Creating BR Country Index

Locked
ramana
Forum Moderator
Posts: 59773
Joined: 01 Jan 1970 05:30

Creating BR Country Index

Post by ramana »

This thread is being started to collect ideas on creating our own BR Idex for countries.

A series of x-posts are presented to give an idea of how it has evolved:
A_Gupta wrote:What set of quantitative measures and trends in those measures would you use to characterize Pakistan's failure? E.g., what measure and trend line, which if it steepens you would say, failure is intensifying, and which if they flatten, you would say, failure is on hiatus?
shiv wrote:
A_Gupta wrote:What set of quantitative measures and trends in those measures would you use to characterize Pakistan's failure? E.g., what measure and trend line, which if it steepens you would say, failure is intensifying, and which if they flatten, you would say, failure is on hiatus?
The only figures that I tend to rely on are human development parameters. Birth rate, infant and maternal mortality, population growth rate, fertility rate, literacy, poverty etc.

These are dry statistics but they are all measures of human misery. Polls and the media tend to exclude what is happening to women and children (say below 10). Women and children below 10 make up a whopping 70% or more of Pakistan - about 100 million or so.

Once you have a population like this governance becomes a problem because large numbers of people don't really care about anything as long as they can get a square meal and some shelter. Furthermore - social structures (families) tend to get broken because men have to go away to find work (or become criminal) leaving women and children defenceless. This sort of society is open to takeover by religion and Islam is perfect. But what happens is that the "islam" that takes over has a local leader for local issues and he is a competitor to the state because the state is providing nothing, and this guy is at least providing succour. If the latter leader also has arms, then the state has a bigger problem getting him under control. So you actually have a virtual or real splitting up of a state.

This is Pakistan. Only nobody believes it yet.
A_Gupta wrote:
Birth rate, infant and maternal mortality, population growth rate, fertility rate, literacy, poverty etc.
If we can close in on what is in the "etc.", then
a. We can start gathering statistics - say from 1990 onwards
b. We can decide what weights to give each of the factors mentioned above, and come up with a single composite failure index.
c. We can compute the failure index over the past; and we can update the failure index whenever a new measurement of one of the components is made.
d. Since we will then have a numerical measure, we can compare our qualitative assessment with the failure index trend, and try to tease out factors we may have overlooked.
shiv wrote:
A_Gupta wrote: Birth rate, infant and maternal mortality, population growth rate, fertility rate, literacy, poverty etc.
If we can close in on what is in the "etc.", then
a. We can start gathering statistics - say from 1990 onwards
b. We can decide what weights to give each of the factors mentioned above, and come up with a single composite failure index.
c. We can compute the failure index over the past; and we can update the failure index whenever a new measurement of one of the components is made.
d. Since we will then have a numerical measure, we can compare our qualitative assessment with the failure index trend, and try to tease out factors we may have overlooked.
Good idea. One problem is lack of accurate statistics from Pakistan. The 1990 census was held in 1998. The 2001 census was not held at all. Don't even ask about 2010 Still they can be obtained.

I did a lot of this stuff while writing my ebook and I still collect stats and tend to follow them. In 10 years the population has gone up by 35 million (20%) but literacy has not gone up significantly, nor has poverty come down. Only the population numbers are up.

Need to keep in mind that these numbers are "strategic secrets" for Pakistan. Pakistan has more poor people now than the entire population of Pakistan in 1947. This is true of india as well - I say it before someone points it out. For this reason the race is to improve your human development statistics so fast that you will eventually (at some future date) "catch up" with the population increase.
ramana
Forum Moderator
Posts: 59773
Joined: 01 Jan 1970 05:30

Re: Creating BR Country Index

Post by ramana »

RajeshA wrote:
A_Gupta wrote:What set of quantitative measures and trends in those measures would you use to characterize Pakistan's failure? E.g., what measure and trend line, which if it steepens you would say, failure is intensifying, and which if they flatten, you would say, failure is on hiatus?
1) Human Development Index (HDI). The HDI is used by the UNDP. It consists of
  • Life Expectancy Index
  • Education Index
  • GDP per capita at PPP
134. India 0.612
141. Pakistan 0.572

2) The World Bank uses a The Worldwide Governance Indicators (WGI). It uses six indicators
  • Voice and Accountability
  • Political Stability and Absence of Violence
  • Government Effectiveness
  • Regulatory Quality
  • Rule of Law
  • Control of Corruption
Please feel free to use the tools. You can call up the various indices for a single country for certain years or make comparisons between countries based on a single index. You can produce Excel sheets if you want.

