Knowledge

*A banking centre is defined as an administrative unit – usually a town/panchayat-town or a village – which has at least 3 bank branches of any scheduled commercial banks of India.

• RBI publishes credit and deposit data for about 6000 banking centres, every quarter.
• Estimation of population for these banking centres basis Census 2011

It may be noted that except for these two information, there is no other standard, defined, socio-economic information for these units in India. Therefore, we rely on these two information types for the estimation of purchasing power and per-capita purchasing power for these banking centres. It is possible that these banking centres could have been village panchayats at the time of Census 2011, or specified as towns.

The number of households and population for the district/town projected based on 2001 and 2011 Census growth rates for rural/urban/town. This is the estimated number of households in the given income group in the selected district/town The number of households and population is obtained by projecting the total households (rural/urban/town) from 2011 based on decennial growth rate for 2001-2011. The growth rates are separately obtained for rural and urban households

The estimation of Purchasing Power (PP) or Disposable Income is based on observed data at the district level from the Reserve Bank of India.

The overarching hypothesis is the observed strong correlation between state per-capita GDP and state level aggregates of banking sector data adjusted for district's population. The State level purchasing power is then a function of the banking sector data. The state purchasing power is broken down for the district level based on the district level observed values of banking sector data. The rural and urban break-up is also available based on banking parameters.

The total number of households of a district/town is projected based on the decadal growth rates in Census, between 2001 and 2011; the rural and urban growth rates are separately obtained for each district, as also for the respective town.

The Purchasing Power estimates include

• Distribution of HHs by Income Groups
  The number of households in different income groups estimated based on the per-capita purchasing power, spread of economic activities, and the NSSO estimates on the expenses.
• Distribution of Purchasing Power by Income Groups:
  This is the combined purchasing power of the households in each income group for the district. In other words the total purchasing power of a district is split among the households in different income groups. This is a function of the number of households in each income group.

Purchasing power for centres is based on analysis of parameters such as credit and deposit of the respective center which is analysed as a function of the total banking business of the district. This relationship is used to estimate the purchasing power at the centre-level .

Purchasing power is estimated for each year. The user can select any set of time periods, and the system computes the compounded annual growth rate (CAGR) in terms of percent per annum.

Where,   EP = End period,   SP = Start period,   n = number of years

According to the Ministry of Statistics and Programme Implementation (MoSPI), private final consumption expenditure (PFCE) is defined as the expenditure incurred on final consumption of goods and services by the resident households and non-profit institutions serving households (NPISH)

Akara’s PP estimates at the district-level amount to more than 90% of the All-India PFCE figures released by the Ministry of Statistics and Programme Implementation (MOSPI) annually. Therefore, a strong relationship is established at this juncture which is also consistent across timeseries.

The overarching hypothesis on estimation of PFCE is the relationship between PP estimates of Akara and the PFCE released annually by MOSPI, through this the district/town and centre PFCE estimates are computed.

Commodity-wise estimation:

For commodity-wise PFCE estimates, state-level data on food and non-food items is used for estimating commodity-wise proportion to total PFCE. Mapping of NSS to PFCE heads is presented in the Annexure. Based on the mapping of commodities, the aggregate PFCE for each commodity.

Mapping of Food & Non-Food items from NSS Data to PFCE

4-Tier estimation from aggregate PFCE:

The aggregate PFCE by commodity is classified into four categories:

1. Basic
2. Intermediate
3. Premium
4. Luxury

The proportion of households in each category in a district (the income distribution) is used to estimate the sub-group-level estimate of PFCE for the district/town. In districts with larger number of households in higher income groups, the estimated PFCE in luxury items is higher. The population effect is taken in to consideration to account for the consumption of essential commodities.

The table below at the level of the relevant geography will appear as follows :

Rationale: In the above example, for the same category of consumption, say, “Clothing”, Ajmer has higher estimated expenditure, compared to Adilabad. That is an obvious conclusion. But the interesting comparison is between two similar towns with similar total purchasing power – that is Ajmer and Agra. However, the redistribution of total purchasing power under the Clothing category will be a function of number of estimated number households (rural and urban, separately projected). Further the redistribution of estimated expenses under “Clothing” into different categories – Basic, Intermediary, Luxury and Premium will depend upon the income distribution of the respective districts.

4-Tier commodity classification:

Sub-classification of PFCE by type of consumption into the following levels

1. Basic
2. Intermediate
3. Premium
4. Luxury

The subclassification into the 4 classes of consumption takes into account the income distribution of the district/town.

