I got here throughout a chart of long run actual gross home product per capita for Australia (1900 to about 2020) that had a few errors in its y axis scale and labelling. I needed to see for myself what it ought to truly appear to be.
In search of information for this drew my consideration to the superb Maddison Undertaking, which publishes long run estimates of actual GDP per capita, in 2011 costs, for a lot of international locations. It’s definitely not excellent (value comparisons over such an extended time period—they usually go properly again earlier than 1900—is a vexed enterprise!), nevertheless it seems like a terrific effort. It’s essential to have one thing like this from the financial historians as a result of most traditional official statistics collection for Australia solely return to 1959, as a result of issues of comparisons over time.
The Maddison Undertaking’s newest launch is 2023.
Right here’s my re-creation of the unique chart (which isn’t proven right here), however with appropriate y axis labels:
A particular function of this chart is that it compares the precise development trajectory to a continuing price of development from 1900 to current. It’s a pleasant comparability. It takes care to attract—it’s undoubtedly not the identical as a line of greatest match. On a logarithmic scale the place the fixed development is a straight line, it offers an excellent level of reference for the sooner and slower intervals of development.
Right here’s my code that attracts the chart. The work is definitely performed by a project-specific perform draw_chart(), which attracts a reasonably respectable chart for any given nation out there within the database. This lets me simply produce comparable charts for different international locations (outcomes proven after the R code).
# obtain newest model of Maddison information from:
# https://www.rug.nl/ggdc/historicaldevelopment/maddison/releases/maddison-project-database-2023
library(tidyverse)
library(readxl)
library(scales)
library(glue)
library(countrycode)
gdppc <- read_excel("mpd2023_web.xlsx", sheet = "GDPpc", skip = 2)
#' Draw time collection chart from 1900 to 2023 for a single nation
#'
#' @param ccode 3 digit ISO nation code
#' @param factors whether or not so as to add factors for every remark(default is to only draw line)
draw_chart <- perform(ccode, factors = FALSE){
# nation identify for this nation code
cname <- countrycode(ccode, origin = "iso3c", vacation spot = "nation.identify.en")
# Knowledge for simply this nation"
data_tc <- gdppc |>
choose(12 months, all_of(ccode)) |>
filter(12 months >= 1900)
names(data_tc)[2] <- "worth"
constant_growth <- data_tc |>
organize(12 months) |>
drop_na() |>
summarise(n = max(12 months) - min(12 months),
begin = worth[1],
finish = worth[n()],
start_year = min(12 months),
# for drawing labels, not truly a 'mid' level:
mid_point = (finish + begin) / 4) |>
mutate(growth_rate = (finish / begin) ^ (1 / n) - 1)
# colors for fixed development and for information:
cgcol <- "crimson"
dcol <- "blue"
# outline plot
p1 <- data_tc |>
ggplot(aes(x = 12 months, y = worth)) +
# Draw fixed development line:
annotate("phase",
x = constant_growth$start_year,
xend = max(data_tc$12 months),
y = constant_growth$begin, yend = constant_growth$finish,
color = cgcol, linetype = 2) +
# draw information line:
geom_line(color = dcol) +
annotate("textual content", x = 1900, y = constant_growth$mid_point,
label = glue("Fixed development of {%(constant_growth$growth_rate, accuracy = 0.1)}"),
color = cgcol, hjust = 0) +
annotate("textual content", x = 2010, y = constant_growth$mid_point,
label = "Precise GDP per capita",
color = dcol, hjust = 1) +
scale_y_log10(label = dollar_format(accuracy = 1),
breaks = c(0, 0.25, 0.5, 1:6) * 10000) +
labs(x = "",
y= "",
title = glue("Long run historic development in GDP per individual in {cname}"),
subtitle = "GDP per capita, buying energy parity, 2011 costs.",
caption = "Supply: Maddison Undertaking Database 2023")
# for some international locations with damaged collection we would wish to draw factors, not simply
# traces:
if(factors){
p1 <- p1 + geom_point(color = dcol)
}
frs::svg_png(p1, glue("../img/0326-{cname}"), w = 9, h = 5)
}
draw_chart("AUS")
draw_chart("NZL")
draw_chart("USA")
draw_chart("DNK")
draw_chart("CHN", factors = TRUE)
draw_chart("IND")
draw_chart("GBR")
draw_chart("IDN")
draw_chart("JPN")Right here’s a few of these different outputs, together with some glib and not-really-thought via observations from myself. Some fascinating issues right here as we see the seen affect of historic occasions: Covid, the second world struggle and nationwide independence throughout a number of international locations; in addition to nationwide catastrophes comparable to China’s Nice Leap Ahead.
New Zealand and the USA have barely slower development than Australia over this era however an identical general trajectory. The USA has a very robust New Deal and World Struggle II increase to development:
Denmark has much less affect from the Thirties Nice Melancholy however skilled the struggle as financial destruction (moderately than a GDP increase as was the case for the USA). Total its skilled pretty quick and regular financial development:
The UK had a foul time recovering from World Struggle I however ultimately attained sooner (albeit periodically interrupted) financial development till the early 2000s:
China’s financial development actually started with its new financial insurance policies within the Seventies:
India’s vital financial development started quickly after 1947 independence however solely accelerated from the Eighties onwards after which farther from the 2000s onwards:
Indonesia’s post-war development path displays its turbulent political historical past however is basically a hit story. Along with the interruption of significant estimates through the struggle and deconolonisation, and the troubled transition from Sukarno to Suharto within the Sixties, the East Asia Financial Disaster of the late Nineteen Nineties actually stands out:
Japan had a very catastrophic World Struggle II, recovered properly till the late twentieth century, and in latest many years has sunk into well-known relative stagnation:
The Maddison Undertaking estimates are offered and mentioned in Bolt and Van Zanden (2024), “Maddison model estimates of the evolution of the world economic system: A brand new 2023 replace”, Journal of Financial Surveys, 1–41.
