India is experiencing the same phenomenon. Absolute poverty has dropped, but inequity has grown.

Dear Emmanuel,

Thank you for your response.

How one can explain that phenomenon ? Can you suggest any paper discussing the same situation and potential explanations?

Cheers,

Davit

It is not strange at all. I recorded the same scenario for Nigeria in a study

Dear Akinyemi,

Can you please explain me why that can happen ? What are your arguments regarding this indirect relationship?

Best regards,

Davit

I believe this is a first step for creating an inclusive social order. When the inequality reduces, then the differences among the communities decrease, leaving to increase social capital among them. Further, when all the people can access markets, as all of them are equally positioned in the income scale, then it is good for the society and wellbeing. All of them might still be relatively poor, but the increased social capital, the equality exists among them and understanding about each other all would possibility help them to enhance their incomes. You may would like to refer to welfare economics theories which talk about subjective wellbeing etc. Goodluck with your research.

Dear Rama,

Thank you for your nice comment.

Can you suggest any paper in which a similar observation is discussed ?

Just if you know. I need to somehow explain this case

Cheers,

Davit

I think that such changes are not impossible, but I also think that it is difficult to measure, especially as national poverty data is influenced by governments, as e.g. in India.

Dear Eberhard,

Thank you for your comment.

Let me clarify that the survey was undertaken by UNICEF in Georgia. It is a longitudinal survey, which is done in every two years since 2009. 2017 data is exceptional, because it shows that poverty measured in terms of consumption increased. HH average expenditure decreased in real terms (deflated using Consumer price index) by 10%, whereas HH income increased by 18% in real terms on average. Gini measured in terms of income PAE shows slight decrease over the last two years (from 0.43 to 0.42). While Gini measured in terms of consumption stayed at the same level.

In my understanding it is not right to compare inequality measured in terms of income and poverty rates measured in terms of expenditure. Sample size is quite large 4,697 households. I also tried some validation like retaining panel Households which participated in 2015 and 2017. Then I looked at % change in terms of expenditure in decile groups. Certain components of expenditure decreased (especially spending on food items).

I do much appreciate if you can share your thoughts about it. How one can explain such a scenario?

I have attached photos of income and expenditure distributions in decile groups. Please note it is panel HH data.

Cheers,

Davit

Isn't this the same trend, i.e. increased inequality and decreased poverty, that Pilketty in his book Capital, talked about? Haven't read the book yet, just commentaries on it. If i remember my MA thesis(long time ago), on the measurement of inequality, Sources of inequality, would be. structure of economy (eg relative share of rents vs. wages, capital vs. labor gains, agriculture vs. manufacturing), while poverty, esp. in terms of thresholds, more sensitive to prices, education and labour market considerations. Reduction in food as share of income signifies some progress and reduced poverty, right? If you read The Economist (or World Bank studies), they have been pointing to this trend with stats. There have been tremendous gains in poverty reduction for the past decades. The challenge in reducing it further is in reaching the 'last mile', as well as those vulnerable to shocks, eg ill health, staple prices. Global inequality is of course related to rent-seeking, hiding money in safe havens, CEO salaries vs. wages, etc..

Dear Maria,

Thank you for your comments.

The case here is that poverty rates increased while inequality measured in terms consumption PAE did not change. In the same study inequality measured in terms of income PAE decreased slightly on national level. HH income increased but consumption decreased for panel households (the same households participating in two period of the survey. If we look at the distribution of different categories of consumption food decreased the most in the poorest decile group. While looking at income distribution in decile groups it shows increase. The question is why households started to spent less when they have more income? and why poverty rates are increased ?

Since income increased and expenditure decreased one may assume that they are saving so why poverty measured in terms of consumption increased. It is not a case at all.

In the previous round of the survey, HH income was 75% of HH expenditure. So they spent more with less income. In that period, income did not increase in real terms. Two years later, income increased substantially, while expenditure also have increased so much that it exceeded the income and balance between these two measures decreased. 3 factors contributed to the overspending pattern of households. Inflation especially (on food, healthcare, alcohol and tobacco, utilities), very strong devaluation of local currency and slow down of economic growth by 50-60%. In another words, HH started to spent less since they spent all their savings or

money borrowed from financial institutions and cent. So they are left with less money to spent on basic needs.

Do you have in mind any other sources that can explain such a scenario? Especially regarding inequality?

Best regards,

Davit

if I recall my colleagues were wont not to use income, but rather expenditures for their studies on economic well-being. Incomes tend to be understated among well off and overstated in others. Expenditures, which may suffer from recall problems, would reflect more the economic state of households. It is possible they spend more than their incomes. With the economic shocks and downturn, indeed, they will still be reflecting more expenditures, more so, even as their incomes do not increase, or may have gone down in real terms. This sounds like common sense; that is, one has to spend in order to survive right? No matter if you don't have income. In poor countries, you have other means of sourcing your necessities.

