goliaths.Rmd

---
title: "Global Goliaths"
author: "Manuel Alcala Kovalski"
date: "`r Sys.Date()`"
output:
  prettydoc::html_pretty:
    theme: cayman
    highlight: github
    fig_width: 4
    fig_height: 6
editor_options: 
  chunk_output_type: console
---

```{r setup, include = FALSE}

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  warning = FALSE,
  message = FALSE,
  results = 'asis',
  tidy.opts=list(width.cutoff=60),
  tidy=TRUE, 
  echo = FALSE,
      fig.ext = "png",
  #cache = TRUE,
  fig.align = "center",
  fig.show = "hold",
  fig.width=8, 
  fig.asp=0.9,
  out.width = "100%",
  fig.path = "figures/",
  fig.retina = 2,
  fig.showtext = TRUE,
  fig.pos = "t",
  dpi = if (knitr::is_latex_output()) 72 else 300,
  dev.args = list(png = list(type = "cairo-png")), 
  optipng = "-o1 -quiet"
)


library(knitr)
knitr::knit_hooks$set(optipng = hook_optipng)
knitr::knit_hooks$set(pngquant = hook_pngquant)
```

```{r packages, include = FALSE}
librarian::shelf('tidyverse', 'glue', 'tsibble', 'readxl', 'janitor', 'ggthemes', 'ggtext','ggrepel', 'RColorBrewer', 'nfa', 'sysfonts', 'showtext', 'readxl', 'scales', 'gghutchins', 'directlabels', 'gghighlight', 'gglaplot')
conflicted::conflict_prefer('lag', 'dplyr')
conflicted::conflict_prefer('filter', 'dplyr')
conflicted::conflict_prefer('geom_line', 'ggplot2')
conflicted::conflict_prefer('geom_col', 'ggplot2')
conflicted::conflict_prefer('geom_text', 'ggplot2')
conflicted::conflict_prefer('geom_point', 'ggplot2')

ggplot2::theme_set(gghutchins::theme_hutchins())

 caption_text <- '<br>**Note:** The original figure appears in *Global Goliaths*, Brookings Institution Press, 2021. <br>
**Source**: Bureau of Economic Analysis.'

```

## Figure 1

```{r fig1}
mnc_share <-
  readxl::read_xlsx("data/role-mnc.xlsx")

p1 <- mnc_share %>% 
  pivot_longer(where(is.numeric)) %>% 
  mutate(name = fct_relevel(name, 
                            "U.S. Imports", "U.S. Exports", "Industrial R&D", "Capital Expenditures", "Manufacturing Employee Compensation", "Manufacturing Employment", "Employee Compensation", "Employment")) %>% 
  ggplot(aes(y = name,
             x = value,
             fill = mnc_type)) +
  geom_col(width = 0.7) +
  stat_summary(fun = sum, aes(x = value , label = scales::percent(after_stat(x), accuracy = 0.1),  group = name), geom = "text", position = position_nudge(x = 0.06), family = 'Roboto Medium', fontface = 'bold') 

p1 +
  scale_x_continuous(
    name = "",
    labels = scales::percent_format(accuracy = 1),
    limits = c(0, 1)
  ) +
  scale_y_discrete(name = '') +
   scale_fill_hutchins(
    'qual',
    rev = TRUE,
    name = '',
    labels = c("U.S. arms of foreign-headquartered firms", "U.S. parents")
  )  +

    guides(fill = guide_legend(reverse = T, nrow = 1)) +
  

    labs(
    title = "**Multinationals are major players in the U.S. economy**",
    subtitle = "Multinationals' share of economic activity in 2017, by category<br>",
    caption = '<br>**Note:** The original figure appears in *Global Goliaths*, Brookings Institution Press, 2021. <br>
**Source**: Bureau of Economic Analysis, National Income and Product Accounts; National Science Foundation, Science and Engineering Indicators; Census Bureau Annual Capital Expenditure Survey.'
  ) +
   theme(legend.title = element_blank(), 
     legend.position = c(-0.467,1.025), 
     legend.justification= 'left',
     legend.box = "horizontal",
     legend.margin = margin(0),
     legend.key = element_blank(),
     legend.direction = "horizontal",
     plot.title.position = "plot",
     plot.caption.position =  "plot",
     plot.caption = element_textbox_simple(hjust = 0)
     )


