A look at income inequality in Bermuda – Slices of the income cake & the Palma Index

Setting the Context

Not too long ago Craig Simmons, a lecturer in Economics at the Bermuda College, was featured in an RG article discussing inequality in Bermuda (and was clearly doing so after reading Thomas Picketty’s masterpiece on the subject).

In response to the warnings being raised by Mr Simmons, the Finance Minister Bob Richards essentially dismissed Mr Simmons thoughtful comments by saying there’s no statistics to back up Mr Simmons argument.

Which isn’t to say that Mr Simmons is wrong (and unless we’re invoking Bermudian exceptionalism, I’d imagine Mr Simmons is more right than wrong, seeing as the points he’s raised have been found globally), it’s just to say that the Government isn’t actively collecting or making available the statistics we’d need to ‘prove’ his points.

Not too long after this, Mr Wayne Furbert (PLP MP) proposed a Quality of Life study, which was rather crudely dismissed by the OBA Chair, Ms Susan Jackson, MP.

Both of these incidents got me thinking about what stats do we have which may illustrate inequality in Bermuda.

Historical Reviews

Both the 1991 and the 2000 census attempted to measure, to a degree, rates of inequality, and the Low Income Thresholds Study, is also informative on this (see Statistics for more information on these).

However, the matter of inequality seems absent from the 2010 census, and the Low Income Thresholds Study – despite a stated commitment to update it regularly – has not been updated since its 2007 publication.

So, I thought I’d try my hand at a first look at matters of inequality in Bermuda.

I’m not a statistician, and I only have limited time and resources (including limited access to data), so I’m only going to say that my findings are, at best, indicative of inequality in Bermuda.

Furthermore, they only speak to income inequality, not total wealth, which would include accumulated savings, inheritances, various assets (property, stocks, etc).

I’m pretty sure that if all those were included to make a more refined model, the inequalities in Bermuda would be much more unequal than the income inequalities alone suggest.

So, what did I do?

I decided that the data regarding income in the 2010 census was the best place to start.

This provides data on total income distribution, and then income distribution broken down according to race and sex.

I’ve decided here to focus on race rather than sex, although I think that’s important too.

The data shows population distribution according to 16 income bands, ranging from ‘$36,000 and below’ to ‘$750,000 and over’.

I’ve worked out:

  • The percentages of population distributions across these bands.
  • Plotted these according to a line graph.
  • Plotted both total racial population distribution and compared it to total racial income distribution.
  • Worked out a proxy of total income wealth distribution per income band (by adopting the mean for each income band multiplied by numbers of people in each band, adopting $36k and $750k for the first and last bands – which I imagine under-represents total wealth).
  • Condensed the 16 income bands into 10 in order to approach a decile breakdown of income distribution (combining income bands 2 and 3 into a single decile, and income bands 12 to 16 into a single decile).
  • Created a Palma Index for Bermuda’s income distribution, for Total population, Blacks, Whites, Mixed and Not Stated racial categories.

I decided to use a Palma measure of inequality simply because it was easier to do than other measures of inequality such as the Gini Coefficient or the Theil Index, and this article indicated it was actually a potentially better index than the others.

It essentially looks at the very rich and the very poor as an indication of inequality, largely ignoring the ‘middle’.

The Palma Index itself is determined by dividing the total wealth of the richest decile by the total wealth of the four poorest deciles.

A Short Discursive

Basically, the data reflects what many of us have known for a while…

  • The working class is overwhelmingly Black, the upper class is overwhelmingly White.
  • The majority of income wealth in Bermuda is commanded by a minority of the population.
  • We live in an unequal society, and, based solely on income distributions, a more unequal society than, say, the USA or the UK.
  • Income inequality is more pronounced within the White population (in that there are greater extremes of wealth inequalities there – there are a lot of VERY rich Whites, the bulk of Whites are ‘middle class’ but there’s also a number of very poor ‘small’ Whites) than in the Black population (Which has a minority of VERY rich Blacks, and the rest are almost evenly split between the lower and middle classes).

Once again, these results are only indicative of inequalities in Bermuda and only speak to income inequality.

Personally, my feeling is that if one were to include a wider metric, covering all facets of wealth, our inequality levels would be much greater than that reflected in income distribution alone.

