Table 1.1.6. Real Gross Domestic Product, Chained Dollars (A) (Q)
Data - BEA
François Geerolf
Info
| source | dataset | .html | .RData |
|---|---|---|---|
| bea | T10105 | 2024-05-19 | 2023-12-17 |
| bea | T10106 | 2024-05-20 | 2024-05-19 |
Data on US macro
| source | dataset | .html | .RData |
|---|---|---|---|
| fred | gdp | 2024-05-10 | 2024-05-10 |
| fred | unr | 2024-05-10 | 2024-05-10 |
| oecd | QNA | 2024-05-05 | 2024-04-15 |
| oecd | SNA_TABLE1 | 2024-04-16 | 2024-04-15 |
LAST_COMPILE
| LAST_COMPILE |
|---|
| 2024-05-20 |
Last
| date | Nobs |
|---|---|
| 2024-03-31 | 27 |
Basic Macroeconomic Facts
U.S. Real GDP (1929-2019)
Linear
plot_linear <- T10106 %>% filter(FREQ == "A") %>%
year_to_date %>%
filter(LineNumber == 1) %>%
ggplot(.) + xlab("") + ylab("U.S. Real GDP") +
geom_line(aes(x = date, y = DataValue/1000000)) +
theme_minimal() +
scale_x_date(breaks = nber_recessions$Peak,
labels = date_format("%y")) +
scale_y_continuous(breaks = 1*seq(0, 50, 2.5),
labels = scales::dollar_format(accuracy = 0.1, suffix = "Tn")) +
geom_rect(data = nber_recessions %>%
filter(Peak > as.Date("1928-01-01")),
aes(xmin = Peak, xmax = Trough, ymin = 0, ymax = +Inf),
fill = 'grey', alpha = 0.5)
plot_linear
Log
plot_log <- plot_linear +
scale_y_log10(breaks = 20*2^seq(-7, 1, 1),
labels = scales::dollar_format(accuracy = 0.1, suffix = "Tn")) +
geom_rect(data = nber_recessions %>%
filter(Peak > as.Date("1928-01-01")),
aes(xmin = Peak, xmax = Trough, ymin = 0, ymax = +Inf),
fill = 'grey', alpha = 0.5)
plot_log
Other
Quarterly
All
T10106 %>% filter(FREQ == "Q") %>%
quarter_to_date %>%
filter(LineNumber == 1) %>%
ggplot(.) + xlab("") + ylab("U.S. Real GDP") +
geom_line(aes(x = date, y = DataValue/1000000)) +
theme_minimal() +
scale_x_date(breaks = nber_recessions$Peak,
labels = date_format("%y")) +
scale_y_log10(breaks = 20*2^seq(-7, 1, 1),
labels = scales::dollar_format(accuracy = 0.1, suffix = "Tn")) +
geom_rect(data = nber_recessions %>%
filter(Peak > as.Date("1945-01-01")),
aes(xmin = Peak, xmax = Trough, ymin = 0, ymax = +Inf),
fill = 'grey', alpha = 0.5)
1970-
T10106 %>% filter(FREQ == "Q") %>%
quarter_to_date %>%
filter(LineNumber == 1,
date >= as.Date("1970-01-01")) %>%
ggplot(.) + xlab("") + ylab("U.S. Real GDP") +
geom_line(aes(x = date, y = DataValue/1000000)) +
theme_minimal() +
scale_x_date(breaks = c(nber_recessions$Peak),
labels = date_format("%Y")) +
scale_y_log10(breaks = 20*2^seq(-7, 1, 0.2),
labels = scales::dollar_format(accuracy = 0.1, suffix = "Tn")) +
geom_rect(data = nber_recessions %>%
filter(Peak > as.Date("1970-01-01")),
aes(xmin = Peak, xmax = Trough, ymin = 0, ymax = +Inf),
fill = 'grey', alpha = 0.5)
1990-
T10106 %>% filter(FREQ == "Q") %>%
quarter_to_date %>%
filter(LineNumber == 1,
date >= as.Date("1990-01-01")) %>%
ggplot(.) + xlab("") + ylab("U.S. Real GDP") +
geom_line(aes(x = date, y = DataValue/1000000)) +
theme_minimal() +
scale_x_date(breaks = c(nber_recessions$Peak),
labels = date_format("%Y")) +
scale_y_log10(breaks = 20*2^seq(-7, 1, 0.1),
labels = scales::dollar_format(accuracy = 0.1, suffix = "Tn")) +
geom_rect(data = nber_recessions %>%
filter(Peak > as.Date("1990-01-01")),
aes(xmin = Peak, xmax = Trough, ymin = 0, ymax = +Inf),
fill = 'grey', alpha = 0.5)
2005-
T10106 %>% filter(FREQ == "Q") %>%
quarter_to_date %>%
filter(LineNumber == 1,
date >= as.Date("2005-01-01")) %>%
