UN Speeches Analysis

The United Nations shares, “It was not until 1968 that environmental issues received serious attention by any major UN organs. The Economic and Social Council on 29 May was the first to include those issues in its agenda as a specific item and decided – later endorsed by the General Assembly – to hold the first United Nations Conference on the Human Environment” (https://www.un.org/en/chronicle/article/stockholm-kyoto-brief-history-climate-change). In this analysis, I want to observe how the frequency and shape of climate change discussions have changed in the UN over time, specifically from 1970-2016. The data used in this study originates from the UN General Debate Corpus (UNGDC) (https://www.kaggle.com/datasets/unitednations/un-general-debates). I hope to analyze the presence frequency of climate change words in all UN speeches over time by countries that are both most impacted by climate change and are the largest contributors to climate change.

Hypothesis

I hypothesize, or rather I optimistically hope, that the nations that contribute to climate change the most discuss climate change more often than the countries that are most impacted by the effects of climate change.

First, I will need to upload the data and necessary packages.

Code

library(tidyverse)
library(tidytext)
library(ggplot2)
library(ggthemes)
library(wordcloud2)

Code

un_general_debates <- read_csv("un-general-debates.csv")

Now, I will create a lexicon of singular climate change words. I used the “Glossary of Climate Change Terms” from the United States Environmental Protection Agency (https://19january2017snapshot.epa.gov/climatechange/glossary-climate-change-terms_.html) to create this lexicon.

Code

un_general_debates |> 
  unnest_tokens(word, text) -> un_words

climate_words <- c('adaptation', 'aerosols', 'afforestation', 'albedo', 'anthropogenic', 'atmosphere', 'biofuels', 'biomass', 'borehole' , 'carbon', 'chlorofluorocarbons', 'climate', 'ecosystem', 'emissions', 'emission', 'footprint', 'fluorocarbons', 'fuel', 'geosphere', 'glacier', 'greenhouse', 'halocarbons', 'hydrocarbons', 'hydrochlorofluorocarbons', 'hydrofluorocarbons', 'hydrosphere', 'inundation', 'landfill', 'methane', 'mitigation', 'oxidize', 'ozone', 'offset', 'pollution', 'polluting', 'pollute', 'perfluorocarbons', 'permafrost', 'phenology', 'photosynthesis', 'precession', 'radiation','recycling', 'recycle', 'reflectivity', 'reforestation', 'resilience', 'respiration', 'vulnerability', 'warming', 'wastewater', 'weather')

Let’s look at how often these words were used in UN Speeches over time.

Code

un_words |> 
  filter(word %in% climate_words) |> 
  count(year, sort = TRUE) |> 
  ggplot(aes(year, n)) + geom_line() -> un_wordsyear

un_wordsyear + 
  ggtitle('Instances of Climate Words in UN Speeches Over Time') +
  ylab('Number of Instances of Climate Words') +
  xlab('Year')

From observing this graph, there are clear spikes of UN speeches that includes words within this lexicon at around 1999, 2007, 2009, and 2013.

Now, with a better understanding of the overall most popular moments for climate discussion overtime, let’s take a deeper dive into the most popular climate change words that were used in each decade.

Code

un_words |> 
  filter(year %in% c( '1970', '1971', '1972', '1973', '1974', '1975', '1976', '1977', '1978', '1979')) |>
  filter(word %in% climate_words) |> 
  count(word, sort = TRUE) |> 
  knitr::kable() 

word	n
atmosphere	689
climate	580
pollution	243
fuel	107
adaptation	43
vulnerability	32
offset	31
weather	26
resilience	16
recycling	14
polluting	13
radiation	12
hydrocarbons	11
reforestation	7
pollute	6
ecosystem	4
carbon	2
mitigation	2
ozone	2
recycle	2
warming	2
afforestation	1
biomass	1
emissions	1

By far, the most popular words from the climate change lexicon during the 1970s were atmosphere, climate, pollution, and fuel.