conclusions.qmd

---
title: "Conclusions"
---

```{r setup}
#| message: false
#| echo: false
library(tidyverse)
library(sf)
library(tigris)
library(gt)
library(ggrepel)

# read in county dataframe
df_counties <- read_rds("intermediate/df_countries_proced.rds")

# store year range in variables}
years <- levels(df_counties$year)
start_year <- first(years)
last_year <- last(years)

start_year_num <- as.numeric(start_year)
last_year_num <- as.numeric(last_year)
```

Overall, dependency rations are increasing from `r start_year_num` to `r last_year_num` in Washington State. This is driven primarily by a strong upward trend in the aged dependency ratio. The statewide child dependency ratio is declining, but with local counter trend

The following maps show how all three dependency ratios are changing across the state over time.

```{r join geometry and demographic data}
#| echo: false
#| warning: false

# Inner join the geometries with the demographic data. This will allow us to
# generate maps that include county boundaries and the dependency ratios as fill
# data.

wa_county <- counties(state = "WA", cb = TRUE, class = "sf", progress_bar = FALSE)

df_joined <- wa_county |>
  inner_join(df_counties, by=c("NAME"="geography"))

base_map <- df_joined |> 
  ggplot() +
  # for the same colour scale for all dependency maps with limits
  scale_fill_continuous(type = "viridis",limits=c(0, 100)) +
  facet_wrap(~year) +
  labs(
    caption="Data source: Washington State County Demography Dashboard"
  ) +
  theme_void() +
  theme(
    plot.caption.position = "plot",
    plot.title.position = "plot",
    plot.subtitle = element_text(size=14),
    plot.title = element_text(size=18),
    strip.text = element_text(size = 12),
    legend.position = "bottom"
  )
```

```{r generate CDR map}
#| code-fold: true
#| code-summary: "Show the code"
#| fig-height: 6
#| fig-width: 8
base_map +
  geom_sf(aes(fill=child_dep_ratio), colour="white") +
  labs(
    title="Child dependency ratios for Washington State counties",
    subtitle=paste("Ratios stay relatively flat from", start_year, "to", last_year, "\n"),
    fill="Child dependency ratio"
  )
```

```{r generate ADR map}
#| code-fold: true
#| code-summary: "Show the code"
#| fig-height: 6
#| fig-width: 8
base_map +
  geom_sf(aes(fill=aged_dep_ratio), colour="white") +
  labs(
    title="Aged dependency ratios for Washington State counties",
    subtitle=paste("Ratios increase from", start_year, "to", last_year, "\n"),
    fill="Aged dependency ratio"
  )
```

```{r generate TDR map}
#| code-fold: true
#| code-summary: "Show the code"
#| fig-height: 6
#| fig-width: 8
base_map +
  geom_sf(aes(fill=total_dep_ratio), colour="white") +
  labs(
    title="Total dependency ratios for Washington State counties",
    subtitle=paste("Ratios increase from", start_year, "to", last_year, "\n"),
    fill="Total dependency ratio"
  )
```

## Conclusions and key takeaways

* There are regional differences in dependency ratios.
* The highest ratios are on the Pacific coast and in the eastern quarter of the state.
* Between these two regions there are distinct differences in dependency ratios with the highest rates of aged and total dependency being on the coast.
* The counties with the highest total dependency ratios have rates of change 2-5 times higher than the statewide rate.
* Given that the statewide total dependency ratio is increasing, the state should consider the long term impacts on revenue and services expenditures.
* The counties with the highest total dependency ratios may need to consider how their population dynamics will impact the tax base.