ggplot(.) + xlab("") + ylab("U.S. Real GDP") +
geom_line(aes(x = date, y = DataValue/1000000)) +
theme_minimal() +
scale_x_date(breaks = "2 years",
labels = date_format("%Y")) +
scale_y_log10(breaks = 20*2^seq(-7, 1, .1),
labels = scales::dollar_format(accuracy = 0.1, suffix = "Tn")) +
geom_rect(data = nber_recessions %>%
filter(Peak > as.Date("2005-01-01")),
aes(xmin = Peak, xmax = Trough, ymin = 0, ymax = +Inf),
fill = 'grey', alpha = 0.5)
U.S. Real GDP Trends (1929-2019) - Log Scale
T10106 %>% filter(FREQ == "A") %>%
year_to_date %>%
filter(LineNumber == 1) %>%
select(date, `Real GDP` = DataValue) %>%
mutate(`Real GDP` = `Real GDP` / 1000000,
year = year(date)) %>%
mutate(`HP (1000) Trend` = exp(hpfilter(log(`Real GDP`), freq = 1000, type = "lambda", drift = FALSE)$trend),
`HP (10000) Trend` = exp(hpfilter(log(`Real GDP`), freq = 10000, type = "lambda", drift = FALSE)$trend),
`Log-Linear Trend` = exp(fitted(lm(log(`Real GDP`) ~ year, na.action = na.exclude)))) %>%
group_by(date) %>%
select(-year) %>%
gather(variable, value, -date) %>%
ggplot(.) + geom_line(aes(x = date, y = value, color = variable, linetype = variable)) +
theme_minimal() +
scale_color_manual(values = viridis(5)[1:4]) +
theme(legend.title = element_blank(),
legend.position = c(0.8, 0.2)) +
scale_x_date(breaks = nber_recessions$Peak,
labels = date_format("%y")) +
scale_y_log10(breaks = 20*2^seq(-7, 1, 1),
labels = scales::dollar_format(accuracy = 0.1, suffix = "Tn")) +
geom_rect(data = nber_recessions %>%
filter(Peak > as.Date("1928-01-01")),
aes(xmin = Peak, xmax = Trough, ymin = 0, ymax = +Inf),
fill = 'grey', alpha = 0.5) +
xlab("") + ylab("U.S. Real GDP")
U.S. Real GDP Cycle (1929-2019)
T10106 %>% filter(FREQ == "A") %>%
year_to_date %>%
filter(LineNumber == 1) %>%
select(date, `Real GDP` = DataValue) %>%
mutate(`Real GDP` = `Real GDP` / 1000000,
year = year(date)) %>%
mutate(`HP (1000) Cycle` = exp(hpfilter(log(`Real GDP`), freq = 1000, type = "lambda", drift = FALSE)$cycle),
`HP (10000) Cycle` = exp(hpfilter(log(`Real GDP`), freq = 10000, type = "lambda", drift = FALSE)$cycle),
`Log-Linear Cycle` = exp(residuals(lm(log(`Real GDP`) ~ year, na.action = na.exclude)))) %>%
group_by(date) %>%
select(-year, - `Real GDP`) %>%
gather(variable, value, -date) %>%
ggplot(.) + geom_line(aes(x = date, y = value, color = variable, linetype = variable)) + theme_minimal() +
theme(legend.title = element_blank(),
legend.position = c(0.8, 0.8)) +
geom_rect(data = nber_recessions %>%
filter(Peak > as.Date("1928-01-01")),
aes(xmin = Peak, xmax = Trough, ymin = -Inf, ymax = +Inf),
fill = 'grey', alpha = 0.5) +
scale_x_date(breaks = nber_recessions$Peak,
labels = date_format("%y")) +
scale_y_continuous(breaks = seq(0.60, 1.40, 0.05),
labels = scales::percent_format(accuracy = 1)) +
xlab("") + ylab("GDP Cycle (% of Trend)")
U.S. Real GDP
fit <- T10106 %>% filter(FREQ == "A") %>%
year_to_date %>%
filter(LineNumber == 1) %>%
select(date, `Real GDP` = DataValue) %>%
mutate(`Real GDP` = `Real GDP` / 1000000,
year = year(date)) %>%
filter(year <= 1971) %>%
lm(log(`Real GDP`) ~ year, data = .) %>%
pluck("coefficients")
fit_2019 <- T10106 %>% filter(FREQ == "A") %>%
year_to_date %>%
filter(LineNumber == 1) %>%
select(date, `Real GDP` = DataValue) %>%
mutate(`Real GDP` = `Real GDP` / 1000000,
year = year(date)) %>%
lm(log(`Real GDP`) ~ year, data = .) %>%
