posc1020-lecture-development-2.Rmd

---
title: "Development Policies and Development Politics"
subtitle: POSC 1020  -- Introduction to International Relations
author: Steven V. Miller
institute: Department of Political Science
titlegraphic: /Dropbox/teaching/clemson-academic.png
date: 
fontsize: 10pt
output:
 beamer_presentation:
    template: ~/Dropbox/miscelanea/svm-r-markdown-templates/svm-latex-beamer.tex
    latex_engine: xelatex
    dev: cairo_pdf
    fig_caption: false
    slide_level: 3
make149: true
mainfont: "Open Sans"
titlefont: "Titillium Web"
---

```{r setup, include=FALSE, cache=F, message=F, warning=F, results="hide"}
knitr::opts_chunk$set(cache=TRUE)
knitr::opts_chunk$set(fig.path='figs/')
knitr::opts_chunk$set(cache.path='cache/')

knitr::opts_chunk$set(
                  fig.process = function(x) {
                      x2 = sub('-\\d+([.][a-z]+)$', '\\1', x)
                      if (file.rename(x, x2)) x2 else x
                      }
                  )
```

```{r loadstuff, include=FALSE}
knitr::opts_chunk$set(cache=FALSE)

library(rvest)
library(WDI)
library(tidyverse)
library(stevemisc)
# library(maddison)

library(fredr)

library(pwt9)
data("pwt9.0")

pwt <- pwt9.0

Mexdebt <- readxl::read_xls("~/Dropbox/data/reinhart-rogoff/18_data.xls", sheet=20, skip=19) %>% slice(-1)
Argdebt <- readxl::read_xls("~/Dropbox/data/reinhart-rogoff/7_data.xls", sheet=3, skip=20) %>% slice(-1)
Bradebt <- readxl::read_xls("~/Dropbox/data/reinhart-rogoff/7_data.xls", sheet=8, skip=19) %>% slice(-1)

Mexdebt %>%
  rename(year = 1, debt = `debt/GDP`) %>%
  select(year, debt) %>%
  mutate(Country = "Mexico") -> Mexdebt

Argdebt %>%
  rename(year = 1, debt = `debt/GDP`) %>%
  select(year, debt) %>%
  mutate(Country = "Argentina") -> Argdebt

Bradebt %>%
  rename(year = 1, debt = `debt/GDP`) %>%
  select(year, debt) %>%
  mutate(Country = "Brazil") -> Bradebt


Trade <- read_csv("~/Dropbox/data/cow/trade/National_COW_4.0.csv")

# Price <- fred(series_id = "CPIAUCNS",
#                       observation_start = "1913-01-01") %>% data.frame %>%
#   data.frame(Date = rownames(.), .) %>% 
#   mutate(Date = as.Date(Date),
#          year = lubridate::year(Date)) %>% 
#   group_by(year) %>%
#   summarize(cpi = mean(CPIAUCNS)) %>%
#   mutate(last = last(cpi),
#        index = (cpi/last)*100) %>% tbl_df()


```


# Introduction
### Puzzle(s) for Today

*What explains development success and development failure?*

###

```{r asian-tigers-vs-latin-america, echo=F, eval=T, fig.width = 14, fig.height = 8.5, warning = F, message = F}

pwt %>%
  filter(country == "Argentina" | country == "Brazil" | country == "Mexico" | 
           country == "Singapore" |
           country == "Japan" | country == "Taiwan" | country == "Republic of Korea") %>%
  mutate(Type = ifelse(country == "Argentina" | country == "Brazil" | country == "Mexico",
                       "Import Substitution Industrialization", "Export-Oriented Industrialization")) %>%
  rename(Country = country) %>%
  mutate(rgdppc = rgdpna/pop) %>%
  ggplot(.,aes(year,rgdppc,group=Country,color=Country,linetype=Type)) + theme_steve_web() +
  geom_line(size=1.1) + 
   scale_color_brewer(palette="Paired") +
  scale_y_continuous(label=scales::comma) +
  xlab("Year") + ylab("Real GDP per Capita at Constant 2005 national prices (in mil. 2005US$") +
  scale_x_continuous(breaks=seq(1950,2015, by=5)) +
  labs(title="Why Did Development in These 'Asian Tigers' Far Exceed the Three Biggest Economies in Latin America?",
       subtitle = "One prominent answer focuses on the different development strategies we saw in East Asia vs. Latin America.",
       caption = "Data: Penn World Table (9.0)") +
  guides(linetype=F)

