Equity distribution (from Survey of Consumer Finances)

README.md

Replicating Matt Bruenig's analysis. To repeat the analysis below run make in a terminal, or download and unzip the .dta file from the url in the Makefile, then open README.Rmd in Rstudio and run render it.

draw <- as_data_frame(read_stata('rscfp2013.dta'))

set.seed(111)
d <- sample_n(draw, nrow(draw) * 10, replace=TRUE, weight=wgt)

breaks_d <- 0:10 * 0.1
d %>% select(equity) %>%
    mutate(raw_quantile=cume_dist(equity), total_equity=sum(equity)) %>%
    mutate(`people who own more equity than __% of the population`=cut(raw_quantile,
                include.lower=TRUE,
                breaks=breaks_d,
                labels=paste0(100 * breaks_d[-length(breaks_d)], "%"))) %>%
    group_by(`people who own more equity than __% of the population`) %>%
        summarize(`percentage owned (of all equity)`=
          round(sum(equity) / unique(total_equity) * 100, 2))

## # A tibble: 5 x 2
##   people who own more equity than __% of the population percentage owned (of all equity)
##                                                  <fctr>                            <dbl>
## 1                                                   50%                             0.10
## 2                                                   60%                             0.79
## 3                                                   70%                             2.82
## 4                                                   80%                             9.63
## 5                                                   90%                            86.65

How do things look by age?

# from: https://www.federalreserve.gov/econresdata/scf/files/bulletin.macro.txt
#*   age of the household head, and categorical variable:
#    1:<35, 2:35-44, 3:45-54, 4:55-64, 5:65-74, 6:>=75;
#    AGE=X14;
#        AGECL=1+(AGE GE 35)+(AGE GE 45)+(AGE GE 55)+(AGE GE 65)+(AGE GE 75);
d$age_cl <- recode_factor(factor(d$agecl), `1`="<35",`2`="35-44",
          `3`="45-54", `4`="55-64", `5`="65-74", `6`=">=75")
d_eqd_age <-  d %>% select(equity, age_cl) %>%
    mutate(raw_quantile=cume_dist(equity), total_equity=sum(equity)) %>%
    mutate(`people who own more equity than __% of the population`=cut(raw_quantile,
                include.lower=TRUE,
                breaks=breaks_d,
                labels=paste0(100 * breaks_d[-length(breaks_d)], "%"))) %>%
    group_by(age_cl, `people who own more equity than __% of the population`) %>%
        summarize(`percentage owned (of all equity)`=
          round(sum(equity) / unique(total_equity) * 100, 2))
ggplot(d_eqd_age) +
    geom_bar(aes(`people who own more equity than __% of the population`,
             `percentage owned (of all equity)`, fill=age_cl),
         stat='identity') +
    ylim(0, 100) +
    guides(fill=guide_legend("Age")) +
    labs(
         title='Who owns the stock market?',
         caption='Data source: https://www.federalreserve.gov/econresdata/scf/\nMethods: https://gist.github.com/ashander/407997acb06eb68fceb3cb68fc9de438\nBuilding on: https://medium.com/@MattBruenig/who-gains-from-dow-20-000-ba07555e5f12'
         )

How do things look with 5-percentile bins?

breaks_v <- c(0, 0.05, 0.1, 0.15, 0.2, 0.25, 0.3, .35, .4,
          .45, .5, .55, .6, .65, .7, .75,  .8, .85,  .9, .95, 1)
d_eqv <- d %>% select(equity, age_cl) %>%
    mutate(raw_quantile=cume_dist(equity), total_equity=sum(equity)) %>%
    mutate(`people who own more equity than __% of the population`=cut(raw_quantile,
                include.lower=TRUE,
                breaks=breaks_v,
                labels=paste0(100 * breaks_v[-length(breaks_v)], "%"))) %>%
    group_by(`people who own more equity than __% of the population`, age_cl) %>%
        summarize(`percentage owned (of all equity)`=
          round(sum(equity) / unique(total_equity) * 100, 2))

ggplot(d_eqv) +
    geom_bar(aes(`people who own more equity than __% of the population`,
             `percentage owned (of all equity)`, fill=age_cl),
         stat='identity') +
    ylim(0, 100) +
    guides(fill=guide_legend("Age")) +
    labs(
         title='Who owns the stock market?',
         caption='Data source: https://www.federalreserve.gov/econresdata/scf/\nMethods: https://gist.github.com/ashander/407997acb06eb68fceb3cb68fc9de438\nBuilding on: https://medium.com/@MattBruenig/who-gains-from-dow-20-000-ba07555e5f12'
         )

README.Rmd

---
title: "Equity ownership distribution"
output: md_document
---



```{r setup, echo=FALSE}
suppressMessages(library(haven))
suppressMessages(library(dplyr))
suppressMessages(library(ggplot2))
options(width=100)
```

