Need to reorder paired bar chart

How do i change the order of the paired bars?

I would like Alaska (light blue) to be above the AM West (dark blue) for every pair.

ggplot(data=Overview_All_flights_by_Airline_and_Destination1, 
       aes(x= Percent_Ontime, y=reorder(Destination, +Percent_Ontime), 
            fill=Airline)) +
  geom_bar(stat="identity", color="black",position=position_dodge()) +
  geom_text(aes(label=Percent_Ontime), vjust=-0.3, 
                position = position_dodge(0.9), size=3.5) +
  scale_fill_brewer(palette="Paired") +
  theme_minimal()  

What makes things difficult is that y- and x-axis are flipped.
One option is to define airline as a factor and setting the order of levels.
Using forcats library allows, you can reverse things back from what was changed by flipping axis. With forcats::fct_rev y- axis and fill needs to be reversed.
The order of the legend items can also be adopted with rev.

I have made a dummy DF, which comes close to your own DF

library(ggplot2)
library(forcats)
set.seed(123)
# Make artificial DF
destination <- rep(LETTERS[1:5], each = 2)
airline <- rep(c("Alaska", "AM West"), times = 5)
percent_ontime <- ceiling(runif(10, 70, 99))

df <- data.frame(destination, airline, percent_ontime)
df
#>    destination airline percent_ontime
#> 1            A  Alaska             79
#> 2            A AM West             93
#> 3            B  Alaska             82
#> 4            B AM West             96
#> 5            C  Alaska             98
#> 6            C AM West             72
#> 7            D  Alaska             86
#> 8            D AM West             96
#> 9            E  Alaska             86
#> 10           E AM West             84
# make airline a factor and define order of levels
# df$airline <- factor(df$airline, levels = c("AM West", "Alaska"))
df$airline <- factor(df$airline, levels = c("Alaska", "AM West"))

ggplot(
  data = df,
  aes(
    x = percent_ontime,
    y = fct_rev(destination),
    fill = fct_rev(airline)
  )
) +
  geom_bar(stat = "identity", color = "black", position = position_dodge()) +
  geom_text(aes(label = percent_ontime),
    vjust = -0.3,
    position = position_dodge(0.9), size = 3.5
  ) +
  scale_fill_brewer(palette = "Paired", breaks = rev) +
  theme_minimal()

<!-- -->