```


# What Can LDCs Do About This Bias?

1. Import-substituting industrialization (ISI)
2. Export-oriented industrialization (EOI)
3. Commodity cartels

### Autos Were Usually Focal Points of ISI

![Some countries, like Brazil, chose to cut off imports and protect/foster nascent industries](fnm-fabrica.jpg)

### The ISI Framework

- Enact trade barriers to shield industries from competition.
- Subsidize modern industrial sector.
- Invest in infrastructure to make industrialization possible.

```{r arg-bra-mex-trade-value, echo=F, eval=F, fig.width = 14, fig.height = 8.5, warning = F, message = F}
Trade %>%
  select(statename, imports, exports, year) %>%
  filter(statename == "Argentina" | statename == "Brazil" | statename == "Mexico") %>%
  left_join(Price, .) %>%
  mutate(importsa = (imports/index)*100,
         exportsa = (exports/index)*100) %>%
  filter(year <= 1990) %>% mutate(Country = forcats::fct_relevel(statename, "Mexico", "Brazil", "Argentina")) %>%
  ggplot(.,aes(year, importsa, color=Country, linetype=Country)) + theme_steve_web() +
  geom_line(size=1.1) +
  scale_x_continuous(breaks = seq(1910, 1990, by = 5)) 
```

###

```{r, echo=F, eval=F, fig.width = 14, fig.height = 8.5, warning = F, message = F}

pwt %>% tbl_df() %>%
  mutate(ccode = countrycode::countrycode(isocode, "iso3c", "cown")) %>%
  filter((ccode > 20 & ccode < 400)) %>%
  mutate(Category = ifelse(ccode >= 200, "Europe", "Latin America"),
         rgdppc = rgdpna/pop) %>%
  select(ccode,Category, year, rgdppc) %>%
  group_by(ccode) %>%
  mutate(growth = (rgdppc - lag(rgdppc, 1))/ lag(rgdppc, 1)*100) %>%
  ungroup() %>%
  group_by(Category, year) %>%
  summarize(growth = mean(growth, na.rm=T)) %>%
   ggplot(.,aes(year, growth, group = Category, color=Category, linetype = Category)) + theme_steve_web() +
  geom_line(size=1.1) + 
  scale_color_brewer(palette="Set1") +
  labs(caption = "Data: World Bank",
       title = "GDP per Capita for Zambia (Rhodesia) and South Korea, 1960-1975",
       subtitle = "Zambia (Rhodesia) was much better endowed and had more growth potential in the 1960s.") +
  xlab("Year") + ylab("GDP per Capita")

```

```{r lat-america-vs-europe-growth-1960-1990, echo=F, eval=T, fig.width = 14, fig.height = 8.5, warning = F, message = F}

WDI(country = c("EU", "ZJ"), indicator = c("NY.GDP.MKTP.KD.ZG"),
    start = 1960, end = 2016, extra = FALSE, cache = NULL) %>%
  rename(gdpgrowth = `NY.GDP.MKTP.KD.ZG`)  %>%
  mutate(Country = country) %>%
  filter(year <= 1990) %>%
  ggplot(.,aes(year, gdpgrowth, group = Country, color=Country, linetype = Country)) + theme_steve_web() +
  geom_line(size=1.1) + 
  scale_color_brewer(palette="Set1") +
  labs(caption = "Data: World Bank",
       title = "ISI's Failure Wasn't Total as Latin America Experienced Better GDP Growth Until the 1980s",
       subtitle = "We can qualify that Latin America had further to go but growth rates were real in the region, especially in the 1970s (even per capita).") +
  scale_x_continuous(breaks = seq(1960, 1990, by = 5)) +
  xlab("Year") + ylab("GDP Growth Rates")
```

