me1.git

From Income Reported in Survey to Factor Income
We used the same instruction method to build DINA using HES data, but it is worth stating the imputation approach here as income sources are labeled differently in this database. Table includes all income sources of HES data is available in appendix.
In group '1', we marked total subgroups '1.1' (income from wage and salaries), '1. The rest of the process constructing IDINA series and finding share of each percentile have been applied in here exactly as they were for the tax data.
28 for this allocation we followed Kergozou (2017) 29 following the definition of national income which is GDP minus consumption of fixed capital plus net income received from abroad, we add overseas income in calculation process. Foreign income is important because because investment income can be significant for top income earners.        has dropped in those years. We should stress again that our methods and results should be viewed not as a final product, but rather as a prototype and part of an ongoing attempt to provide more and more complete and transparent inequality statistics.
As better sources and methods become available, the results always can be improved accordingly.

Future directions
There are numerous manners in which these distributional national accounts could be improved upon and developed further. Firstly, this paper uses only two categories (labour and capital income) as source of income and imputes each single income for each individual to one of these two categories producing estimates. However, producing 5.2 Future directions DINA with more elaborated imputation and unfolded these sources of income to more accurate sources of income would allow for more precise estimates. Secondly, this paper only produces estimates of pre-tax national and factor income. Estimates of post-tax national and factor income would provide a comprehensive view of how government redistribution affects inequality. In addition, breaking down the series to find series which describe share of gender, share of different age groups, and share of capital (or labour) from National Income for each income percentile can reveal valuable facts in terms of distributional national accounts.