It would be interesting to situate these figures historically too, to see if the income inequalities have changed over time, and under different governments, or if they’ve essentially stayed the same.  Perhaps someone could explore that angle in the future.

I’ll just put forward the hypothesis (and thus the challenge!) that under the PLP (so between the 2000 and 2010 censuses) inequality would have marginally decreased, and that Blacks increased their numbers in the upper income bands.

And now, the tables and graphs!

I worked these out on Excel, and I welcome people to go through my sums and check to make sure I’ve done it right.  It’ll take too much time to explain exactly what I did and why, but I hope it’s mostly self-explanatory.  I’ll try to answer any questions in the comments section, and look forward for any schoolboy errors being pointed out…

The Excel spreadsheet opens by clicking the below link:

Bda Income by Race

Figure One shows the racial compositions (in percentages) of each income band (disregarding the ‘Not Stated’ category).  WFigure Onehites make up about 30% of the lower class and the lower and mid middle class.  Blacks make up around 60% of these classes.  As one gets into the upper middle class (which arguably is ‘lower upper class really, based on the income band!) this begins to switch, with Whites increasing the composition and Blacks decreasing, with the income band of $235k-349k being roughly equal in racial composition.  Blacks then make up around 30% of the upper class, with Whites making up just over 60%.  Mixed/Other seem to be mostly steady at all income bands, at around 10%.

Figure Two shows theFigure Two racial composition of income wealth according to racial group.  Blacks actually command the majority of income wealth (at 47%), but this is lower than what would be expected if income wealth was equally distributed according to total population size of each racial group.  Blacks make up 51% of the population in question, but only command 47% of the income.  Whites only make 32.4% of the population, but command 41% of the income.  Mixed/Other seem more equal, being 11.7% of the population, and commanding 11% of all income.  Not Stated make up only 0.72% of the population, but command 1% of all income.  In other wordFigure Threes, Whites command 1.27 times their ‘fair’ share of income wealth, and Not Stated command 1.4 times their ‘fair’ share.  Blacks command only 0.9 times their fair share, Mixed/Other 0.94 their fair share.

Figure Three shows the population breakdown of each decile of income band (I use the term ‘decile’ loosely here).  If each decile commanded their fair share of income wealth, the income distribution should mirror this.  It doesn’t.

Figure Four shows the totFigure Foural income slices for each of those same deciles as in Figure Three.  The richest deciles (nine and ten) alone command 60% of all income wealth in Bermuda, while only making up 30% of the population.  Decile ten alone, at 14% of the population, command 38% of all income wealth.  Or to flip it around, 70% of the population only commands 40% of all income wealth.  And if one looks at just deciles one to six, which represent 50% of the population, they only command 22% of all income wealth.  It’s not quite the 99% versus the 1%, at least in terms of income distribution, but it certainly is inequality.

Figure Five shows the total income wealth according to the Palma format, showing the income distribution between the pooresFigure Fivet four deciles, the middle five deciles and the top tenth decile.  According to the Palma equation, where one divides the wealth of the top tenth decile to the wealth of the lowest four deciles, Bermuda has a Palma Index, for income distribution, of 2.65.

Figure Six shows the income distribution within the Black population, along the same Palma format as in Figure Five.  The Black population has a Palma Index of 1.56.

Figure Seven shows the income distribution within the White population, as above.  The White population has a Palma Index of 5.68.

Figure Eight shows the income distribution within the Mixed/Other population; they have a Palma Index of 2.

Figure NineFigures Six to Nine shows the income distribution within the Not Stated population; they have a Palma Index of 2.25.

Figure Ten shows Bermuda’s Palma Index in reference to other countries (strictly these are not comparable, as I’ve only measured income, not total wealth; I include it here solely as an indication of our minimum inequality in comparison with other countries).  Based on this, we’re a more unequal country than the USA and the UK, as well as Malaysia, Venezuela, China, Morocco, Russia, Tanzania, Albania, India, Bangladesh, Egypt and Denmark.  Jamaica is the most unequal country, followed by South Africa.

I have a gut feeling that a truer metric of our inequalities (including those beyond income) would see Bermuda place somewhere in between South Africa and Brazil.  I leave that as a hypothesis and a challenge for others.

