Derive an adequate (residual) component in the time series of unemployment rates

[Sofiya Volvakova]

Hello Stata Community,

I have been asked to derive an adequate (residual) component in the time series of unemployment rates from the attached dataset. However, I am struggling to understand how exactly this needs to be done - could you please help? This is what I have done so far:

reg UE Year
predict resid, residuals
egen Mean = mean(resid)
line resid Mean Year, xla(1955(5)1990) name(G1, replace)

Thank you so much in advance!

Sofiya

[Justin Niakamal]

You can start with an Unobserved-components model. I suggest you read the helpfile for details (see help ucm). Preferably you'd use a higher frequency series like monthly data as I'm not sure how appropriate this is with annual data. Here's a start:

use "https://www.statalist.org/forums/filedata/fetch?id=1699354", clear

tsset Year
ucm UE, cycle(1)

predict residual, residual 
predict cycle, cycle
predict trend, trend

// ssc install multiline
multiline UE trend residual cycle  Year, xlab(#4) xtitle("") scheme(s1color)

[Sofiya Volvakova]

You can start with an Unobserved-components model. I suggest you read the helpfile for details (see help ucm). Preferably you'd use a higher frequency series like monthly data as I'm not sure how appropriate this is with annual data. Here's a start:

use "https://www.statalist.org/forums/filedata/fetch?id=1699354", clear

tsset Year
ucm UE, cycle(1)

predict residual, residual
predict cycle, cycle
predict trend, trend

// ssc install multiline
multiline UE trend residual cycle Year, xlab(#4) xtitle("") scheme(s1color)

Thanks so much, Justin. That's been of great help!