Based on price, the same commodity can be classified into different categories

4-Tier commodity classification – Use-cases

Case 1: Coffee, tea and cocoa

For instance, a coffee in a road-side stall at Rs. 15 will constitute consumption as a basic good, while the same consumed in an ambient 5-star hotel is a luxury good. While we cannot provide an estimate of commodity consumed – for instance the consumption of coffee, we shall provide an estimate of PFCE on “coffee, tea and cocoa” at the district (with rural/urban breakup) and towns with breakup of consumption in the 4 broad categories.

Case 2: Cereals and pulses

Similar to the case of coffee, cereals can be classified into 4 groups. Basic consists of coarse grains, intermediate consists of unbranded grains available in loose-packaging, premium refers to branded cereals (FMCG companies) and premium consists of organic cereals. It is to be noted that the price-point is the key differentiator.

From the latest update of District Metrics, we have the data of total purchasing power in different income groups in the table below:

As can be observed, the total purchasing power in Tiruppur and Vellore is similar. However, the income distribution is different which is shown in table below:

Both the districts have the largest distribution of purchasing power in the highest income group. Yet, the category in the income group Rs. 3,00,000 to Rs. 4,00,000 per year there is a difference. Further, the estimated number of households in the different income groups for the districts as shown in table below:

So, the estimate of expenses under different categories of goods as in the PFCE table is a function of the total number of households, the distribution of households in different income groups and their respective purchasing power. In the given example, “Clothing” expenditure in stores such as Manyavar or purchase of silk sarees of Rs. 12,000/- and more could be classified as “Luxury/Premium”, while expenditure for shirts/chudidhars at Rs. 300 per piece could constitute “Basic”. What our estimate provides is: Given the total households in the district, and the distribution of the households into different income groups, what is the total quantum the district/town in question spends in different category f goods/services? Of this expense, for each product, what is the estimated potential to be spent in different price ranges?

How is Index of Economic Activity (IEA) estimated ?

Having provided the estimates of the purchasing power, and the potential expenditure into different category of goods, the major macroeconomic component at the district level is the source of income. That is, what are the economic activities that drive the purchasing power in a district? Akara provides the “Index of Economic Activities” to answer this question.

The index gives the relative contribution of the district to all India, the total India being equal to 1000 for the select economic activity. The data can be analysed across the columns as well as rows. A district's IEA for a chosen activity indicates its share in the business compared to all-India total for that business. This is for a summing on columns. Individual IEA value across a row indicates the relative importance of that activity in the district's purchasing power. For instance, Namakkal district in Tamil Nadu will have a lower value in most economic activities but a high value in transport (across row); Namakkal will have a larger share in transport IEA compared to other districts for the Transport IEA (along a column)

IEA is available for the following economic activities:

The composite index of economic activity is based on the total business activity reflected in the banking sector. The composite index is useful in understanding the total business potential of the district.

1. The scorecard is an index number for the district, with index assigned for certain key parameters.
2. Our objective is to bring out the significant factors which have an impact on per-household purchasing power; as such, in our analysis per-household purchasing power is kept as the dependent variable.
3. Districts which have the complete data for all parameters are selected.
4. A multiple regression analysis is done, which reveals the strength and direction of relationship between the factors
5. Results of multiple regression indicated a strong relationship between per-household purchasing and the following parameters, with the least muliti-collinearity amongst them have been used as the independent variables
   1. Rural female literacy rate: Obtained from Census 2011
   2. Urbanisation rate: Obtained from Census 2011 (total urban population as a percentage of total district population)
   3. Spread of economic activities: Obtained from Index of Economic Activity (IEA). This indicates the number of activities where the district’s rank is within top 100 of the country’s 610 districts.
   4. Growth rate of Purchasing Power: Obtained from Purchasing Power (PP) estimates for the district from 2004 and each district’s growth rate to reach 2015 January value.
   5. Growth rate of Economic activity: Obtained from IEA estimates for the district from 2003 and the district’s growth rate to reach 2014 March value.
   6. Volatility of Purchasing Power: Coefficient of variation for annual growth rates of purchasing power during the decade 2004-2014
   7. Volatality of economic activity: Coefficient of variation for annual growth rate of IEA during the decade 2004 – 2014.
6. Principal component analysis (PCA) technique was used to assign weights to the parameters, including per-household purchasing power. For this, logarithmic values were obtained in order to standardize the data.
7. Weights obtained are then assigned to individual district level parameters to obtain the final index value.