How to create a geom line plot with single geom point at the end with legend

For geom_point and geom_text, you can subset your dataframe by keeping only the last year (2018):

ggplot(df, aes(x = `Gss year for this respondent`, y = Percent*100))+
  geom_line(aes(color = `Political party affiliation`), show.legend = FALSE)+
  geom_point(data = df[df$`Gss year for this respondent` == max(df$`Gss year for this respondent`),],
             aes(x =`Gss year for this respondent`, y = Percent*100, color = `Political party affiliation`), show.legend = FALSE)+
  geom_text(data = df[df$`Gss year for this respondent` == max(df$`Gss year for this respondent`),],
            aes(x =`Gss year for this respondent`, 
                y = Percent*100, 
                colour = `Political party affiliation`,
                label = `Political party affiliation`), show.legend = FALSE, hjust = -0.1)+
  xlim(1975, 2026)+
  ylim(0,80)+
  scale_color_manual(values = c("blue","green","red"))

Here are a few updates to @dc37's answer that get OP closer to desired plot. To be clear, @dc37's answer solves the main problem: how to plot just the final dot in the series. All of this was too much to add as a comment to @dc37's answer, so I'm posting it as a separate solution - although it's really more an extension of @dc37's answer, rather than a distinct alternative.

Updates

• Fixes x-axis range so that annotations occur after axis stops

• Removes color from annotations

• Eliminates space between x-axis line and y=0 on plot

• Creates space around x-axis tick marks (space between axis line and tick text)

• Major y-axis grid lines displayed

• Made x- and y- ticks match the original

  ggplot(df, aes(x = `Gss year for this respondent`, y = Percent*100, 
             color = `Political party affiliation`)) +
  geom_line(show.legend = FALSE, size = 1) +
  geom_point(data = point_df, aes(x =`Gss year for this respondent`), 
             show.legend = FALSE, size = 2) +
  scale_x_continuous(breaks = seq(1980, 2015, by = 5)) +
  scale_y_continuous(limits = c(0, 80), 
                     breaks = seq(0, 80, by = 20),
                     expand = c(0, 0)) +
  scale_color_manual(values = c("deepskyblue1","gray80","firebrick3")) +
  geom_text(data = point_df,
            aes(x = 2019, y = Percent*100, 
                label = `Political party affiliation`), 
            color = "black",
            show.legend = FALSE, 
            hjust = -0.1) +
  coord_cartesian(xlim = c(1978, 2016), clip = 'off') +
  theme(plot.margin = unit(c(1,7,1,1), "lines"),
        panel.grid.major.y = element_line(color = "gray90"),
        axis.line.y = element_blank(),
        axis.ticks.length.x = unit(.3, units = "cm"),
        axis.ticks.x = element_line(linetype = 2)) +
  labs(y = "", x = "")  
  

  

Important
The graph produced from OP's data - replicated in OP's code, as well as this answer and @dc37's answer - looks slightly different than the original screenshot. (Look at how the lines don't overlap each other in the same way.) I haven't dug into this - the updates here are focused on cosmetics.