###

```{r arg-bra-mex-debt-1945-1990, echo=F, eval=T, fig.width = 14, fig.height = 8.5, warning = F, message = F}
bind_rows(Mexdebt, Argdebt, Bradebt) %>%
  filter(year >= 1945 & year <= 1990) %>%
  mutate(Country = forcats::fct_relevel(Country, "Mexico", "Brazil", "Argentina")) %>%
  ggplot(.,aes(year,debt, color=Country, linetype=Country)) + theme_steve_web() +
  geom_line(size=1.1) +
  scale_x_continuous(breaks = seq(1945, 1990, by = 5)) +
  labs(title = "ISI Usually Meant These Governments Took on Debt Well Beyond Their Ability to Service It",
       subtitle = "ISI amounts to massive subsidies or even outright ownership of enterprises, which is compounded by the fact these exports weren't competitive.",
       caption = "Data: Various places, compiled by Reinhart and Rogoff (2009). I promise these statistics were presented without Excel errors, though. :P") +
  xlab("") + ylab("Central Government Debt/GDP")
```

### The ISI Framework

Problems:

- Industries were inefficient.
- Lack of competition usually meant products were low quality.
- Uncompetitive exports compounded economic downturns.

**Washington Consensus** to address ISI problems largely ended those policies.

###

```{r arg-tariff-rate-1980-2016, echo=F, eval=T, fig.width = 14, fig.height = 8.5, warning = F, message = F}
tribble(
  ~Year, ~Country, ~tariff,
  1980, "Argentina", 44,
  1981, "Argentina", 44,
  1982, "Argentina", 33,
  1983, "Argentina", 33,
  1984, "Argentina", 33,
  1985, "Argentina", 29.97,
  1986, "Argentina", 42.28,
  1987, "Argentina", 42.28,
  1988, "Argentina", 46.77,
  1989, "Argentina", 46.77,
  1990, "Argentina", 22.48,
  1991, "Argentina", 13.06,
  1992, "Argentina", 12.63,
  1993, "Argentina", 11.19,
  1994, "Argentina", 11.65,
  1995, "Argentina", 12.10,
  1996, "Argentina", 12.72,
  1997, "Argentina", 12.83,
  1998, "Argentina", 15.26,
  1999, "Argentina", 15.09,
  2000, "Argentina", 15.05,
  2001, "Argentina", 12.71,
  2002, "Argentina", 13.8,
  2003, "Argentina", 14.15,
  2004, "Argentina", 12.35,
  2005, "Argentina", 11.21,
  2006, "Argentina", 11.19,
  2007, "Argentina", 11.21, 
  2008, "Argentina", 9.98,
  2009, "Argentina", 10.16,
  2010, "Argentina", 11.31,
  2011, "Argentina", 11.29,
  2012, "Argentina", 11.14,
  2013, "Argentina", 12.17,
  2014, "Argentina", 12.58,
  2015, "Argentina", 12.48,
  2016, "Argentina", 12.58
) %>% 
  mutate(tariffp = tariff/100,
         lab = paste0(round(tariff, 1),"%")) %>%
  ggplot(.,aes(Year, tariffp)) + theme_steve_web() + 
  scale_y_continuous(labels=scales::percent) + 
  geom_text(aes(label=lab), vjust=-.5, colour="black",
            position=position_dodge(.9), size=3, family = "Open Sans") +
  scale_x_continuous(breaks = seq(1980, 2015, by = 5)) +
  geom_bar(stat="identity", alpha=0.8, fill="#619cff", color="black") +
  xlab("Year") + ylab("Tariff Rate (Simple Mean)") +
  labs(title = "Tariff Rates, Like Argentina's Tariff Rate Here, Were Some of the First Targets of Washington Consensus Reforms",
       subtitle = "Tariffs were a core component to ISI but reducing them were among the first orders of business for economic recovery.",
       caption = "Data: various sources, primarily World Bank (1980-1984, 2010-2016). See also: Lora's (2012) report for the Inter-American Development Bank.\nNote: 1980-1984 figures are actually means for 1980-1 and 1982-4 via Laird and Nogues' (1989) article in the World Bank Economic Review.")
  