Figure Ten


9 thoughts on “A look at income inequality in Bermuda – Slices of the income cake & the Palma Index

  1. “I’ll just put forward the hypothesis (and thus the challenge!) that under the PLP (so between the 2000 and 2010 censuses) inequality would have marginally decreased, and that Blacks increased their numbers in the upper income bands”.

    That’s probably true, if only because (a) the economy grew in those years, and (b) no doubt many of those who became employed in both the “local” sector (e.g. Law firms, Accountants etc) and also IB, were coming out of unemployment into work.

  2. The fact that Bermuda’s black majority continues to find itself behind in terms of wealth inequality is the legacy of Bermuda’s racial past. The great irony in this is that if education is suppose to be the pathway to upward mobility; the black population which statistics have shown to have more educational degrees than the white population yet continues to lag behind when it’s comes to access to higher paying jobs and wealth acquisition. This also had it’s impact on the bringing in of white immigrants which are able to take advantage of the system of white privilege which has been to the detriment of the natural aspirations of Bermuda’s black population. Hence the great divide between the black and white view of the government’s new policy concerning PRC pathway to Bermuda status which it now supports and why it continues a focal point of sharp disagreement as it relates to income inequality in Bermuda.

  3. There are a lot of wealthy Black Bermudians if you include assets such as property. I suspect in the middle and working classes (I hate those expressions), that Black Bermudians are more wealthy than whites.

    I certainly don’t own my own property. One of the great unwashed that have to rent, and always will.

    I suspect that White immigrants – in the main – do not have the ability/opportunity etc to acquire property let alone the desire to pay inflated prices.

  4. Hi Mike,

    I agree that this analysis only provides part of the picture. At best it indicates – and conservatively so – the income inequalities only. To the degree that income inequalities may serve as a proxy for total wealth inequalities, it may be seen as indicating that too – although as stated, I think it generally underestimates inequality in Bermuda.

    I’m going to see if I can’t do an analysis, even a limited one, of wealth/assets inequalities, but I’m not confident that I’ll be able to access the relevant data. I’ll take a look.

    I imagine home ownership is quite high within the Black population, although the question then becomes to what quality (in terms of monetary value) these homes compare to other groups. After all, I’m sure most of us would agree that the property value of homes in Fairylands are an order of magnitude greater than the property value of homes back-a-town.

  5. I just wanted to note here that, in a discussion on Facebook, based on this article, some have argued I’ve used the wrong data set.

    I selected what I felt was the most appropriate one, but I will look at the data set they propose and see if I can perform a similar analysis on it.

  6. Pingback: A look at income inequality in Bermuda – Gini Coefficients & Lorenz Curves | "catch a fire"

  7. If the people around 42nd Street swapped houses with the people in Tucker’s Town, I predict that within 3 years Tucker’s Town would be the new 42nd Street and 42nd Street would be the new Tucker’s Town. So what are you trying to prove when you say that houses in Fairylands are worth more than houses in the back of town?

  8. One of the questions to be asked, is to what factors does one attribute income disparity within each race, i.e. within Whites and also within Blacks.

    Somewhere within that lot must be “old money” gained by families in years gone by – however obtained.

    1. Should wealth inspire us to achieve more, the American Dream concept?
    2. To what extent does lack of ambition – lack of education – lack of “being in the right place at the right time” contribute?
    3. Would we have a worse situation if we didn’t have the creative power of a free market (warts and all)?
    4. To what extent do we want “sameness” by reducing the income gap?

  9. Hi Justin,

    I was discussing that all the graphs show is just income inequalities in Bermuda; they don’t show a full picture of total wealth inequalities, although they likely give an indication of such inequalities.

    In order to get a more complete picture, one would have to complement the data on income inequalities with one on wealth and assets, including physical property such as houses. I believe Mike had argued that a large proportion of Blacks own houses, and he appeared to argue that would moderate the inequalities. My point however is that even if the majority of Blacks own their houses, those houses may be an order of magnitude lower in value than those of Whites – I gave the extremes for an example. It may not be such a drastic difference in value, but it might be. Until we see the data we cannot say.

    As such, all I’m ‘trying to prove when [I] say that houses in Fairylands are worth more than houses in the back of town’ is that simply owning one’s house doesn’t necessarily negate the inequality gap, if even a handful of the population own houses worth more than a thousand of the population combined, for example.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s