3) Terrorism
There may be several sources, especially Indian, however one relatively reliable source is the National Counter Terrorism Center in USA. Every year they compile the statistics.
2008 Report on Terrorism (pdf)

Fatalities in 2008 in Terrorist related violence:
  • 1. Iraq: 5016
  • 2. Pakistan: 2293
  • 3. Afghanistan: 1989
  • 4. Somalia: 1278
  • 5. India: 1113 :evil:
Kidnappings in 2008
  • 1. Pakistan: 1264
  • 2. Afghanistan: 584
  • 3. Iraq: 371
  • 4: India: 366 :evil:
4. Small Arms Proliferation
There have been some surveys. Here is one by Regional Center for Strategic Studies, Colombo
SMALL ARMS AND HUMAN INSECURITY: Reviewing Participatory Research in South Asia (July 2002) (pdf)

5. Religious Education
There are some numbers flying around. Here is one proper study by Harvard University; Pamona College, Claremont, California; and World Bank - Madrassa Metrics: The Statistics and Rhetoric of Religious Enrollment in Pakistan

A good article The AfPak Madrassa Threat: What Are We to Believe?

6. Population Explosion
One can look at UN Population Division or in Wikipedia

In 2050 A.D.
  • 1. India: 1.6 billion
  • 2. China: 1.4 billion
  • 3. United States: 439 million
  • 4. Pakistan: 309 million
7. Not all indicators can be quantified as such. I would consider the fear psychosis of a country very important. How quickly an area adopts the Talibanic way, starting from when the Taliban start to openly threaten and intimidate people in the area, gives an indication of how quickly the whole country could fall. Only certain regions would be immune to an immediate surrender, but most of the country would start to obey the code quite quickly.

Hope it helps!
A_Gupta wrote:We can create our own BRF Index I think, relying on such measures as mentioned by Shiv and RajeshA.

To the list, I'd like to add (if the data is available)
1. %age unemployment/underemployment among males 18-45 age group
2. Males 18-45 age group as %age of the total population

The above 2 measure the potential pool of troublemakers/revolutionaries/jihadis/criminals.

Then I'd like to add something to do with unequal income and asset distribution.
1. Gini coefficient of income distribution
2. Gini coefficient of asset distribution (this simply may not be available :( )
A_Gupta wrote:Will do. I'm thinking of a multiply-owned, publicly accessible Google Docs spreadsheet as the primary document. Will lay out a template and share. I assume we'll have some secure way of exchanging gmail IDs so that the volunteers can be set up as owners of the document.
A_Gupta
BRF Oldie
Posts: 12056
Joined: 23 Oct 2001 11:31
Contact:

Re: Creating BR Country Index

Post by A_Gupta »

There is an important post from Shiv on the "Managing Pakistan's failure" thread, an excerpt of which I'm quoting here:
Pakistan has more poor people now than the entire population of Pakistan in 1947. This is true of india as well - I say it before someone points it out. For this reason the race is to improve your human development statistics so fast that you will eventually (at some future date) "catch up" with the population increase.

This is a slightly complex calculation - I am not sure I can do it. A statistician/demographer probably could - but readymade stats are available on the net - I have often quoted them

You need to look at
1) the rate of increase of population
2) the rate at which the rate of increase of population is being reduced (meaning that if the growth rate was 3.5 in the last decade and it is 2.8 per thousand in this decade - the rate has fallen by 20% in 10 years) You might want to extrapolate that same rate of fall of growth rate to compute decadal population growth rate in 10 years and 20 years form now.
3) Using those population statistics see the literacy rate 10 years ago and what it is now and assume the same rate of increase in literacy and make a prediction of how many are going to be literate in 10, or 20 years of the rate of increase of literacy remains the same.
4) From this compute the number of extra schools and schoolteachers needed and check the finances required for that. Check if the rate of increase in numbers of schools and teachers matches the required number and then see if the same rate of increase in literacy can be maintained over the next decade.
5) Ditto for all other parameters
I think the key insight here is that a nation is failing if the (absolute size of the problem)/(the absolute resources available to solve the problem) is a quantity that increases with time.

So e.g., the absolute size of the maternal mortality problem may be going down because of a better mortality rate and/or because fewer women have births. This is not reflected in maternal mortality rates statistics. If I have to find medical care for 10 million women now and only 9 million women a year ten years from now, though I may not have made any progress in the maternal mortality rate, the problem I have to solve has shrunk, and I am more distant from failure after ten years than I am now.