```



### ISI Products Usually Became Punch-Lines

![ISI products like the Yugo usually became punch-lines](yugo-00.jpg)

###

```{r excel-tercel-yugo-sales-data-1985-1992, echo=F, eval=T, fig.width = 14, fig.height = 8.5, warning = F, message = F}

tribble(
  ~Year, ~Car, ~Sales,
  1985, "Hyundai Excel", NA,
  1986, "Hyundai Excel", 168882,
  1987, "Hyundai Excel", 263610,
  1988, "Hyundai Excel", 264282,
  1989, "Hyundai Excel", 148563,
  1990, "Hyundai Excel", 100590,
  1991, "Hyundai Excel", 66376,
  1992, "Hyundai Excel", 42324,
  1985, "Yugo", 3895,
  1986, "Yugo", 35910,
  1987, "Yugo", 48617,
  1988, "Yugo", 31583,
  1989, "Yugo", 10391,
  1990, "Yugo", 5659,
  1991, "Yugo", 2509,
  1992, "Yugo", 1311,
  1985, "Toyota Tercel", 88841,
  1986, "Toyota Tercel", 76914,
  1987, "Toyota Tercel", 100590,
  1988, "Toyota Tercel", 104655,
  1989, "Toyota Tercel", 97577,
  1990, "Toyota Tercel", 90808,
  1991, "Toyota Tercel", 102043,
  1992, "Toyota Tercel", 96173
) %>%
#  filter(Car == "Yugo") %>%
  mutate(narr = prettyNum(Sales, big.mark=",")) %>%
  ggplot(., aes(factor(Year), Sales, group=Car, color=Car)) +
  geom_bar(aes(fill=Car), stat="identity", position="dodge", alpha=0.8, color="black") +
  theme_steve_web() + scale_fill_brewer(palette="Set1") +
  scale_y_continuous(labels=scales::comma) +
  xlab("Year") +
  geom_text(aes(label=narr, group=Car), color="black",
            position=position_dodge(width=.9), size=3, vjust = -.5, family="Open Sans") +
  labs(title = "The Yugo Was Easily Outsold by Similar Models from Asian Automakers and Bottomed Out the U.S. Market Quickly",
       subtitle ="Despite its competitive price, the Yugo was a junker from the start and became a punchline for car aficionados as well as development scholars critical of ISI.",
       caption = "Data: carsalesbase.com. Note: I'm aware the inclusion of the Tercel is questionable since the third generation of Tercels were quite different from the first and second generations. Just roll with it.")

```

### The EOI Framework

Latin America generally turned inward. The "Asian Tigers" looked outward.

- Market products for quality for richer countries (esp. America).
- Depress currency to make exports cheap.
- Extend low-interest loans and tax breaks to exporters toward that end.

###

```{r bra-rok-auto-export-quality-passenger-cars, echo=F, eval=T, fig.width = 14, fig.height = 8.5, warning = F, message = F}
  
eq_passengercars %>%
  filter(ccode %in% c(140, 732)) %>%
  filter(category == "Export Quality Index") %>%
  rename(Country = country) %>%
  ggplot(.,aes(year, value, color=Country, group=Country,linetype=Country)) + 
  theme_steve_web() +
  geom_line(size=1.1) +
  scale_x_continuous(breaks=seq(1964, 2014, by =2)) +
  scale_color_manual(values=c("#009B3A", "#C60C30")) +
  xlab("Year") + ylab("Export Quality Index") +
  labs(title = "South Korean Passenger Automobiles Surpassed Brazil's Autos in Export Quality Early into the 1970s",
       subtitle = "South Koreas's first foray into the distant foreign market may have been Guam (through Hyundai) but cracking the U.S. market soon set it apart from Brazil.",
       caption = "Data: International Monetary Fund")