```

## Figure 2

```{r fig2,fig.asp=.9}
df <- readxl::read_xlsx("data/employment_share.xlsx") %>%
    janitor::clean_names() %>%
    rename_with( ~snakecase::to_title_case(.),
                 ends_with('income')) %>% 
    fill(everything(), .direction = "up") %>%
    mutate(
      across(
        .cols = !year & !all_countries,
        .fns = ~ .x / all_countries,
        .names = "{.col}"
      )
    ) %>%
    select(-all_countries) %>%
    pivot_longer(-year) 


  ggplot(df) +
  geom_line(aes(x = year, y = value, group = name), alpha = 0.5) +
    coord_cartesian(clip = 'off') +
geom_dl(aes(x = year, y = value, label = stringr::str_remove(name, ' Income'), size = 0.1),method = list(dl.trans(x = x + .1),'last.bumpup'), cex = 0.2)+
  geom_line(df %>% filter(name == 'High Income'),
           mapping =  aes(x = year, y = value), 
           color = '#003A79', size = 1.3) +
       scale_x_continuous(
      name = "",
      breaks = seq(1982, 2021, 5),
      limits  = c(1982, 2020),
      expand = expansion(mult = c(0, 0.1))
    ) +
    scale_y_continuous(
      name = "",
      labels = scales::percent,
      limits = c(0, 0.8),
      breaks = seq(0, 0.8, 0.15),
      expand = expansion(mult = c(0.05, .1))
    ) +
    labs(
      title = "The bulk of multinationals' employees are in high-wage economies",
      subtitle = "Share of U.S. multinational foreign affiliate employment, by host country income",
      caption = caption_text 
    ) 
  
  
```

## Figure 3

```{r fig3, fig.asp = 0.9}
df <-
  readxl::read_xlsx("data/fig_3.xlsx") %>% 
  clean_names() 




df <- 
  df %>% 
  mutate(
    across(
      -c('date', 'total'),
      ~ .x / total
    )
  ) %>% 
  select(-total, -other) %>% 
  pivot_longer(-date) %>% 
  mutate(name = factor(name,
                       levels = c(
                         "asia",
                         "canada",
                         "europe",
                         "latin_am"
                       ),
                       labels = c(
                         "Asia",
                         "Canada",
                         "Europe",
                         "Latin America"
                       )
  ))
df_minus <- df %>% mutate(name2 = name) %>% select(-name)

df %>%  ggplot(aes(x = date, y = value, group = name)) +
   geom_line(alpha = 2, size = 1.5, aes(group = name), color = hutchins_colors['green']) +
 #  geom_line(data = df_minus, aes(x = date, y = value, group = name2), color = 'grey', alpha = 0.4, size = 1.1) +
   theme_hutchins() +
    coord_cartesian(clip = 'off') +
   labs(title = "Multinationals' markets have shifted toward Asia",
        subtitle = 'Share of U.S. multinational foreign affiliate sales, by affiliate location',
        caption = caption_text) +
   scale_y_continuous(name = '', labels = scales::percent_format(accuracy = 1),  limits = c(0,0.65),
                      breaks = seq(0, 0.65, 0.1),
                      expand = expansion(mult = c(0, 0))) +
    scale_x_continuous(name = '',
                       limits = c(1982, 2017),
                       breaks = seq(1982, 2017, 5),
                         expand = expansion(mult = c(0.1, .2))) +
        geom_dl(aes(label = name), method = list(dl.trans(x = x + .2),'last.bumpup'), cex = 0.8) +
     gghighlight(name == 'Asia',
                  use_group_by = FALSE,
                  use_direct_label = FALSE,
                  unhighlighted_params = list(size = 0.2, alpha = 0.9, color = 'gray50'))
  
   

 
```

## Figure 4

```{r fig4, fig.asp = 0.9}
readxl::read_xlsx('data/data.xlsx', sheet = 'figure 4') %>% 
  summarize(date, 
            `Employment` = parent_employment / total_employment,
            `Capital spending` = parent_capital_expenditures /total_capital_expenditures) %>% 
  
  pivot_longer(-date) %>% 
  mutate(label = if_else(date == max(date), as.character(name), NA_character_)) %>%

  ggplot(aes(x = date, y = value, color = name)) +
  geom_line(size = 1.4, alpha = 0.8) +
  scale_x_continuous(name = '', limits = c(1981, 2020), breaks = seq(1982, 2017, 5),
                       expand = expansion(c(0.1, 0.15))) +
  scale_color_hutchins() +
  scale_y_continuous(name = '',labels = scales::percent_format(accuracy = 1), 
                     limits = c(0,0.45), breaks = seq(0, 0.6, 0.05),
                     expand = expansion(0)) +
    coord_cartesian(clip = 'off') +
  labs(title = 'U.S. operations of U.S. multinationals have grown at the same pace as the economy',
       subtitle = 'U.S. multinational parent share of U.S. private sector activity',
       caption = caption_text) +
  geom_dl(aes(label = name), method = list(dl.trans(x = x + .2),'last.bumpup'), cex = 0.8,
          color = 'black') +
  theme(legend.position = 'none')
  
