top of page

Steps of Econometric Model Formation

Econometric modelling bridges economic theories with real-world data to estimate relationships and predict outcomes. Below are the key steps involved in forming an econometric model, accompanied by numerical examples for better understanding.


Steps of Model Formation
Steps of Model Formation

Step 1: Economic or Financial Theory (Previous Studies)

Study the literature and make a conceptual note or start with an existing economic theory. For instance, literature suggests that consumption depends on income.



Step 2: Formulation of an Estimable Theoretical Model

Translate the theory, given in the literature or conceptual understanding into a mathematical form that can be estimated. So, we can write

The mathematical expression of the consumption function illustrates the relationship between consumption (C) and income (Y) with parameters alpha (α) and beta (β).



Step 3: Collection of Data

Collect sufficient and relevant data. Complete enumeration or sampling method can be used for data collection. However, the method of data collection depends on the study objectives, available resources and feasibility. (See, details of data collection method)


Table 1: Sample data

Year

Income (₹ '000)

Consumption (₹ '000)

2020

50

40

2021

60

48

2022

70

55

2023

80

65

2024

90

72


Step 4: Model Estimation

Estimate the model (formulated in Step 2), using the collected data and appropriate analytical tools and techniques. Let us estimate the model using the sample data (shown in Step 3). The estimated model can be expressed as follows:

The results show that when there is no income, consumers' expenditure is ₹10,000. As the income increases by 1 unit (here, it is ₹1,000), consumption rises by ₹700.



Step 5: Is the Model Statistically Adequate?

Here, comes the importance of post-diagnostic tests of the model. In our example, we can check the statistical significance using and the p-value of the beta coefficient.

Here, R² = 0.95 implies a high explanatory power of the model. The p-value for β = 0.002 (<0.05) indicates the independent variable i.e., income is statistically significant.


Step 6: Interpret Model

As our estimated model satisfies the required condition, we can proceed further with the interpretation of the results - ₹1,000 increase in income leads to a ₹700 rise in consumption or the consumption increases by 70% due to one unit increase in income.


If the estimated model satisfies the required condition, we need to reformulate the model. For example, we may add variables like interest rates or savings if the model lacks explanatory power.


Step 7: Interpret Model

Use the validated model for prediction, policy analysis, or forecasting. For example, the projected average income of Indian citizens for 2025 is ₹1,00,000 per month. What will be the estimated average consumption of Indians for 2025? We can identify the answer in the following way:




Final Thoughts

This structured approach ensures that the econometric model is theoretically sound, statistically reliable, and practically useful for economic decision-making.

Recent Posts

See All
Cumulative Frequency Distribution

Cumulative frequency is a statistical concept used to determine the total number of observations that fall below or at a given value in a...

 
 
 

Comments


  • Facebook
  • Twitter
  • LinkedIn

©2022 by Dr. Dona Ghosh

bottom of page