Assignment 2

Click the Original, Code and Reconstruction tabs to read about the issues and how they were fixed.

A full list of references is given on the References tab.

Original

Source: Financial Times (2013).

Objective

As mentioned by the Global Times (2013), these source statistics were revisions on

Explain the objective of the original data visualisation and the targetted audience.

The visualisation chosen had the following three main issues:

Unethical data selection
- The y-axis limits for the first plot are
- Each subplot uses a different time period
  - GDP uses 2010 - 2012, despite data being available
  - GINI uses 2003-2012, as released the previous week
  - If like years were used as-is, it would imply a steep drop in inequality
No clear purpose
- The visualisation doesn’t clearly and easily connect to the bulk of the article
- Neither subplots clearly connect to each other, and their relevance isn’t explained
- There is no clear narrative linking the two pieces of information, and only the second subplot is discussed clearly in the article
Visualisation Style/Organisation
- GDP data
  - Subplot states that it is ‘Annual % change’, but there are four colums per year given
  - This may refer to quarter-on-quarter change, but NBS China’s ‘Quarterly Data’ (n. d.) shows a far lower rate
  - This may mean that these are a rolling twelve month rate of change, but this is not clearly indicated
  - Because I have included the identical years for each subplot, I have instead used the actual annual rate of change, and labelled it so.
- GINI data
  - Scale clarification
    - With the inclusion of the UN Warning Level, there was the opportunity to include an additional equality level to orient and inform the reader
    - This was taken from the UNICEF China (2018) report, indicating equality being 0.0 - 0.2, so the report was not available to the author at the time of writing, but similar information would’ve existed
    - This allowed me to remove the difficult-to-read and unnecessarily verbose note on the subplot

Target Audience

According to their readership surveys (FT 2022), The Financial Times is marketed towards an audience that has a high net worth, and is involved in high-level decision-making in their respective businesses and industries. While a high proportion of the readership is involved in the finance industry (at 31%), the audience does vary across a wide spectrum of the public. These factors combined, means that the average reader will have a higher level of domain expertise, and a higher proportion will have expertise specific to business strategy, finance, and economics.

The target audience of the article in question will follow a similar profile. The article focuses on the development of the Chinese economy broadly, with a specific focus on the economic concerns of the general public. It uses the rise in the gambling industry as a focal point to examine the living conditions of the Chinese low- and middle-income earners, in a period of overall high economic growth.

Code

The following code was used to fix the issues identified in the original.

# For individual plots
library(tidyverse)
library(readxl)
library(magrittr)
library(here)

# For final visualisation
library(patchwork)
library(ggtext)


# Import data
data_gini <- read.csv(here("Data", "GINI.csv"))
data_gdp <- read_excel(here("Data", "imf-dm-export-20221119.xls"), sheet = 1)


# Wrangle GDP data
# IMF data has blank lines and a footer row
# NOTE: might not work for countries with missing data!
data_gdp <- filter_all(data_gdp, all_vars(!is.na(.)))

# Pivot to the same format as GINI data
data_gdp <- pivot_longer(
    data_gdp,
    colnames(data_gdp)[-1],
    names_to = "Year",
    values_to = "Rate",
    names_transform = as.integer
  )

# Filter data to the same years as GINI data
data_gdp <- filter(data_gdp, Year %in% data_gini$Year)


# Theme colours taken from FT website
ft_colour_red <- "#99103d"
ft_colour_bg <- "#fff1e5"
ft_colour_bgemph <- "#f2dfce"
ft_colour_unemph <- "#fff7ef"


# Create GDP subplot
plot_gdp <- ggplot(data = data_gdp, aes(x = Year, y = Rate)) +
  
            geom_bar(stat = "identity", fill=ft_colour_red) +
  
            scale_y_continuous(name = "% YoY", breaks = (seq(0, 15, 2))) +
            scale_x_discrete(name = "Year", limits = data_gdp$Year[seq(2, nrow(data_gdp), 2)]) +
  
            theme(
              panel.background = element_rect(fill = NA),
              axis.line = element_line(color = 'grey'),
              
              panel.grid.major.x = element_blank(),
              panel.grid.major.y = element_line(size = 0.1, color = "grey"),
              panel.grid.minor.y = element_blank()
            )