Replicating [Matt Bruenig's analysis](https://medium.com/@MattBruenig/who-gains-from-dow-20-000-ba07555e5f12).
To repeat the analysis below run `make` in a terminal, or download and unzip
the .dta file from the url in the Makefile, then open README.Rmd in Rstudio and
run render it.

```{r data, cache=TRUE}
draw <- as_data_frame(read_stata('rscfp2013.dta'))
```
```{r reweight, cache=TRUE}
set.seed(111)
d <- sample_n(draw, nrow(draw) * 10, replace=TRUE, weight=wgt)
```

My result agrees with Bruenig's (up to some error introduced by resampling the data based on the weights).
```{r deciles}
breaks_d <- 0:10 * 0.1
d %>% select(equity) %>%
	mutate(raw_quantile=cume_dist(equity), total_equity=sum(equity)) %>%
	mutate(`people who own more equity than __% of the population`=cut(raw_quantile,
			    include.lower=TRUE,
			    breaks=breaks_d,
			    labels=paste0(100 * breaks_d[-length(breaks_d)], "%"))) %>%
	group_by(`people who own more equity than __% of the population`) %>%
        summarize(`percentage owned (of all equity)`=
		  round(sum(equity) / unique(total_equity) * 100, 2))
```

How do things look by age?
```{r equity-deciles-by-age, fig.cap="Distribution of equity owned by each 5%-ile bin broken into age classes."}
# from: https://www.federalreserve.gov/econresdata/scf/files/bulletin.macro.txt
#*   age of the household head, and categorical variable:
#    1:<35, 2:35-44, 3:45-54, 4:55-64, 5:65-74, 6:>=75;
#    AGE=X14;
#        AGECL=1+(AGE GE 35)+(AGE GE 45)+(AGE GE 55)+(AGE GE 65)+(AGE GE 75);
d$age_cl <- recode_factor(factor(d$agecl), `1`="<35",`2`="35-44",
	      `3`="45-54", `4`="55-64", `5`="65-74", `6`=">=75")
d_eqd_age <-  d %>% select(equity, age_cl) %>%
	mutate(raw_quantile=cume_dist(equity), total_equity=sum(equity)) %>%
	mutate(`people who own more equity than __% of the population`=cut(raw_quantile,
			    include.lower=TRUE,
			    breaks=breaks_d,
			    labels=paste0(100 * breaks_d[-length(breaks_d)], "%"))) %>%
	group_by(age_cl, `people who own more equity than __% of the population`) %>%
        summarize(`percentage owned (of all equity)`=
		  round(sum(equity) / unique(total_equity) * 100, 2))
ggplot(d_eqd_age) +
	geom_bar(aes(`people who own more equity than __% of the population`,
		     `percentage owned (of all equity)`, fill=age_cl),
		 stat='identity') +
	ylim(0, 100) +
	guides(fill=guide_legend("Age")) +
	labs(
	     title='Who owns the stock market?',
	     caption='Data source: https://www.federalreserve.gov/econresdata/scf/\nMethods: https://gist.github.com/ashander/407997acb06eb68fceb3cb68fc9de438\nBuilding on: https://medium.com/@MattBruenig/who-gains-from-dow-20-000-ba07555e5f12'
	     )
```

How do things look with 5-percentile bins?
```{r equity-vigesimals, fig.cap="Distribution of equity owned by each 5%-ile bin."}
breaks_v <- c(0, 0.05, 0.1, 0.15, 0.2, 0.25, 0.3, .35, .4,
	      .45, .5, .55, .6, .65, .7, .75,  .8, .85,  .9, .95, 1)
d_eqv <- d %>% select(equity, age_cl) %>%
	mutate(raw_quantile=cume_dist(equity), total_equity=sum(equity)) %>%
	mutate(`people who own more equity than __% of the population`=cut(raw_quantile,
			    include.lower=TRUE,
			    breaks=breaks_v,
			    labels=paste0(100 * breaks_v[-length(breaks_v)], "%"))) %>%
	group_by(`people who own more equity than __% of the population`, age_cl) %>%
        summarize(`percentage owned (of all equity)`=
		  round(sum(equity) / unique(total_equity) * 100, 2))

ggplot(d_eqv) +
	geom_bar(aes(`people who own more equity than __% of the population`,
		     `percentage owned (of all equity)`, fill=age_cl),
		 stat='identity') +
	ylim(0, 100) +
	guides(fill=guide_legend("Age")) +
	labs(
	     title='Who owns the stock market?',
	     caption='Data source: https://www.federalreserve.gov/econresdata/scf/\nMethods: https://gist.github.com/ashander/407997acb06eb68fceb3cb68fc9de438\nBuilding on: https://medium.com/@MattBruenig/who-gains-from-dow-20-000-ba07555e5f12'
	     )
```

equity-deciles-by-age-1.png

equity-vigesimals-1.png

Makefile

README.md: README.Rmd
	Rscript -e 'rmarkdown::render("README.Rmd")'
	cp README_files/figure-markdown_strict/equity-vigesimals-1.png equity-vigesimals-1.png
	cp README_files/figure-markdown_strict/equity-deciles-by-age-1.png equity-deciles-by-age-1.png

rscfp2013.dta: scfp2013s.zip
	unzip scfp2013s.zip

scfp2013s.zip:
	wget https://www.federalreserve.gov/econresdata/scf/files/scfp2013s.zip