```

## Figure 5

```{r fig5, fig.asp = .9}
read_xlsx('data/data.xlsx', sheet = 'figure 5') %>% 
  summarize(date,
            share = parents / total) %>% 
  pivot_longer(share) %>% 
  ggplot(aes(x = date, y = value, label = value)) +
  geom_line(size = 1.1, alpha =  0.8, color = hutchins_colors['brookings blue']) +
  
  scale_y_continuous(name = '', labels = scales::percent_format(accuracy = 1), limits = c(0.55, 0.85), breaks = seq(0.55, 0.85, 0.05),
                     expand = expansion(mult = c(0.15, 0))
                 ) +
  labs(title ='U.S. multinationals still do the bulk of U.S. private sector R&D spending',
       subtitle = 'U.S. parent share of U.S. business R&D',
       caption = '**Note:** The original figure appears in *Global Goliaths*, Brookings Institution Press, 2021. This figure also previously appeared in "The Rise of Global Innovation by U.S. Multinationals Poses Risks and Opportunities", Peterson Institute for International Economics, 2019.<br>**Source:** Bureau of Economic Analysis.') +
  scale_x_continuous(name = '', limits = c(1989, 2018), breaks = seq(1989, 2017, 5),
                       expand = c(0.01, 0)) 

```

## Figure 6

```{r fig6, fig.asp = .9}
get_label <- function(df){
  df %>% 
    mutate(label = if_else(date == max(date), as.character(name), NA_character_))
}


add_label <- function(){
    geom_label_repel(aes(label = label),
                  nudge_x = 1,
                  na.rm = TRUE)
}
read_xlsx('data/data.xlsx', sheet = 'figure 6') %>% 
  summarize(date,
            Exports = exports / all_exports,
            Imports = imports /all_imports) %>% 
  pivot_longer(-date) %>% 
  ggplot(aes(x = date, y = value, color = name)) +
  geom_line(size = 1.3, alpha = 0.75) +
  scale_color_manual(values = c(hutchins_colors['vivid blue'] ,"#FF9E1B" )) +
  scale_y_continuous(labels = percent_format(accuracy = 1),
                     name = '',
                     limits = c(0.25, 0.45),
                     breaks = seq(0.25, 0.5, 0.05),
                     expand = expansion(mult = c(0.2, 0))) +
  geom_dl(aes(label = name), method = 'last.points', color = 'black') +
  labs(title = 'Intra-company trade has grown at the same pace as overall international trade',
       subtitle = 'Related party share of U.S. imports and exports',
       caption = caption_text) +
  scale_x_continuous(name = '', limits = c(1981, 2020), breaks = seq(1982, 2017, 5),
                       expand = c(0, 0.5)) +
  theme(
    legend.text = element_blank(),
    legend.position = 'none'
  )


```

## Figure 7

```{r fig7, out.width='100%', fig.asp = .9}
read_xlsx('data/data.xlsx', sheet = 'figure 7') %>% 
  pivot_longer(investment) %>% 
  get_label() %>% 
  ggplot(aes(x = date, y = value,  color = name)) +
  geom_line(size = 1.2,alpha=0.7)+
  scale_color_hutchins(pal = 'mixed') +
  scale_y_continuous(labels = percent_format(accuracy = 1), limits = c(0, 0.35), breaks = seq(0, 0.35, 0.05),
                     expand = expansion(mult = 0)) +
  scale_x_continuous(limits = c(1982, 2017), breaks = seq(1982, 2017, 5)) +
  labs(title = "A growing share of multinationals' profits come from abroad",subtitle = "U.S. firms' income on equity investment abroad as share of worldwide profits",
       caption = caption_text) +
  theme(legend.position =  'none')


```

```{r fig8}

caption_text_fig8 <- '<br>**Note:** The original figure appears in *Global Goliaths*, Brookings Institution Press, 2021. <br>
**Source**: Bureau of Economic Analysis and Dhammika Dharmapala.'

read_xlsx('data/data.xlsx', sheet = 'figure 8') %>% 
  mutate(label = glue::glue('{100 * round(value, 3)}%')) %>% 
  ggplot(aes(x = reorder(name, value), y = value )) +
  geom_col(fill = hutchins_colors['gray'], alpha = 1, width = 0.4) +
  scale_fill_hutchins() +
  scale_y_continuous(labels = percent_format(accuracy = 1),
                     limits = c(0,0.55),
                     breaks = seq(0,0.55, 0.1),
                     expand = expansion(0.02))+
  coord_flip() +
  guides(fill = guide_legend(reverse = T)) +
  labs(
       title = "Percentage of U.S. multinationals' foreign activity in tax havens",
       caption = caption_text_fig8) +
  geom_dl(aes(label = label), hjust = -0.25, method = list(dl.trans(x = x + .2),'last.bumpup'), cex = 0.8,  font = 'Robot Medium', fontface = 'bold',
          color = 'black')