pluck("coefficients")
fit_1929_1971 <- T10106 %>% filter(FREQ == "A") %>%
year_to_date %>%
filter(LineNumber == 1) %>%
select(date, `Real GDP` = DataValue) %>%
mutate(`Real GDP` = `Real GDP` / 1000000,
year = year(date)) %>%
filter(year <= 1971) %>%
lm(log(`Real GDP`) ~ year, data = .) %>%
pluck("coefficients")
fit_1929_1971# (Intercept) year
# -83.52871738 0.04327969fit_1972_2019 <- T10106 %>% filter(FREQ == "A") %>%
year_to_date %>%
filter(LineNumber == 1) %>%
select(date, `Real GDP` = DataValue) %>%
mutate(`Real GDP` = `Real GDP` / 1000000,
year = year(date)) %>%
filter(year >= 1972) %>%
lm(log(`Real GDP`) ~ year, data = .) %>%
pluck("coefficients")U.S. Real GDP Trends (1929-2019) - Log Scale
All
T10106 %>% filter(FREQ == "A") %>%
year_to_date %>%
filter(LineNumber == 1) %>%
select(date, `Real GDP` = DataValue) %>%
mutate(`Real GDP` = `Real GDP` / 1000000,
year = year(date)) %>%
mutate(`Trend to 2019 (g = 3.5%)` = lm(log(`Real GDP`) ~ year) %>% fitted %>% exp,
`Trend to 1971, extrapolated` = ifelse(year >= 1971, (fit[1] + fit[2] * year) %>% exp, NA),
`Trend to 1971 (g = 4.1%)` = ifelse(year <= 1971, (fit[1] + fit[2] * year) %>% exp, NA)) %>%
group_by(date) %>%
select(-year) %>%
gather(variable, value, -date) %>%
ggplot(.) + geom_line(aes(x = date, y = value, color = variable, linetype = variable)) +
theme_minimal() +
scale_color_manual(values = c(viridis(4)[3], viridis(4)[1], viridis(4)[1], viridis(4)[2])) +
scale_linetype_manual(values = c("solid", "solid", "dashed", "solid")) +
theme(legend.title = element_blank(),
legend.position = c(0.8, 0.2)) +
scale_x_date(breaks = nber_recessions$Peak,
labels = date_format("%y")) +
scale_y_log10(breaks = 20*2^seq(-7, 1, 1),
labels = scales::dollar_format(accuracy = 0.1, suffix = "Tn")) +
geom_rect(data = nber_recessions %>%
filter(Peak > as.Date("1928-01-01")),
aes(xmin = Peak, xmax = Trough, ymin = 0, ymax = +Inf),
fill = 'grey', alpha = 0.5) +
xlab("") + ylab("U.S. Real GDP, $2012")
Figure 1: U.S. Real GDP Trends (1929-2019) - Log Scale.
U.S. Real GDP Trends (1929-2019)
T10106 %>% filter(FREQ == "A") %>%
year_to_date %>%
filter(LineNumber == 1) %>%
select(date, `Real GDP` = DataValue) %>%
mutate(`Real GDP` = `Real GDP` / 1000000,
year = year(date)) %>%
mutate(`Trend to 2019 (g = 3.5%)` = lm(log(`Real GDP`) ~ year) %>% fitted %>% exp,
`Trend to 1971, extrapolated` = ifelse(year >= 1971, (fit[1] + fit[2] * year) %>% exp, NA),
`Trend to 1971 (g = 4.1%)` = ifelse(year <= 1971, (fit[1] + fit[2] * year) %>% exp, NA)) %>%
group_by(date) %>%
select(-year) %>%
gather(variable, value, -date) %>%
ggplot(.) + geom_line(aes(x = date, y = value, color = variable, linetype = variable)) +
theme_minimal() +
scale_color_manual(values = c(viridis(4)[3], viridis(4)[1], viridis(4)[1], viridis(4)[2])) +
scale_linetype_manual(values = c("solid", "solid", "dashed", "solid")) +
theme(legend.title = element_blank(),
legend.position = c(0.2, 0.8)) +
scale_x_date(breaks = nber_recessions$Peak,
labels = date_format("%y")) +
scale_y_continuous(breaks = 20*2^seq(-4, 1, 1),
labels = scales::dollar_format(accuracy = 0.1, suffix = "Tn")) +
geom_rect(data = nber_recessions %>%
filter(Peak > as.Date("1928-01-01")),
aes(xmin = Peak, xmax = Trough, ymin = 0, ymax = +Inf),
fill = 'grey', alpha = 0.5) +
xlab("") + ylab("U.S. Real GDP, $2012")
Figure 2: U.S. Real GDP Trends (1929-2019).