```


###

```{r bra-rok-auto-production-1950-2017, echo=F, eval=T, fig.width = 14, fig.height = 8.5, warning = F, message = F}


mvprod %>%
  filter(country %in% c("South Korea", "Brazil")) %>%
  rename(Country = country) %>%
  mutate(value = mvprod/1000,
         lbl = prettyNum(round(value), ",")) %>%
  ggplot(.,aes(as.factor(year), value, fill=Country)) + 
  theme_steve_web() +
  geom_bar(stat="identity", position = "dodge", color="black") +
  scale_y_continuous(labels = scales::comma) +
  xlab("Year") + ylab("Motor Vehicle Production (in Thousands)") +
  geom_text(aes(label=lbl, group=Country), color="black",
            position=position_dodge(width=.9), size=3, vjust = -.5, family="Open Sans") +
  labs(title = "It Didn't Take That Long For South Korea's Motor Vehicle Production to Outpace Brazil",
       subtitle = "South Korea's model found an audience abroad, producing products for foreign consumption in contrast to Brazil's domestic focus.",
       caption = "Data: Organisation Internationale des Constructeurs d'Automobiles, among other sources.\nNote: production includes figures for passenger cars, light commercial vehicles, minibuses, trucks, buses and coaches.\nProduction is later defined in the 21st century as where vehicle was assembled.") +
  scale_fill_manual(values=c("#009B3A", "#C60C30"))
  
```



### Commodity Cartels

Poorer countries sharing common resource have found cartels useful

- OPEC is clearly best example of this.

###

```{r wti-crude-price-1947-present, fig.height=8.5, fig.width = 14, echo=F, eval = T, message =F, warning = F}

# CPIAUCSL: urban consumers
# CPALCY01USM661N: Consumer Price Index: Total, All Items for the United States 
# Urban consumers, all items, has longer reach so I'll go with that.
fredr(series_id = "CPIAUCSL",
                      observation_start = as.Date("1947-01-01")) %>%
  rename(cpi = value) -> Price

last_date <- last(Price$date)
last_date_ym <- paste(lubridate::month(last_date, label=T, abbr=F), lubridate::year(last_date))
# zoo::as.yearmon(last_date)
my_ylab = paste0("Inflation Adjusted Price, in ", last_date_ym, " USD Dollars")
my_caption = paste0("Data: Federal Reserve Bank of St. Louis\nNote: prices are for West Texas Intermediate (WTI), manually converted from nominal to readl prices with a ", last_date_ym, " index.")
my_title = paste0("Average Crude Oil Price, 1947-",lubridate::year(last_date))
  

fredr(series_id = "WTISPLC",
             observation_start = as.Date("1947-01-01")) %>%
  rename(wti = value,
         seriesid = series_id) %>%
  left_join(., Price, by="date") %>%
  select(date, wti, cpi) %>%
  mutate(last = last(cpi),
         index = (cpi/last)*100,
    real = (wti/index)*100) %>%
  ggplot(.,aes(date, real, group=1)) + geom_line() + theme_steve_web() +
  scale_x_date(date_breaks = "4 years",
               date_labels = "%Y") +
  geom_ribbon(aes(ymin=0, ymax=real),
              alpha=0.3, fill="blue") +
  ylab(my_ylab) + xlab("") +
  annotate("text", x = as.Date("1973-07-01"), y = 75, 
           label = "Arab oil embargo,\nYom-Kippur War\n(1973)",
           family="Open Sans") +
 annotate("text", x = as.Date("1980-01-01"), y =135, 
          label = "Iranian revolution,\nIran-Iraq War\n(1979/80)",
          family = "Open Sans") +
  annotate("text", x = as.Date("2005-01-01"), y = 140, 
           label = "Energy crisis,\nrecession\n(2007/8)",
           family = "Open Sans") +
  annotate("text", x = as.Date("2013-01-01"), y = 140, 
           label = "Arab Spring\n(2011)",
           family = "Open Sans") +
  labs(title = my_title,
       subtitle = "We observe prominent spikes that coincide with the Arab oil embargo, Iran-Iraq War onset, and the worst consequences of the Iraq War and corollary financial crisis.",
       caption = my_caption) 
  # theme(
  #   axis.text.x = element_text(angle=45,size=7, vjust = 0.5)
  # )

