I am using the lag function (package plm) which is returning me the contemporaneous value, not the lagged one.
In this code regarding a panel of data (years, agencies), the chg variables is filed with zeros.
group_by(Cod) %>%
mutate(orcam = orc*100/real,
chg = (orcam-lag(orcam))/lag(orcam)) %>%
ungroup()
I tried to create a function growth rate, this way
growth_rate <- function(x)(x/lag(x)-1)*100
chg = growth_rate(orcam)
The problem persisted.