Telling a story about gas prices

Define what it means for gas to be expensive, then finish and visualize a story about how gas prices have changed over time.

Note

This application exercise is completed in class and submitted via a worksheet.

ImportantGetting started

This application exercise is designed to be run in your web browser using the {webr} framework. Simply work through the exercises and use the provided code cells to execute live R code in your browser.

library(tidyverse)
library(scales)
library(colorspace)

theme_set(
  brand.yml::theme_brand_ggplot2() +
    theme(plot.caption = element_text(hjust = 1))
)

Rising concerns about gas prices

Since 2021, Fox News polls have asked registered voters, “How much of a problem are current gas prices for you and your family?”

# Fox News poll results among registered voters
# "(Don't know)" responses are omitted (1% or less in every poll)
fox_poll <- tribble(
  ~date        , ~major , ~minor , ~not_problem , ~wording             ,
  "2021-10-19" ,     50 ,     34 ,           15 , "Rising gas prices"  ,
  "2022-03-21" ,     52 ,     36 ,           12 , "Rising gas prices"  ,
  "2022-06-13" ,     67 ,     23 ,            9 , "Current gas prices" ,
  "2023-08-14" ,     49 ,     36 ,           14 , "Current gas prices" ,
  "2024-05-13" ,     49 ,     35 ,           16 , "Current gas prices" ,
  "2024-09-16" ,     48 ,     36 ,           15 , "Current gas prices" ,
  "2025-09-09" ,     33 ,     43 ,           24 , "Current gas prices" ,
  "2026-04-20" ,     60 ,     29 ,           11 , "Current gas prices" ,
  "2026-09-14" ,     61 ,     29 ,           10 , "Current gas prices"
) |>
  mutate(date = ymd(date))

fox_poll |>
  pivot_longer(
    cols = c(major, minor, not_problem),
    names_to = "response",
    values_to = "pct"
  ) |>
  mutate(
    response = factor(
      response,
      levels = c("major", "minor", "not_problem"),
      labels = c("Major problem", "Minor problem", "Not a problem")
    )
  ) |>
  ggplot(mapping = aes(x = date, y = pct / 100, color = response)) +
  geom_line() +
  geom_point() +
  scale_y_continuous(labels = label_percent(), limits = c(0, NA)) +
  scale_color_viridis_d(end = 0.8) +
  labs(
    title = "Gas prices as a problem for families, 2021-2026",
    subtitle = "\"How much of a problem are current gas prices for you and your family?\"",
    x = NULL,
    y = "Percent of registered voters",
    color = NULL,
    caption = "Source: Fox News polls.\nOctober 2021 and March 2022 polls asked about \"rising gas prices.\""
  )

In recent months Americans have expressed increasing concerns about gas prices at the pump. Gas prices are one of the most visible prices in the economy. They are posted in large numbers on nearly every street corner, and most households pay them every week.

Our story starts with a simple question: how have gas prices changed over time?

Your turn: Before looking at any data, discuss with your group:

• How do we turn this question into a meaningful communication?
• Is the price per gallon over time sufficient to answer it?
• What else should we account for when deciding whether gas is “expensive”?

For each way of measuring gas prices your group comes up with, identify at least one strength and one weakness. Record your discussion on the worksheet.

Framing the story

Recall the four components of a story: Opening, Challenge, Action, and Resolution. We will start with the first two.

• Opening: Americans are concerned about current prices at the pump.
• Challenge: How have gas prices changed over time? Is this anxiety warranted?

Your group will finish the story by identifying the Action and Resolution.

Data

The dataset contains quarterly measures of gas prices and wages in the United States from 1979 through the second quarter of 2026. All series were obtained from FRED, the Federal Reserve Bank of St. Louis’s economic data portal.

• Gas prices are the average retail price of a gallon of regular unleaded gasoline across U.S. cities (APU000074714).
• Wages are the median usual weekly earnings of full-time wage and salary workers age 16 and older (LEU0252881500Q).
• Inflation-adjusted (“real”) values are converted to second quarter 2026 dollars using the Consumer Price Index for All Urban Consumers (CPIAUCSL).

