My research is about the relation of public spending on economic growth so I am using proxy data like economic growth rate and public spending as share of GDP, but initial regressions show some heteroskedacity problems so I am wondering if it would be necesary to seasonaly adjust the public spending and other variables as share of gdp before using them in the regression. thanks in advance
Since government activities such as spending and public revenue or debt generally do not show any seasonal fluctuations, a seasonal adjustment would make no sense. The statistical problems must therefore have other causes. If the time series are sufficiently long, these also fade into the background. Then valid empirical results can be achieved. See for example the following links:
https://www.researchgate.net/publication/273966820_Taxes_Tax_Systems_and_Economic_Growth
https://www.researchgate.net/publication/273973721_Impact_of_the_Tax_System-The_Case_of_the_Federal_Republic_of_Germany
If you detect seasonality, it should be removed for computing growth rates which are not inflated by seasonal patterns.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Regression