In general, "problem rate" * "relevant population" might serve for absolute size of the problem.
GDP, or (government budget - amount spent on debt service) might be a proxy for the "resources available".

Coming up with good measures is going to require some research and experimentation. Let's not go half-cocked into data gathering straightaway.
brihaspati
BRF Oldie
Posts: 12410
Joined: 19 Nov 2008 03:25

Re: Creating BR Country Index

Post by brihaspati »

There are some standard ways of meta-modeling relevant here: I actually think that raw data of the widest variety possible is the way to to start.

step 1 : make a list of all relevant quantities you can think of [some discussion as to relevance needed]
step 2: make a time series [multivariable] say annual
step 3: run within year, and cross year, or summed correlation estimates. This will give a first order idea of linkages.
step 4: depending on this initial study form tentatiev ideas of dependent variables and independent one
step 5: run multivariable statistical models - say a linear factor model, linear regression, principal components etc. modeling weighted linear combinations of independent variables as explanatory for dependent ones.

This step itself could give a weighted index that has predictive value.
shiv
BRF Oldie
Posts: 34982
Joined: 01 Jan 1970 05:30
Location: Pindliyon ka Gooda

Re: Creating BR Country Index

Post by shiv »

The least I can do would be to trawl for statistics and dump them on here. Will do that.
shiv
BRF Oldie
Posts: 34982
Joined: 01 Jan 1970 05:30
Location: Pindliyon ka Gooda

Re: Creating BR Country Index

Post by shiv »

The following appears to be a great resource for country comparison and lists all sorts of details that you can call up and make country comparisons
http://www.nationmaster.com/index.php
arnab
BRFite
Posts: 1136
Joined: 13 Dec 2005 09:08

Re: Creating BR Country Index

Post by arnab »

I believe we were looking at constituents of 'middle class' between India and Pak. I think this paper has some great statistics.
What is middle class about the middle classes around the world?

Abhijit V. Banerjee and Esther Duflo
http://econ-www.mit.edu/files/3107
shiv
BRF Oldie
Posts: 34982
Joined: 01 Jan 1970 05:30
Location: Pindliyon ka Gooda

Re: Creating BR Country Index

Post by shiv »

From Nationmaster
http://www.nationmaster.com/index.php

If you put these stats in a graph form you will know the slope and how each country's slope compares. You can also put a "target" and see when the target will be reached from the slope.

2003 - Birth rate
  • India -23.28 per thousand
    Pakistan - 29.59 per thousand
    Bangladesh - 29.9 per thousand
    Sri Lanka -16.12 per thousand
    Nepal - 32.46 per thousand
2008- Birth rate
  • India -22.22 per thousand
    Pakistan - 28.35 per thousand
    Bangladesh - 28.86 per thousand
    Sri Lanka -16.63 per thousand
    Nepal - 29.92 per thousand

-------------------------

Population growth annual % - note how Pakistan's has hardly shown any decline since 1995

1975
  • India - 2.28
    Pakistan - 3.17
    Bangladesh - 2.36
    Sri Lanka - 1.61
    Nepal - 2.21
1985
  • India - 2.04
    Pakistan - 2.68
    Bangladesh - 2.39
    Sri Lanka - 1.51
    Nepal - 2.3

1995
  • India - 1.78
    Pakistan - 2.46
    Bangladesh - 2.18
    Sri Lanka - 1.36
    Nepal - 2.51
2005
  • India - 1.37
    Pakistan -2.41
    Bangladesh -1.86
    Sri Lanka - 0.84
    Nepal - 2.02
-------------
Annual growth of population (percentage increase in size of population per year)

2003
  • India - 1.47%
    Pakistan - 2.01%
    Bangladesh - 2.06%
    Sri Lanka - 0.83%
    Nepal - 2.26%
2008
  • India - 1.578%
    Pakistan - 1.999%
    Bangladesh - 2.022%
    Sri Lanka - 0.943%
    Nepal - 2.095%
-------------

Number of births per woman

1995
  • India -3.4
    Pakistan -5.2
    Bangladesh -3.77
    Sri Lanka - 2.2
    Nepal - 4.62
2005
  • India - 2.84
    Pakistan - 4.12
    Bangladesh - 2.98
    Sri Lanka - 1.91
    Nepal - 3.46
brihaspati
BRF Oldie
Posts: 12410
Joined: 19 Nov 2008 03:25

Re: Creating BR Country Index

Post by brihaspati »

TFR Life Expec(F/M) growth-rate ending-popn(mill)
1961-65 6.8 43.3/44.4 2.6 48.7
1966-70 6.8 47.1/47.7 2.8 56.0
1971-75 6.8 50.8/51.1 3.0 65.1
1976-80 6.8 54.5/54.5 3.2 76.4
1981-85 6.8 58.2/57.9 3.4 90.6
1986-90 6.6 61.4/60.5 3.5 107.8
1991-95 5.7 61.4/60.5 3.0 125.2
1996-00 4.8 61.4/60.5 2.6 142.3

G Feeney, I. Alam, (Sept, 2003) Estimates and Projections of Population growth in Pakistan, Population Council, p485.

