TL;DR – Really. But there’s still more to be done.
Recently, there has been a lot of talk about inequality. Mdm Halimah’s speech at the opening of the second session of Parliament talked about tackling “inequality vigorously”.
Many MPs and ministers also spoke about it. Minister Ong Ye Kung said that tackling inequality is a national priority. In Minister Ong’s speech, he said that Singapore’s Gini coefficient, after taxes and transfers, is less than that of US and UK.
Wait. Simi Gini coefficient?
Gini coefficient unscrambled
Gini coefficient measures the spread of household incomes over all the households in Singapore. Truth be told, I had attempted to explain it over a year ago here.
It is a way of comparing how distribution of income in a society compares with a similar society in which everyone earned exactly the same amount.
The Gini coefficient is between 0, where everyone earns the same, and 1, where one person earns all the money. In simpler terms, there is more income equality when the Gini coefficient is closer to 0 and less income equality when the coefficient is closer to 1.
Anyway, let’s do this again.
To understand it, we must first understand the Lorenz curve. The curve shows the spread of household incomes plotted on a single graph. First, organise households according to their respective household incomes, from the lowest to the highest.
Understanding the Lorenz curve. (via)
A line is then drawn through the points where the cumulative proportion of households is plotted against the corresponding cumulative proportion of income:
Plotting the Lorenz curve. (via)
And that is the Lorenz curve. Now let’s say there are only five households, and every household has the same income, then this is what the Lorenz curve will look like:
Perfect Equality Line. (via)
That 45-degree line is known as the perfect equality line.
The other extreme is when one household has all the income and the other households earn nothing. The Lorenz curve is flat at the horizontal axis before it rises to 100% where all the income is earned by one household. This is the perfect inequality line.
Perfect inequality line. (via)
Of course, in reality, the Lorenz curve won’t be at either extremes. The Gini coefficient is calculated based from the Lorenz curve.
Calculating the Gini coefficient using the Lorenz curve. (via)
Let’s call the area between the Lorenz curve and the perfect equality line as Area A, and the area between the Lorenz curve and the perfect inequality line as Area B. The Gini coefficient is then defined as:
Definition of Gini coefficient. (via)
If there is perfect equality, Area A will be 0. So the Gini coefficient is 0:
Gini coefficient when perfectly equal. (via)
Conversely, when there is perfect inequality, Area B is 0, and so the Gini coefficient is 1:
Perfectly unequal society has Gini coefficient of 1. (via)
So the Gini coefficient goes from 0 to 1, with 0 reflecting a society that is perfectly equal, and 1 reflecting a society that is perfectly unequal.
The Gini coefficient is used globally as a measure of income inequality. But comparing income inequality across countries using Gini coefficients is not straightforward. This is because different countries use different equivalence scales, definitions of household income, and household coverage.
What are equivalence scales?
They are used to adjust household incomes, allowing us to compare and analyse households of different compositions and sizes. Equivalence scales take into account the needs of the household increases with each additional member, but not proportionally so. For instance, while a household with one member uses one unit of electricity, but a household with four members doesn’t necessarily use four units of electricity.
There is a variety of equivalence scales and different countries use different scales. There are generally three models:
To explain each scale, let’s take a household with four members, two of whom are adults, two of whom are children, with a household income of $4,000 as an example.
Using the per household member scale, the “equivalence value” is 4 (i.e. the number of members in the household), and the equivalised household income is the household income divided by the equivalence value. So in the example, it’s $4,000/4, i.e. $1,000.
Using the modified OECD scale, the equivalence value is calculated by assigning 1 to the first adult member of the household, 0.5 points for each additional adult, and 0.3 points for each child. So in the example, the equivalence value is 2.1. The equivalised household income is the household income divided by the equivalence value, which is $4,000/2.1, i.e. about $1,905.
Using the square root scale, the equivalence scale is the square root of the number of members in the household. So in our example, the equivalence value is 2. So the equivalised household income is $4,000/2, i.e. $2,000.
Three popular equivalence scales. (via)
Internationally, there is no one standard equivalence scale recommended for general use. The OECD uses the square root scale to compare Gini coefficients across countries.
This video by Singstat is a wonderful resource to learn about the Gini coefficient:
How does Singapore measure up?
After accounting for taxes and social transfers, Singapore’s Gini coefficient calculated using the square root equivalence scale is 0.356, which is lower than that of US (0.39) and UK (0.36).
Comparison of Gini coefficients amongst different countries. (via)
So yes, it appears that Singapore, after our tax and social transfers, isn’t all that bad in terms of inequality.
It’s worth noting that the countries that have lower Gini coefficients after tax and transfers than Singapore have higher overall taxes than Singapore. Other than social transfers, Singapore is also unique in that we enjoy other subsidies through various government policy intervention, such as subsidies on HDB flats.
But inequality is more than just Gini coefficient
Two MPs who spoke in Parliament after Minister Ong pointed out that inequality is more than just the Gini coefficient. Mr Cedric Foo said that being obsessed with the Gini coefficient to measure inequality creates the wrong impression that “societal inequality can be reduced to just one number”.
Mr Murali Pillai raised the argument by Harvard University professor Steven Pinker measures of income inequality don’t necessarily fundamentally reflect well-being of households. Professor Pinker had highlighted that a decrease in inequality is not always good, because some of the most effective levellers of income equality include epidemics, massive wars and state collapse. In other words, one great way to reduce the Gini coefficient is to make everyone equally poor.
Extending from this point, Mr Murali highlighted that in tackling income inequality, the focus should be on addressing poverty and access to services like education and health. He pointed out that rising costs of living negatively affect low income households more than households with higher incomes. He reminded the Government to be mindful of such pressures, especially if the wages of households with low income do not keep pace with inflation.
I agree with what Mr Murali has said. It would be tragic if all Singaporean households have roughly the same income, but that income is low by international standards.
Reducing inequality while ensuring that the income of the poorest Singaporean households keep rising in a sustainable manner is a difficult challenge. If we get it right, then Singapore will have many more good years to come. If we get it wrong, then a sovereign Singapore might not exist for much longer. Let’s hope we get it right.