Wage data is not available for the fourth quarter of 2025.

gas_wage <- read_csv("data/gas-wages.csv")
gas_wage

# A tibble: 190 × 14
   date       gas_price_nominal median_wage_nominal gas_price_real
   <date>                 <dbl>               <dbl>          <dbl>
 1 1979-01-01             0.734                 234           3.53
 2 1979-04-01             0.849                 239           3.96
 3 1979-07-01             0.986                 240           4.45
 4 1979-10-01             1.04                  249           4.58
 5 1980-01-01             1.20                  256           5.04
 6 1980-04-01             1.27                  257           5.16
 7 1980-07-01             1.26                  262           5.06
 8 1980-10-01             1.25                  271           4.88
 9 1981-01-01             1.37                  278           5.17
10 1981-04-01             1.40                  279           5.20
# ℹ 180 more rows
# ℹ 10 more variables: median_wage_real <dbl>,
#   median_wage_per_hour_nominal <dbl>, median_wage_per_hour_real <dbl>,
#   gallons_per_hour <dbl>, minutes_per_gallon <dbl>, below_mean <lgl>,
#   gas_price_nominal_index <dbl>, gas_price_real_index <dbl>,
#   median_wage_nominal_index <dbl>, median_wage_real_index <dbl>

Variable	Description
date	First day of the quarter
gas_price_nominal	Average price per gallon of regular gasoline (dollars)
median_wage_nominal	Median weekly earnings (dollars)
gas_price_real	Average price per gallon of regular gasoline (2026 Q2 dollars)
median_wage_real	Median weekly earnings (2026 Q2 dollars)
median_wage_per_hour_nominal	Median hourly earnings, assuming a 40-hour work week (dollars)
median_wage_per_hour_real	Median hourly earnings, assuming a 40-hour work week (2026 Q2 dollars)
gallons_per_hour	Gallons of gas that one hour of work at the median wage can purchase
minutes_per_gallon	Minutes of work at the median wage needed to purchase one gallon of gas
below_mean	Is gallons_per_hour below its average over the full time period?
gas_price_nominal_index	Nominal gas price, indexed so 1979 Q1 = 100
gas_price_real_index	Inflation-adjusted gas price, indexed so 1979 Q1 = 100
median_wage_nominal_index	Nominal median weekly earnings, indexed so 1979 Q1 = 100
median_wage_real_index	Inflation-adjusted median weekly earnings, indexed so 1979 Q1 = 100

ImportantInteractive table of the data

Sample charts

The charts below are a starting point for your analysis. They are intentionally plain. Use them to decide which measures best answer the challenge and what your story should be, not as models for your final design.

Price of gas

ggplot(data = gas_wage, mapping = aes(x = date, y = gas_price_nominal)) +
  geom_line() +
  scale_y_continuous(labels = label_currency()) +
  labs(
    title = "Price of regular gasoline, 1979-2026",
    x = NULL,
    y = "Price per gallon",
    caption = "Source: U.S. Bureau of Labor Statistics via FRED"
  )

ggplot(data = gas_wage, mapping = aes(x = date, y = gas_price_real)) +
  geom_line() +
  scale_y_continuous(labels = label_currency()) +
  labs(
    title = "Inflation-adjusted price of regular gasoline, 1979-2026",
    x = NULL,
    y = "Price per gallon (2026 dollars)",
    caption = "Source: U.S. Bureau of Labor Statistics via FRED"
  )

Wages

ggplot(data = gas_wage, mapping = aes(x = date, y = median_wage_nominal)) +
  geom_line() +
  scale_y_continuous(labels = label_currency()) +
  labs(
    title = "Median weekly earnings, 1979-2026",
    x = NULL,
    y = "Median weekly earnings",
    caption = "Source: U.S. Bureau of Labor Statistics via FRED"
  )

ggplot(data = gas_wage, mapping = aes(x = date, y = median_wage_real)) +
  geom_line() +
  scale_y_continuous(labels = label_currency()) +
  labs(
    title = "Inflation-adjusted median weekly earnings, 1979-2026",
    x = NULL,
    y = "Median weekly earnings (2026 dollars)",
    caption = "Source: U.S. Bureau of Labor Statistics via FRED"
  )

Gas prices and wages together

ggplot(data = gas_wage, mapping = aes(x = date)) +
  geom_line(mapping = aes(y = gas_price_nominal_index, color = "Gas price")) +
  geom_line(
    mapping = aes(y = median_wage_nominal_index, color = "Median wage")
  ) +
  scale_color_manual(values = c("grey30", "orange")) +
  labs(
    title = "Gas prices and wages, 1979-2026",
    x = NULL,
    y = "Index (1979 Q1 = 100)",
    color = NULL,
    caption = "Source: U.S. Bureau of Labor Statistics via FRED"
  )

ggplot(data = gas_wage, mapping = aes(x = date)) +
  geom_line(mapping = aes(y = gas_price_real_index, color = "Gas price")) +
  geom_line(mapping = aes(y = median_wage_real_index, color = "Median wage")) +
  scale_color_manual(values = c("grey30", "orange")) +
  labs(
    title = "Inflation-adjusted gas prices and wages, 1979-2026",
    x = NULL,
    y = "Index (1979 Q1 = 100)",
    color = NULL,
    caption = "Source: U.S. Bureau of Labor Statistics via FRED"
  )