```

###

```{r opec-production-1972-2018, echo=F, eval=T, fig.width = 14, fig.height = 8.5, warning = F, message = F}

#tibble(year = seq(1960, 1987),
#       opecv = c(8.7, 9.36, 10.51, 11.51, 12.98, 14.34, 15.77, 16.85, 18.79, 20.91, 23.3, 25.21, 26.89,
#                 30.63, 30.35, 26.77, 30.33, 30.89, 29.46, 30.58, 26.61, 22.48, 18.78, 17.5, 17.44, 16.18, 18.28, 18.52))

Oil <- readxl::read_xlsx("~/Dropbox/teaching/posc1020/development-2/opec-oil-production.xlsx", skip=10) %>% slice(-1)

Oil %>%
  mutate_at(vars(starts_with("Crude")),funs(as.numeric)) %>%
  select(Month, `Crude Oil Production, Total OPEC`) %>%
  ggplot(.,aes(Month, `Crude Oil Production, Total OPEC`)) + 
  theme_steve_web() +
  geom_line() + 
  xlab("") +
  ylab("Crude Oil Production, Total OPEC") +
  scale_x_datetime(breaks = seq(as.POSIXct("1972-01-01"), 
                            as.POSIXct("2018-01-01"), by = "2 years"),
    # date_breaks = "2 years",
                   date_labels = "%Y") +
  scale_y_continuous(label = scales::comma) +
  annotate("text", x = as.POSIXlt("1973-07-01"), y = 33000, 
           label = "Arab oil embargo,\nYom-Kippur War\n(1973)",
           family="Open Sans", size=3) +
 annotate("text", x = as.POSIXlt("1980-01-01"), y =32000, 
          label = "Iranian revolution,\nIran-Iraq War\n(1979/80)",
          family = "Open Sans", size=3) +
  annotate("text", x = as.POSIXlt("2007-01-01"), y = 35000, 
           label = "Energy crisis,\nrecession\n(2007/8)",
           family = "Open Sans", size=3) +
  annotate("text", x = as.POSIXlt("2011-01-01"), y = 35000, 
           label = "Arab Spring\n(2011)",
           family = "Open Sans", size=3) +
  geom_ribbon(aes(ymin=-Inf, ymax=`Crude Oil Production, Total OPEC`),
              alpha=0.3, fill="blue") +
  labs(title = "OPEC is a Clear Example of a Commodity Cartel Changing the Terms on Which They Trade Goods",
       subtitle = "OPEC's first success was doubling the price of oil in 1973 in response to the Yom Kippur War. Overall oil production was halved in the 10 years from 1973 to 1983.",
       caption = "Data: U.S. Energy Information Administration (Feb. 2018 Monthly Energy Review)")

```

### Commodity Cartels

However, commodity cartels are no quick fix.

- Cartels are hard to govern!
- Oil is a unique commodity too.