# Create GINI subplot
plot_gini <- ggplot(data = data_gini, aes(x = Year, y = GINI)) +
  
             geom_area(fill = ft_colour_unemph) +
             geom_line(stat = "identity", colour = ft_colour_red) +
             geom_point(colour = ft_colour_red) +
  
             coord_cartesian(ylim=c(0, 0.5)) +
             scale_y_continuous(name = "GINI Coefficient", breaks = seq(0.0, 0.5, 0.075)) +
             scale_x_discrete(name = "Year", limits = data_gdp$Year[seq(2, nrow(data_gdp), 2)]) +
  
             theme(
                panel.background = element_rect(fill = NA),
                panel.ontop = TRUE,
                axis.line = element_line(color = 'grey'),
                
                panel.grid.major.x = element_blank(),
                panel.grid.major.y = element_line(size = 0.1, color = "grey"),
                panel.grid.minor.y = element_blank()
             ) +
  
             geom_hline(yintercept = 0.4, linetype = "dashed", color = ft_colour_red, size = 0.5) +
             annotate("text", x = mean(data_gini$Year), y = 0.42, label = "UN Warning Level", size = 3) +

             geom_hline(yintercept = 0.1, linetype = "dashed", color = ft_colour_red, size = 0.5) +
             annotate("text", x = mean(data_gini$Year), y = 0.12, label = "Income Equality", size = 3) 


# Create final visualisation
patch <- plot_gdp +
         plot_gini +
  
         plot_annotation(
            theme = theme(plot.title = element_markdown(lineheight = 1.1)),
            title = "<span style='color:#99103d;'>GDP Growth</span> versus <span style='color:#cf9969;'>Income Equality</span>",
            subtitle = "Growth slows, but inequality remains stubbornly high",
            caption = "Source: IMF (2022), UNICEF China (2018), Global Times (2013)"
         )

Reconstruction

The following plot fixes the main issues in the original.

patch

Choices

Unethical data selection
- Identical time periods chosen
  - The GDP data would have always been available at the time, as they are derived from the Statistical Yearbooks, which only ever excluded the GINI coefficient data
- Changed GINI subplot to use accurate proportions
  - Result gives a radically different impression of stability in inequality levels, but the addition information reorients the message back to support the article’s premise
  - This warning level was stated by the director of China’s National Bureau of Statistics (NBS) in the press conference for the release of these statistics (Global Times 2013)
  - This may not have been intentional, because the default ggplot2 settings will result in a plot similar to the final article, but it still seems like a major oversight, given that it was not the case with the GDP plot.
No clear purpose
- Colours intentionally reused, as these represent organisation scheme
- Title and subtitle added to illustrate relevance to the article
  - This was chosen to give it relevance to the general tone and support the overall purpose of the article
- Leverage existing colours, to highlight title to each subplot
  - Because two distinct colours were used in the original, this colour scheme was copied, then leveraged in the title, to allow the audience to easily distinguish the purpose of each plot without being overly verbose.
Visualisation Style/Organisation
- GDP data
  - Subplot states that it is ‘Annual % change’, but there are four colums per year given
  - This may refer to quarter-on-quarter change, but NBS China’s ‘Quarterly Data’ (n. d.) shows a far lower rate
  - This may mean that these are a rolling twelve month rate of change, but this is not clearly indicated
  - Because I have included the identical years for each subplot, I have instead used the actual annual rate of change, and labelled it so.
- GINI data
  - Scale clarification
    - With the inclusion of the UN Warning Level, there was the opportunity to include an additional equality level to orient and inform the reader
    - This was taken from the UNICEF China (2018) report, indicating equality being 0.0 - 0.2, so the report was not available to the author at the time of writing, but similar information would’ve existed
    - This allowed me to remove the difficult-to-read and unnecessarily verbose note on the subplot

References

FT (The Financial Times Ltd) (2022) Audience, Financial Times website, accessed 20 November 2022. https://commercial.ft.com/audience/

IMF (International Monetary Fund) (2022) Real GDP growth: Annual percent change, [Excel], IMF website, accessed 19 November 2022. https://www.imf.org/external/datamapper/NGDP_RPCH@WEO/

Mazzocco I (2022) ‘How Inequality Is Undermining China’s Prosperity’, Center for Strategic & International Studies, accessed 14 November 2022. https://www.csis.org/features/how-inequality-undermining-chinas-prosperity

NBS China (National Bureau of Statistics of China) (n. d.) Quarterly Data, NBS China website, accessed 19 November 2022. http://www.stats.gov.cn/english/statisticaldata/Quarterlydata/index.html

Rabinovitch S (23 January 2013) ‘The Chinese lottery: a tax on hope’, Financial Times, accessed 13 November 2022. https://www.ft.com/content/11ddb6d6-fe44-324b-8b9e-e4ecf5868e08

UNICEF China (2018) Children in China: An Atlas of Social Indicators 2018, UNICEF China, accessed 18 November 2022. https://www.unicef.cn/en/atlas-2018-en

Global Times (20 January 2013) ‘China’s inequality index highlights urgency for distribution reforms’, Global Times, accessed 20 November 2022. https://www.globaltimes.cn/content/756903.shtml