U.S. Real GDP Cycles (1929-2019)
T10106 %>% filter(FREQ == "A") %>%
year_to_date %>%
filter(LineNumber == 1) %>%
select(date, `Real GDP` = DataValue) %>%
mutate(`Real GDP` = `Real GDP` / 1000000,
year = year(date)) %>%
mutate(`Cycle, Detrending to 2019 (g = 3.5%)` = lm(log(`Real GDP`) ~ year) %>% residuals %>% exp,
`Cycle, Detrending to 1971, extrapolated` = ifelse(year >= 1971, exp(log(`Real GDP`) - fit[1] - fit[2] * year), NA),
`Cycle, Detrending to 1971 (g = 4.1%)` = ifelse(year <= 1971, exp(log(`Real GDP`) - fit[1] - fit[2] * year), NA)) %>%
group_by(date) %>%
select(-year, -`Real GDP`) %>%
gather(variable, value, -date) %>%
ggplot(.) + geom_line(aes(x = date, y = value, color = variable, linetype = variable)) +
theme_minimal() +
scale_color_manual(values = c(viridis(4)[1],viridis(4)[1], viridis(4)[2])) +
scale_linetype_manual(values = c("solid", "dashed", "solid")) +
theme(legend.title = element_blank(),
legend.position = c(0.7, 0.8)) +
scale_x_date(breaks = nber_recessions$Peak,
labels = date_format("%y")) +
scale_y_continuous(breaks = seq(0, 2.25, 0.10),
labels = scales::percent_format(accuracy = 1),
limits = c(0.4, 1.55)) +
geom_rect(data = nber_recessions %>%
filter(Peak > as.Date("1928-01-01")),
aes(xmin = Peak, xmax = Trough, ymin = -Inf, ymax = +Inf),
fill = 'grey', alpha = 0.5) +
xlab("") + ylab("U.S. Real GDP, $2012")
Figure 3: U.S. Real GDP Cycles (1929-2019).
Compare two part-linear and quadratic
dataset <- T10106 %>% filter(FREQ == "A") %>%
year_to_date %>%
filter(LineNumber == 1) %>%
select(date, `Real GDP` = DataValue) %>%
mutate(`Real GDP` = `Real GDP` / 1000000,
year = year(date)) %>%
select(year, `Real GDP`)
fit <- dataset %>%
filter(year <= 1971) %>%
lm(log(`Real GDP`) ~ year, data = .) %>%
pluck("coefficients")
fit_2019 <- dataset %>%
lm(log(`Real GDP`) ~ year, data = .) %>%
pluck("coefficients")
fit_1929_1971 <- dataset %>%
filter(year <= 1971) %>%
lm(log(`Real GDP`) ~ year, data = .) %>%
pluck("coefficients")
fit_1972_2019 <- dataset %>%
filter(year >= 1972) %>%
lm(log(`Real GDP`) ~ year, data = .) %>%
pluck("coefficients")Residuals
ssq1 <- dataset %>%
lm(log(`Real GDP`) ~ year, data = .) %>%
pluck("residuals") %>%
norm(., type = "2")
ssq2 <- dataset %>%
mutate(year2 = year^2) %>%
lm(log(`Real GDP`) ~ year + year2, data = .) %>%
pluck("residuals") %>%
norm(., type = "2")
ssq3a <- dataset %>%
filter(year >= 1972) %>%
lm(log(`Real GDP`) ~ year, data = .) %>%
pluck("residuals") %>%
norm(., type = "2")
ssq3b <- dataset %>%
filter(year <= 1971) %>%
lm(log(`Real GDP`) ~ year, data = .) %>%
pluck("residuals") %>%
norm(., type = "2")
ssq1# [1] 1.365111# [1] 0.7920961# [1] 1.1385931938, 1958, 1978, 1998, 2018 Table
Percent
T10106 %>%
filter(FREQ == "A") %>%
select(-FREQ) %>%
year_to_date %>%
mutate(year = year(date)) %>%
filter(year %in% c(1938, 1958, 1978, 1998, 2018)) %>%
group_by(year) %>%
mutate(value = round(100*DataValue/DataValue[1], 1)) %>%
ungroup %>%
select(2, 3, 6, 7) %>%
spread(year, value) %>%
{if (is_html_output()) datatable(., filter = 'top', rownames = F) else .}Billions
T10106 %>% filter(FREQ == "A") %>%
select(-FREQ) %>%
year_to_date %>%
mutate(year = year(date)) %>%
filter(year %in% c(1938, 1958, 1978, 1998, 2018)) %>%
group_by(year) %>%
mutate(value = round(DataValue/1000)) %>%
ungroup %>%
select(2, 3, 6, 7) %>%
spread(year, value) %>%
{if (is_html_output()) datatable(., filter = 'top', rownames = F) else .}