Gas prices relative to wages

ggplot(data = gas_wage, mapping = aes(x = date, y = gallons_per_hour)) +
  geom_line() +
  labs(
    title = "Gallons of gas per hour of work, 1979-2026",
    x = NULL,
    y = "Gallons per hour of work",
    caption = "Source: U.S. Bureau of Labor Statistics via FRED"
  )

ggplot(data = gas_wage, mapping = aes(x = date, y = minutes_per_gallon)) +
  geom_line() +
  labs(
    title = "Minutes of work per gallon of gas, 1979-2026",
    x = NULL,
    y = "Minutes of work per gallon",
    caption = "Source: U.S. Bureau of Labor Statistics via FRED"
  )

# interpolate a series onto a fine grid so the shading changes color exactly
# where the line crosses the average for the full time period
interpolate_vs_mean <- function(data, var) {
  data <- drop_na(data, {{ var }})
  values <- pull(data, {{ var }})

  approx(x = as.numeric(data$date), y = values, n = 2000) |>
    as_tibble() |>
    transmute(
      date = as.Date(x),
      value = y,
      avg = mean(values),
      position = if_else(value < avg, "Below average", "Above average"),
      # separate ribbon per contiguous run so groups don't bridge gaps
      run = consecutive_id(position)
    )
}

gph_fine <- interpolate_vs_mean(gas_wage, gallons_per_hour)

gph_plot <- ggplot(data = gph_fine, mapping = aes(x = date)) +
  geom_ribbon(
    mapping = aes(
      ymin = pmin(value, avg),
      ymax = pmax(value, avg),
      fill = position,
      group = run
    )
  ) +
  geom_line(data = gas_wage, mapping = aes(y = gallons_per_hour)) +
  geom_hline(
    yintercept = gph_fine$avg[[1]],
    color = "grey50",
    linetype = "dashed"
  ) +
  labs(
    title = "Gallons of gas per hour of work, 1979-2026",
    subtitle = "Compared to the average for the full period",
    x = NULL,
    y = "Gallons per hour of work",
    fill = NULL,
    caption = "Source: U.S. Bureau of Labor Statistics via FRED"
  )

gph_plot +
  scale_fill_manual(values = c("grey70", "orange"))

gph_plot +
  scale_fill_manual(values = c("orange", "grey70"))

mpg_fine <- interpolate_vs_mean(gas_wage, minutes_per_gallon)

mpg_plot <- ggplot(data = mpg_fine, mapping = aes(x = date)) +
  geom_ribbon(
    mapping = aes(
      ymin = pmin(value, avg),
      ymax = pmax(value, avg),
      fill = position,
      group = run
    )
  ) +
  geom_line(data = gas_wage, mapping = aes(y = minutes_per_gallon)) +
  geom_hline(
    yintercept = mpg_fine$avg[[1]],
    color = "grey50",
    linetype = "dashed"
  ) +
  labs(
    title = "Minutes of work per gallon of gas, 1979-2026",
    subtitle = "Compared to the average for the full period",
    x = NULL,
    y = "Minutes of work per gallon",
    fill = NULL,
    caption = "Source: U.S. Bureau of Labor Statistics via FRED"
  )

mpg_plot +
  scale_fill_manual(values = c("grey70", "orange"))

mpg_plot +
  scale_fill_manual(values = c("orange", "grey70"))

Finish the story

Your turn: With your group, use the sample charts to finish the story.

• Action: What does the data show? How does it answer the challenge?
• Resolution: What should the audience take away? What, if anything, should they do or think differently?

Record the action and resolution on the worksheet.

Design two sequential charts

Your turn: Sketch two charts that, shown in order, carry your audience from the challenge to the resolution. The first chart should set up the problem, and the second should deliver the action or resolution. For each chart, write the title you would use. Titles should communicate the takeaway of the chart, not just describe its contents.

Once you have a sketch, implement your charts using the code cells below. The data is loaded as gas_wage, and {tidyverse}, {scales}, {ggrepel}, and {ggtext} are available.

Chart 1

Chart 2

Acknowledgments

• Exercise inspired by Gasoline price in hours of work from Eco3min.