TFR decreases sharply from constant high for decades???? All the while LE remains constant!!!!

Indirect estimates required.
abhischekcc
BRF Oldie
Posts: 4277
Joined: 12 Jul 1999 11:31
Location: If I can’t move the gods, I’ll stir up hell
Contact:

Re: Creating BR Country Index

Post by abhischekcc »

Let's clarify first - what should be the purpose of the index:

Is it going to be purely political risk index (country risk index), or a political-economic combined index (country and investment risk index), or a socio-political-economic index?

This is important because the way I see it, BR's comparative advantage (over other index makers) is in political and related matters, not in economic topics standalone.

------------

If we decide to have a composite index covering several aspects of a nation's life (not just political), we could have several sub-indices to go with this general index - terror risk index, political uncertainity, medium term or long term risk index, etc.

--------

And we also need to consider the matter of who will give how much time to construct and maintain the indices - before we decide how complex to make the whole system.

---------

Some suggestions for the kind of indices we can build:
1. Freedom of religious expression. This is important to counter western hypocrisy on this issue. Incidents like the French ban on burqa/sikh symbols, Swiss ban on minarets, etc - would reduce the level of tolerance for these countries. Right now, they get away with saying that these are important for preserving their 'culture'.

2. Terror index. Based on number of terror attacks, people killed, draconian laws, distance from pakistan :), etc

3. Country stability index. Based on number of riots/protest against authority, legal avenues of protest, freedom of media, foreign debt, etc.

4. Investment safety/climate index (foreign and domestic can be same or separate). Measuring the prospects of safely investing in a country.

5. Social/human development index. This has to be different from the current western influenced UN index. For reference I am giving the UN Millenium Development Goals:
http://ddp-ext.worldbank.org/ext/ddprep ... EWADVANCED

and

http://unstats.un.org/unsd/mdg/Data.aspx
A_Gupta
BRF Oldie
Posts: 12056
Joined: 23 Oct 2001 11:31
Contact:

Re: Creating BR Country Index

Post by A_Gupta »

The index I'm interested in is something to capture the idea of a failing Pakistan in numerical terms. Presumably the same method, applied to other nations may tell us something.
abhischekcc
BRF Oldie
Posts: 4277
Joined: 12 Jul 1999 11:31
Location: If I can’t move the gods, I’ll stir up hell
Contact:

Re: Creating BR Country Index

Post by abhischekcc »

That would mean a country political risk index.

However, it would also require defining 'failure'. In the thread 'Managing Pakistan's failure', an interesting argument made was that Pakistan's descent into chaos need not be understood as simply failure (of the modern state), but as success of the islamist state in overthrowing the modernist state.

Other examples are: Somalia and post-Shah Iran. Both underwent traumatic revolution, and the modern state failed. And we ended up with islamist regimes there, which keep the peace with brutal violence.

Is break up (like Yugoslavia) a necessary component of defining failure, or is it optional?

PS
Perhaps we can assume that islamist takeover itself will define failure.
A_Gupta
BRF Oldie
Posts: 12056
Joined: 23 Oct 2001 11:31
Contact:

Re: Creating BR Country Index

Post by A_Gupta »

1. The CIA World Fact Books 2000-2010 are available for download.
2. I started gathering data out of them for various countries - India, Pakistan, Sri Lanka, Nepal, Bangaldesh, Afghanistan; for each year
a. Population
b. Percentage of population between 15 and 64 (supposedly working age population)
c. GDP (PPP)
I then compute the GDP per working population.
Because of the demographic bulge the working age population can increase faster than the population. In India, the working age population peaked at 64.3% in 2006, in Bangladesh 63.6(2006), Pakistan as of 2007 still rising at 58.8%, (55% in 2000, 58.8% in 2007).

From 2000-2007 the trends would show India, Sri Lanka, Bangladesh doing well by this measure, and Pakistan, Nepal and Afghanistan doing poorly (Afghanistan data is bad, though).