```



```{r fig9}
read_xlsx('data/data.xlsx', 'figure 9') %>% 
  summarize(date,
            share = tax_haven / total) %>%
  mutate(date = as.character(date)) %>%
pivot_longer(share) %>% 
  ggplot(aes(x = date, y = value, group = 1)) +
  geom_line(color = 'grey60', size = 1) + geom_point(color = hutchins_colors['brookings blue'], size = 4)+
  #geom_col(fill=hutchins_colors['brookings blue'], width = 0.7) + 
  scale_y_continuous(label  = percent_format(accuracy = 1),
                     limits = c(0,0.6),
                     breaks = seq(0,0.6, 0.1),
                     expand = expansion(mult = c(0,0))) +
  labs(title = "Multinationals' use of tax havens has increased" ,
       subtitle = 'Share of U.S. multinational parent firms with affiliates in tax havens',
       caption = '<br>**Note:**  The original figure appears in *Global Goliaths*, Brookings Institution Press, 2021. Tax Haven countries are those identified in James R. Hines Jr. "Treasure islands." Journal of Economic Perspectives 24.4 (2010): 103-26, with the exception of the Cook Islands, the UK Channel Islands, and Niue, which cannot be separately identified in the BEA data. 
       <br> **Source:** Bureau of Economic Analysis. 
    ') +
  theme(
    plot.caption = element_textbox_simple(color = '#666666')
  )

```

```{r fig10}
read_xlsx('data/data.xlsx', sheet = 'figure 10') %>%
  summarize(date, traditional, non_traditional) %>%
  pivot_longer(-date) %>%
  ggplot(aes(x = date, y = value, fill = name,  group  = name)) +
  geom_area(position = 'stack') +
  scale_fill_manual(labels =  c('Non-traditional Hubs', 'Traditional Hubs'),
                    values = c("#f26d00", "#003A70")) +
  scale_y_continuous(labels = dollar_format(), limits = c(0, 60000), breaks = seq(0, 60000, 10000 / 2),
                     expand = expansion(0)) +
  scale_x_continuous(expand = expansion(0)) +
  labs(title = "U.S. multinationals' foreign R&D spending increasingly done in non-traditional hubs",
       subtitle = "Total R&D spending by U.S. multinationals' foreign affiliates, not adjusted for inflation (millions of US $)",
       caption = '<br>**Note:** Traditional hubs are the United Kingdom, Germany, France, Canada, and Japan. Non-traditional hubs are all other countries (mainly Israel, China, and India). The original figure appears in *Global Goliaths*, Brookings Institution Press, 2021. <br>
       **Source:** Bureau of Economic Analaysis.') +
     theme(legend.title = element_blank(), 
     legend.position = 'top', 
     legend.justification='left',
     legend.box = "horizontal",
     legend.margin = margin(0),
     legend.key = element_blank(),
     legend.direction = "horizontal",
     plot.title.position = "plot",
     plot.caption.position =  "plot",
     plot.caption = element_textbox_simple(hjust = 0)
     ) +
   guides(fill = guide_legend(reverse = T, nrow = 1)) 
  

```

## Figure 8

```{r fig11}
df <- readxl::read_xlsx('data/data.xlsx', sheet = 'figure 1-7') 

df %>% summarize(date, 
                 share = foreign / total) %>% 
  pivot_longer(share) %>% 
  ggplot(aes(x = date, y = value, color = name)) +
  geom_line(size = 1.2) +
  scale_color_hutchins() +
  scale_y_continuous(labels = scales::percent_format(accuracy = 1), 
                     limits = c(0, 0.4),
                     breaks = seq(0, 0.4, 0.05),
                     expand = expansion(0)) +
  scale_x_continuous(limits = c(1982, 2017),
                     breaks = seq(1982, 2017, by = 5)) +
  labs(title = "A growing share of U.S. multinationals' workers are overseas",
       subtitle = "Share of employees of U.S. multinationals who work abroad", 
       caption = caption_text)  +
  theme(legend.position = 'none')
```