An Income Model (R version)

An income model from socio-demographic variables

Published on Oct 02, 2022

Reading time: 3 minutes.

Github

Built with

R

Introduction

Back in the day I used to work at a financial institution where my main focus was to select potential new customers for credit cards joining many sources of information and applying business rules. One of this sources was an income model: when offering a credit card to a new customer we would like to do it at a minimal risk so getting to know the income becomes crucial, but sometimes this sensitive information is not available to us. Another source was socio-demographic information.

I never knew how this income model came to be, but I gotta say it was way off. Some other models were proposed but they never really took off mostly because they were not interpretable.

Here I propose a model based on some socio-demographic variables at a household level from INEGI using linear regression where we can interpret the resulting coefficients. Next we improve this model using expenditure information, which in a business scenario would be helpful to adjust the credit or to offer a new one.

Data

The data comes from the ENIGH: Encuesta Nacional de Ingresos y Gastos de los Hogares 2020 (National Survey of Household Income and Expenses 2020) taken from INEGI^[The National Institute of Statistics and Geography is an autonomous agency of the Mexican Government dedicated to coordinate the National System of Statistical and Geographical Information of Mexico]

The data will be taken from four .csv files:

1. conjunto_de_datos_concentradohogar_enigh_2020_ns.csv

Household variables:

  • folioviv: Household identifier

  • tam_loc: Population size in which the household is located

    • 1 - 100,000 or more population
    • 2 - 15,000 to 99,999 population
    • 3 - 2,500 to 14,999 population
    • 4 - 2,500 or less population

  • est_socio: Socioeconomic status

    • 1 - Low
    • 2 - Medium-low
    • 3 - Medium-high
    • 4 - High

  • clase_hog: Type of home

    • 1 - Unipersonal
    • 2 - Nuclear
    • 3 - Extended
    • 4 - Composite
    • 5 - Co-resident

  • sexo_jefe: Head of family gender

    • 1 - Male
    • 2 - Family

  • edad_jefe: Head of family age

  • tot_integ: Total number of members

  • estim_alqu: Rent estimate

2. conjunto_de_datos_hogares_enigh_2020_ns.csv

Household characteristics

  • folioviv: Household identifier

  • acc_alim2: No food

    • 1 - Yes
    • 2 - No

  • alim17_2: Days in which tubers were eaten

  • alim17_3: Days in which vegetables were eaten

  • alim17_4: Days in which fruits were eaten

  • alim17_5: Days in which meat was eaten

  • alim17_6: Days in which egg was eaten

  • alim17_7: Days in which fish was eaten

  • alim17_8: Days in which legumes were eaten

  • alim17_9: Days in which dairy products were eaten

  • alim17_11: Days in which sugar was eaten

  • telefono: It has landline

    • 1 - Yes
    • 2 - No

  • cellphone: It has cellphone

    • 1 - Yes
    • 2 - No

  • tv_paga: It has cable

    • 1 - Yes
    • 2 - No

  • conex_inte: It has internet

    • 1 - Yes
    • 2 - No

3. conjunto_de_datos_gastoshogar_enigh_2020_ns.csv

Household expenditure.

  • folioviv: Household identifier

  • lugar_comp: Type of establishment

    • 0 - Non-applicable
    • 1 - Market
    • 2 - Flea market
    • 3 - Street vendor
    • 4 - Grocery store
    • 5 - Specialty store
    • 6 - Supermarkets
    • 7 - Department store
    • 8 - Purchases outside the country
    • 9 - Wholesale club
    • 10 - Corner store
    • 11 - Restaurants
    • 12 - Other food places
    • 13 - Coffee shops
    • 14 - Bars
    • 15 - Diconsa
    • 16 - Liconsa store
    • 17 - Individual sale
    • 18 - Internet

  • gasto_tri: Quartely expenditure

4. conjunto_de_datos_ingresos_enigh_2020_ns.csv

Income of each member in the household.

  • folioviv: Household identifier

  • ing_tri: Quaterly income

R Markdown

This is an R Markdown document. Markdown is a simple formatting syntax for authoring HTML, PDF, and MS Word documents. For more details on using R Markdown see http://rmarkdown.rstudio.com.

When you click the Knit button a document will be generated that includes both content as well as the output of any embedded R code chunks within the document. You can embed an R code chunk like this:

1

summary(cars)

##      speed           dist       
##  Min.   : 4.0   Min.   :  2.00  
##  1st Qu.:12.0   1st Qu.: 26.00  
##  Median :15.0   Median : 36.00  
##  Mean   :15.4   Mean   : 42.98  
##  3rd Qu.:19.0   3rd Qu.: 56.00  
##  Max.   :25.0   Max.   :120.00

Including Plots

You can also embed plots, for example:

Note that the echo = FALSE parameter was added to the code chunk to prevent printing of the R code that generated the plot.