I have figured out a formula in order to calculate the area under the graph for these lines. I will explain it for the blue line, but this will work for the orange line as well. Step 1: Take the sum of the column '% of population'. Step 2: Multiply this number with 10. Step 3: Subtract 500 of this value. And now you have the same answer as in the example. With this method, you kind of integrate the area under the curve!
Sorry but, I have a question on how this method applies to my data. I am compiling information on income distribution of 30 countries and they don't offer me the possibility to form triangles in any spot. For example, here is the data for Burundi: (0;0), (10;0.3), (20;1.9), (30;5.2), (40; 9.2); (50; 4.7), (60; 20.2), (70; 27.7), (80; 37.3), (90; 53), (100;100). I don't have any point where there are straight lines so I can form traingles. How could I calculate the area of the lorenz curve in this situation? Thank you very much.
It seems silly to calculate the area of all the shapes under the Lorentz curve. Wouldn't it make more sense to do do power regression or exponential regression of the observed data and then integrate the function to find the area under the curve?
It might, if you have a function for the Lorenz Curve. But in the real world you will likely have data broken up by deciles or quintiles in which case calculating the area of triangles is going to be your best bet.
that makes sence. However, if the observations of the income of the population wasn't set in jumps from 10 to 20%, but rather 1 to 2%, then we would have a 100 observations, but most likely also a smoother curve, which would make it very tedious to calculate the triangles between every observation.
That's true. And, of course, for modeling purposes, you would want to have a smooth function that you could easily integrate. One of the interesting things that people who study this have discovered is that the income distribution for most people most closely resembles a log-Normal distribution while the income distribution at the very top is best modeled with a Pareto power distribution (as is wealth). Some physicists who have turned their attention to the question have viewed this as a natural law, but it seems more likely to be the result of human-made institutions.
Thanks for the explanation. I'm curious... Isn't A+B always 5000? Why then do we have do divide A with this static number, instead of just using A as our figure for comparison? (My guess is that while not necessary for comparisons it's more convenient and pedagogical to have a number between 0 and 1 instead of one between 0 and 5000 ?)
No, you need the Lorenz Curve to calculate the Gini coefficient. If you have equations describing (or estimating) the Lorenz Curve you can use calculus, but that isn't how real data works.
@@liammalloy sir i was saying by using percentage data? Actually That are in the boxes, sir i am thinking that this method is bit difficult 😅😅 and . Ya I don't have equations.
Sir we are not finace economs to make gini coefficients and the correlation between working places and amount of unemployment. We are just idiots in Parlament .
I have figured out a formula in order to calculate the area under the graph for these lines. I will explain it for the blue line, but this will work for the orange line as well. Step 1: Take the sum of the column '% of population'. Step 2: Multiply this number with 10. Step 3: Subtract 500 of this value. And now you have the same answer as in the example. With this method, you kind of integrate the area under the curve!
could you please show this way of solution in a process in numbers? I just cant get it. thank you!
A Million Thanks to u Brother ...💯
I was literally wandering for the Concept Clarity 😭...
[ Respect from India 🇮🇳 🙏]
Very easy and concise explanation. I really like it.
Thank youuuu so much!,... I needed this
OMG 😭😭😭😭😭😭😭😭 thank you thank you thank you !!! this id sooo helpful !! God bless you
Excellent and perfectly explain
great explanation .....so cleared .... thank u very much
Excellent Video.. Thank you.. and expecting some more videos
So clear... thank you!
thank you so much for the video, it helped a lot
Thank you! This is so helpful.
Thank you! Great explanation !!
Sorry but, I have a question on how this method applies to my data. I am compiling information on income distribution of 30 countries and they don't offer me the possibility to form triangles in any spot. For example, here is the data for Burundi: (0;0), (10;0.3), (20;1.9), (30;5.2), (40; 9.2); (50; 4.7), (60; 20.2), (70; 27.7), (80; 37.3), (90; 53), (100;100). I don't have any point where there are straight lines so I can form traingles. How could I calculate the area of the lorenz curve in this situation? Thank you very much.
It seems silly to calculate the area of all the shapes under the Lorentz curve. Wouldn't it make more sense to do do power regression or exponential regression of the observed data and then integrate the function to find the area under the curve?
It might, if you have a function for the Lorenz Curve. But in the real world you will likely have data broken up by deciles or quintiles in which case calculating the area of triangles is going to be your best bet.
that makes sence. However, if the observations of the income of the population wasn't set in jumps from 10 to 20%, but rather 1 to 2%, then we would have a 100 observations, but most likely also a smoother curve, which would make it very tedious to calculate the triangles between every observation.
That's true. And, of course, for modeling purposes, you would want to have a smooth function that you could easily integrate. One of the interesting things that people who study this have discovered is that the income distribution for most people most closely resembles a log-Normal distribution while the income distribution at the very top is best modeled with a Pareto power distribution (as is wealth). Some physicists who have turned their attention to the question have viewed this as a natural law, but it seems more likely to be the result of human-made institutions.
why you start to 40 and not to 60 in gini coefficient
May Allah bless your mother
Thanks for the explanation. I'm curious... Isn't A+B always 5000? Why then do we have do divide A with this static number, instead of just using A as our figure for comparison? (My guess is that while not necessary for comparisons it's more convenient and pedagogical to have a number between 0 and 1 instead of one between 0 and 5000 ?)
That's also my question because 5,000+3,150=8150...gini index =22.69%
Thanks you, amazing!
why didn't he just do it using trapeziums
Thank u so much sir,,❤
can i ask? y u start to 40 and not to 60?
Excellent presentation...Thank u sir for clearing my doubts .
Can i get your literature broo?
Sir any other way to find gini coefficient?
No, you need the Lorenz Curve to calculate the Gini coefficient. If you have equations describing (or estimating) the Lorenz Curve you can use calculus, but that isn't how real data works.
@@liammalloy sir i was saying by using percentage data? Actually That are in the boxes, sir i am thinking that this method is bit difficult 😅😅 and . Ya I don't have equations.
@@rachittyagiyoutube You can use any data you have. Percentages, dollars, etc. You just need to make sure you use consistent units.
@@liammalloy ok sir
What did you use to make the graph
Ameerah, I just used excel. The key is that the values have to be the cumulative amount of income or wealth for the bottom 10%, 20%, 30%, etc.
Perfect
Sir we are not finace economs to make gini coefficients and the correlation between working places and amount of unemployment. We are just idiots in Parlament .
Hekoed a kitm thanks!
I love u
thanks a lot