###

```{r uae-opec-production-quota-1982-2000, echo=F, eval=T, fig.width = 14, fig.height = 8.5, warning = F, message = F}

Oil %>%
  mutate_at(vars(starts_with("Crude")),funs(as.numeric)) %>%
  rename_at(vars(starts_with("Crude")), funs(str_sub(., 23, str_length(.)))) %>%
  select(Month, `United Arab Emirates`) %>%
  filter(Month > "1982-03-01") %>%
  mutate(quota = ifelse(Month >= "1982-03-01" & Month <= "1983-03-01", 1000, NA),
         quota = ifelse(Month >= "1983-03-01" & Month <= "1984-10-01", 1100, quota),
         quota = ifelse(Month >= "1984-10-01" & Month <= "1986-08-01", 950, quota),
         quota = ifelse(Month >= "1986-08-01" & Month <= "1986-12-01", 950, quota),
         quota = ifelse(Month >= "1986-12-01" & Month <= "1987-06-01", 902, quota),
         quota = ifelse(Month >= "1987-06-01" & Month <= "1988-12-01", 948, quota),
         quota = ifelse(Month >= "1988-12-01" & Month <= "1989-06-01", 988, quota),
         quota = ifelse(Month >= "1989-06-01" & Month <= "1989-09-01", 1041, quota),
         quota = ifelse(Month >= "1989-09-01" & Month <= "1989-12-01", 1094, quota),
         quota = ifelse(Month >= "1989-12-01" & Month <= "1990-07-01", 1095, quota),
         quota = ifelse(Month >= "1990-07-01" & Month <= "1991-03-01", 1500, quota), # Nominally just Aug. 1990
         quota = ifelse(Month >= "1991-03-01" & Month <= "1991-09-01", 2320, quota),
         quota = ifelse(Month >= "1992-01-01" & Month <= "1992-09-01", 2244, quota),
         quota = ifelse(Month >= "1992-12-01" & Month <= "1993-02-01", 2260, quota),
         quota = ifelse(Month >= "1993-02-01" & Month <= "1993-09-01", 2161, quota),
         quota = ifelse(Month >= "1993-09-01" & Month <= "1997-12-01", 2161, quota),
         quota = ifelse(Month >= "1997-12-01" & Month <= "1998-03-01", 2366, quota),
         quota = ifelse(Month >= "1998-03-01" & Month <= "1998-06-01", 2125, quota),
         quota = ifelse(Month >= "1998-06-01" & Month <= "2000-06-01", 2157, quota),
         quota = ifelse(Month >= "1999-03-01" & Month <= "2000-03-01", 2244, quota),
         quota = ifelse(Month >= "2000-06-01" & Month <= "2000-09-01", 2219, quota),
         quota = ifelse(Month >= "2000-09-01" & Month <= "2000-10-01", 2289, quota),) %>%
  rename(Quota = quota,
         Production = `United Arab Emirates`) %>%
  filter(Month < "2000-01-01") %>%
  gather(Category, Value, Quota:Production) %>%
  ggplot(.,aes(Month, Value, color=Category)) + theme_steve_web() +
  geom_line(size=1.1) +
  scale_x_datetime(breaks = seq(as.POSIXct("1982-04-01"), 
                            as.POSIXct("2000-01-01"), by = "1 year"),
    # date_breaks = "2 years",
                   date_labels = "%Y") +
  scale_y_continuous(label = scales::comma) +
  xlab("") + ylab("Crude Oil Production (Thousand Barrels per Day)") +
  labs(title = "Cartels Are Inherently Unstable. OPEC is No Different. And Everyone---like the United Arab Emirates---Cheats",
       subtitle = "Some OPEC members (e.g. Indonesia) cheat less than others but everyone does it and it kinda defeats the purpose of the cartel.",
       caption = "Data: U.S. Energy Information Administration (Feb. 2018 Monthly Energy Review)
       Quota Data: 2016 OPEC Annual Statistical Bulletin")
  

```

###

```{r rsa-kuw-uae-production-1972-2018, echo=F, eval=T, fig.width = 14, fig.height = 8.5, warning = F, message = F}

Oil %>%
  mutate_at(vars(starts_with("Crude")),funs(as.numeric)) %>%
  rename_at(vars(starts_with("Crude")), funs(str_sub(., 23, str_length(.)))) %>%
  select(Month, `Saudi Arabia`, `United Arab Emirates`, Kuwait) %>%
  gather(Country, Value, `Saudi Arabia`:ncol(.)) %>%
  ggplot(.,aes(Month, Value, linetype=Country, color=Country)) + theme_steve_web() +
  geom_line(size=1.1) +
  scale_y_continuous(labels = scales::comma) +
  scale_x_datetime(breaks = seq(as.POSIXct("1972-01-01"), 
                            as.POSIXct("2018-01-01"), by = "2 years"),
    # date_breaks = "2 years",
                   date_labels = "%Y") +
  ylab("Crude Oil Production (Thousand Barrels per Day") + xlab("") +
  labs(title = "Saudi Arabia's Response to Widespread Cheating in the 1980s: Beat Everyone to the Floor",
       subtitle = "Saudi Arabia has the largest oil reserves, a rather small population, and huge currency reserves.\nIt can take a bloody nose for the cause of the cartel and beat anyone to the floor when it's concerned about market share.",
       caption = "Data: U.S. Energy Information Administration (Feb. 2018 Monthly Energy Review).")
# Saudi Arabia found that its position in the 1980s as 'swing producer' just meant it was losing market share. Thereafter, it started beating OPEC states to the floor.
```




# Conclusion
### Conclusion

- While everyone prefers development, powerful groups can block it.
- Domestic institutions play an important role—they may promote or hinder development.
- Rich countries adopt policies that hurt the poor.
- Successful development requires a country to overcome both internal and external obstacles.