But now I've come to 2008, and there is a big anomaly in the CIA data.
e.g.,
India 2007 exchange rate GDP $805.5 billion, PPP - $4164 billion
India 2008 exchange rate GDP $1099 billion, PPP - $2966 billion
There is a big change in the exchange rate to PPP conversion; not so severe in the case of other countries. Any idea of why that might be?

My theory is that at a minimum, the GDP of a country ought to grow at least 1 to 1 with each person in the workforce. That is keeping productivity fixed. If productivity grows, the working population gets more and is happier. If the GDP doesn't keep pace, it means there is increasing unemployment/underemployment/productivity decline and the working population gets more and more unhappy.

I think the idea is borne out by the data below, just need a better, more consistent source.

PS: The plot for years 2000-2007
http://www.flickr.com/photos/macgupta/4 ... 4/sizes/l/
Image
abhischekcc
BRF Oldie
Posts: 4277
Joined: 12 Jul 1999 11:31
Location: If I can’t move the gods, I’ll stir up hell
Contact:

Re: Creating BR Country Index

Post by abhischekcc »

Some economic data, which has socio political implications, that we can consider are:

1. Foriegn aid received, both total and per capita. This is usually a good indicator of foreign interference in a country as well as the fragility and incapability of local admin institutions to take care of their own problems.

2. FDI, total and per capita. An indicator of the confidence of foreign capital/governments/people in the subject country. Preferably, FDI should be in non-extractive industries. This is because FDI in extractive industries is more out compulsion than out of choice - and is a poor indicator.

3. No of graduates. This is a broad indicator of how many middle level managers a country can produce, and *should* help measure its ability to tackle unemployment.

4. Various numbers for food security - food production, grain reserves, level of hunger, etc.
Unemployment and hunger together would give a good indication of chronic instability (as opposed to episodic instability). Data about major riots in India show that these usually take place when unemployment has stayed high in the affected area for a long period. (Hmmm, this applies to US as well). Moreover, data on ancient China shows that internecine wars took place after several years of famine.

5. Inequality - gini coefficient.

6. Tax base, taxation rate of economy?

---------------

Some violence indicators:

1. Number of guns per capita, gangs per capita (can be used for political activity - as experience in India has shown), murders per capita.

2. Varios factors which indicate resistance to government authority - attacks on police/army/government property, no of protest riots, no of PILs filed (but this also indicates faith in judiaciary).

3. Specifically terrorist activity: no of bomb blasts, severity (casualties per blast) of these bomb blasts, random or targetted blasts, number of pakistanis being issued visa (I say this seriously, not in sarcasm).

4. Is the country a host of terror outfits (Pakistan, UK, Canada), how many of these exist, what kind of activity is allowed or tolerated (head office, propaganda, fund raising, recruitment, training, open government support, etc).
A_Gupta
BRF Oldie
Posts: 12056
Joined: 23 Oct 2001 11:31
Contact:

Re: Creating BR Country Index

Post by A_Gupta »

Some things are difficult to measure:
http://pakteahouse.wordpress.com/2010/0 ... anisation/

RajeshA has already mentioned attacks on CD stores. How do we get to see self-censorship?
Last year, the petrified traders of Lahore’s Hall Road burnt objectionable CDs after receiving threats from extremists. A year later, low-intensity blasts took place in the crowded Hall Road — a market for electronics and kosher and non-kosher DVDs. This week, two internet cafes were targeted in densely populated areas of Lahore and some time back Peeru’s was also bombed. Reports have suggested that the cafes had received threats from unidentifiable numbers asking them to stop their businesses as they were turning into hubs of ‘immoral activities’. Just because no one died there, media attention has been patchy. A younger female colleague told me how tailors are hesitant to take orders for sleeveless shirts and other designs that may offend the purist dress code.
A_Gupta
BRF Oldie
Posts: 12056
Joined: 23 Oct 2001 11:31
Contact:

Re: Creating BR Country Index

Post by A_Gupta »

Which is right?
[1] Pakistan Federal Bureau of Statistics
Pakistan Demographic Survey Year XXXX
http://www.statpak.gov.pk/depts/fbs/sta ... stics.html
[2] CIA World Fact Book Year XXXX
https://www.cia.gov/library/publication ... index.html
Population of Pakistan
Year [1] [2]

2001 133.65 million 141.55 million
2003 138.98 million 150.69 million
2007 149.86 million 164.74 million

PS: This will grow:
http://spreadsheets.google.com/ccc?key= ... y3lEfCb7eQ
Locked