As for references, check out recent Development Economics textbook, particularly one citing recent empirical studies. I think i perused one by de Janvry. There are other authors as well. Sorry, i'm outside of homebase currently so cannot recall the books on the shelf.

Another way, is looking up articles that used panel data on income and expenditures. If it is related to health, check out the work of Angus Deaton (Nobel winner); i think he had a broad historical piece on changes in well being. Or the works of Esther Duflo. She's best.

Dear Maria,

I fully agree that expenditure is a better measure of poverty than income. All poverty rates in the present analysis are measured in terms of consumption PAE.

I also agree that households tend to spend more with less. Many other studies show the same trend. That is a normal behavior due to positive expectations. Maybe you take a loan and live now and pay later just because you expect that you earn some income in the future. In 2015 round, income was observed to be 75% of consumption. In 2017, it is observed that income is 97% of consumption. Some people may think that it is a good development that they started to spend less and earn more so they are saving. But then how poverty rates could have gone up. The logic here is again what I mentioned. Since in 2015 HH tend to spend more with less and overspending was compensated by loans or savings. Since 2015 happened shock in Georgia. Local currency depreciated strongly against the US dollar (on average 30%), prices on food and non-food items started to increase and real economic growth slowed. According to the data, real income increased by 18%, while consumption dropped by 10%. Even though real income increased (which already is corrected for annual inflation) still prices on individual products during the survey period increased substantially. Moreover, local currency depreciated against the dollar pushing HH expenditure up since some people had loans in dollars. The cost of living increased over the last two years and HH tend to spend more. So in 2017, balance between income and expenditure decreased. Increase in income was not translated to HH prosperity due to economic shocks which lead them to change the structure of their consumption.

We also measured inequality in terms of income and consumption. The first showed decrease in inequality and the second measure no change. Of course inequality measured in terms of income is not comparable for poverty measured in terms of consumption. But I am trying to understand the role of reduced income inequality in present case

Thank you for your contribution and nice comments

I will check for that author

Cheers,

Davit

The book: Development Economics: Theory and Practice is by A de Janvry and E Sadoulet. Published by Routledge.

Oops, i got confused. The general question for the topic above was: Is it possible for poverty rates to decrease and inequality to increase? Your last sentence in the recent message is 'understand the role of reduced income inequality'. So, did inequality increase or decrease? To explain either way is to cite structure of the economy and the share of different factors of production. It could be possible government expanded, bringing in more wage earners vs. rent earnings, in which case inequality will be decreased. Being agricultural can also influence inequality, either way depending on mechanization; but agriculture households in poor economies will usually form the bulk of the poor. Although note that 4 years may not bring much changes in structure. You just need to describe current structure and how it sustains or lessens inequality. I suspect, if not data issues, the problem will be that your poverty basket may have become more expensive, but with the reduced inequality which you said may arise from the increased real incomes, households were able to cope. When you say households spend 25% more than incomes, then this declined to 3%, i can only read this to mean non-income sources (which has allowed them to spend more than incomes in the first place) drying up, therefore poverty must be increasing. I am sure a lot is 'lost in the translation', so do simple charts for trends first, then if this does not help, just find something from the references to explain similar or contradictory trends.

alternatively, You can also do Ginis for health inequality, with the extent of inequality signified by how far or close to the diagonal is health care spending. You can read early works of A Wagstaff on this.

Dear Maria,

Thank you for sources. It is much appreciated. Yes income inequality measured in terms of Gini index decreased

(Gini=0.43 in 2015 ; Gini = 0.42 in 2017).

Consumption inequality did not change

(Gini=0.36 in 2015; Gini=0.36 in 2017).

However, consumption poverty rates are increased. This is a case that I have.

Not only non-income source could be case here. Since overspending can be compensated not only by non-income sources but also taking loans, borrowing in general and dissaving. Survey also shows that over this two years especially in the poorest groups borrowing increased substantially which is a good indicator since Georgia faces serious household indebtedness. Especially online micro finance organizations give people money without any prove of income source.

I thought measuring Gini in terms of different income components. I will do that and share with you the insights

I think we have a very interesting discussion here. Thank you for following

Cheers,

Davit

It's a matter of logic. The old sayihng,"the rich get richer and poor get poorer?" is, instead, "the rich get richer, the poor get less poor." In other words, as has happened here in US, the gap between rich and poor was already very wide (10% of pop. were the wealthiest, had the most wealth in the nation), so even as those in the lowest percentiles increased in wealth/less poverty, the number in the upper wealth perecentiles inrceased, thus wideing the gap, i.e growth in inequality. In addition, the middle class wealth stays at the same level, while poverty at the lowest levels decreases and the wealth at the other end increases. See:

https://www.theatlantic.com/business/archive/2017/09/new-census-data-shows-more-americans-emerging-from-poverty/539589/

well. This is the same with Pakistan as well. Poverty has been decreased with an increase in inequality.

Dear Sanwal,

Do you have in mind a paper that describes the same case?

Best,

Davit