I possess data from various countries, including a diverse mix such as the USA, Slovenia, and Chile, covering a wide range of GDP per capita levels. This data encompasses individual and household income levels, alongside information on whether respondents saved money in the past year. My primary objective is to analyze how savings behavior varies across different segments of the income distribution, particularly to ascertain if individuals with comparable purchasing power exhibit distinct savings patterns across these countries.

Initially, I considered segmenting the income data into four quartiles to compare savings behavior across these segments after adjusting for inflation. However, following a suggestion, I'm contemplating adjusting for purchasing power parity (PPP) and converting all income data into a common currency. The recommendation further proposed categorizing income into specific groups (e.g., Group 1: 0-20,000; Group 2: 20,001-40,000; Group 3: 40,001-70,000; Group 4: 70,001 and above) rather than quartiles. This approach, however, introduces challenges, such as the median income in Chile being $19,000, which would place half of the households in the lowest income group, leaving few in the higher income categories. Conversely, in the USA, the distribution would skew towards the higher income groups.

This discrepancy raises concerns about the feasibility of establishing a uniform income categorization that accurately reflects the economic realities across different countries. Given these challenges, I am inclined to continue using the quartile method but am open to suggestions. Does anyone have insights or alternative methods to address this issue more effectively? Your input would be greatly appreciated.



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.