The declining importance of American cement production

Posted on Updated on

A simple rule of thumb: name a material fundamental to the global economy and China makes or consumes half of it. This dominance by China was made famous last year by Bill Gates, who tweeted the following about cement:

There are even alternative formulations of this that are equally staggering. Consider this: China increased its cement consumption more this century than the rest of the world’s did since the invention of cement. Don’t believe me? Check the data!

The dominance of China raises the question of historical precedent. For many things these are easily found in the Britain of the Industrial Revolution. China consumes half of the world’s coal today, something Britain easily achieved for decades.

For cement we can compare China with the country Bill Gates compared it with: America. Today China consumes roughly 60% of the world’s cement. Reliable historical statistics show that American cement production in the first half of the 20th century rivals China’s today in terms of its proportion of global production. (Though not in terms of absolute or per-capita figures, more of that in a later post perhaps.)

Here is the history of American cement production in relation to total global production:

us_history

In the mid-1920s, just prior to the Great Depression, America produced almost half of the world’s cement. But the Wall Street Crash of 1929 resulted in America going from producing almost half of the world’s cement to 20% of it in barely five years. The reason, of course, was simple: no one erects buildings during a depression.

There was then a brief recovery, with America rising to around 40% of global production in the mid 1940s. Thank FDR and the New Deal, and the decision of Hitler to level half of Europe.

But the story after this is clear. America has continued to become less and less relevant in terms of global cement production (and consumption, for that matter). Today, America makes up less than 4% of global cement production.

This relative decline was not matched by a decline in absolute cement production, which kept rising. Here is the history of absolute production:

us_history_total

Absolute production increased by a factor of four in the time it took America to go from 40 to 4% of global production.

As you can see, production peaked in the mid-2000s, before slumping in response to reality setting in after the full delusions of early 21st Century Wall Street had taken effect.

Whether American cement production will ever reach those highs again is not known.

Note on data

R code to produce the above plots, and the data source (USGS), is shown below.

# packages required. gdata for reading stupid excel files in. stringr for dealing with strings. ggplot2 for plotting
require(gdata);require(stringr);require(ggplot2)

options(stringsAsFactors = F)
## Download this file http://minerals.usgs.gov/minerals/pubs/historical-statistics/ds140-cemen.xlsx
## Read the data in
raw <- read.xls("ds140-cemen.xlsx", skip = 4)
raw <- raw[,c(1, 2,9)]
## Subset the data so that it only has the years needed
raw <- subset(raw, Year %in% 1900:2013)
## USGS have the wisdom to use commas in their excel file. Deal with it
raw[,1] <- as.integer(raw[,1])
raw[,2] <- as.integer(str_replace_all(raw[,2], ",", ""))
raw[,3] <- as.integer(str_replace_all(raw[,3], ",", ""))
## Restrict to years with global production data
raw <- subset(raw, !is.na(World.production))
## Do the plots
gg <- ggplot(raw, aes(Year, 100*Production/World.production))+
  geom_line()+
  ylim(c(0,50))+
  scale_x_continuous(breaks = seq(1920, 2010, by = 10))+
  ylab("Cement production (% of global total)")+
  labs(title = "American cement production")
gg

gg1 <- ggplot(raw, aes(Year, Production/10^6))+
  geom_line()+
  scale_x_continuous(breaks = seq(1920, 2010, by = 10))+
  ylab("Cement production per year (Mt)")+
  ylim(c(0, 100))
gg1



Advertisements

One thought on “The declining importance of American cement production

    peter2108 said:
    May 14, 2015 at 12:59 pm

    I applaud the posting of code to reproduce the figures … but what are you doing to embed the R-script – including the syntax highlighting?

    Like

Comments are closed.