1) join income deprivation score to LSOA shapefiles. If you could produce a map that would be great.
2) Count dwellings in LSOAs with have EPC < C (D-G) and main fuel type not mains gas. If you could map all EPCs and then the EPCs that match the criteria.
3) Join tenure to LSOA shapefiles. Multiply count by LSOA level census figure for % of private dwellings.
4) Multiply by 1/0.687 to correct for dwellings without an EPC.
For the Isle of Wight, can we estimate the number of private (owned or rented) addresses in high income deprivation areas which do not have gas?
Our method:
1. Join LSOA data to EPC data to
join IMD income deprivation score to LSOA.
2. Count dwellings in LSOAs with have EPC < C (D-G) and main fuel type not mains gas. If you could map all EPCs and then the EPCs that match the criteria.
3. Join tenure to LSOA shapefiles. Multiply count by LSOA level census figure for % of private dwellings.
4. Multiply by 1/0.687 to correct for dwellings without an EPC.
```{r loadData}
# this will break here if you don't have the data obvs
...
...
@@ -427,21 +432,26 @@ Table \@ref(tab:depr10pc) shows the same results but for the 10% most income dep
Next we assume that the EPC addresses of interest are as likely to be social/private rented as owner-occupied and multiply the EPC address counts by the Census-derived household ownership tenure rate for each LSOA. Without knowing the tensure status of the EPC we cannot do anything else.
> In summary using the EPC data we have estimated the number of addresses in low income areas which are owned and do not have gas to be:
> In summary using the EPC data we have estimated the number of addresses in high income deprivation areas which do not have gas to be:
* 10% most deprived: `r res10pc` or, if we uprate to allow for the `r epcMissingPC` % missing EPCs, `r round(res10pc * epcMissingFactor)`;
* 20% most deprived: `r res20pc` or, if we uprate to allow for the `r epcMissingPC` % missing EPCs, `r round(res20pc * epcMissingFactor)`;
* 10% most deprived:
* Owned: `r res10pcOwned` or, if we uprate to allow for the `r epcMissingPC` % missing EPCs, `r round(res10pcOwned * epcMissingFactor)`
* Private (owned & private rent): `r res10pcPrivate` or, if we uprate to allow for the `r epcMissingPC` % missing EPCs, `r round(res10pcPrivate * epcMissingFactor)`
* 20% most deprived:
* Owned: `r res20pcOwned` or, if we uprate to allow for the `r epcMissingPC` % missing EPCs, `r round(res20pcOwned * epcMissingFactor)`
* Private (owned & private rent): `r res20pcPrivate` or, if we uprate to allow for the `r epcMissingPC` % missing EPCs, `r round(res20pcPrivate * epcMissingFactor)`
Table \@ref(tab:lsoaTable) shows the LSOA level results.
As a rough check, if we multiply the total number of EPC addresses which fit the criteria by the mean ownership rate we get:
As a rough check, if we multiply the total number of EPC addresses which fit the criteria by the mean private ownership rate (owned + private rental = `r round(100*meanPrivate)`%) we get:
* 10% of households = `r round(res10)`
* 20% of households = `r round(res20)`
* 10% of households:
* Private: `r round(res10priv)`
* Owned: `r round(res10own)`
* 20% of households:
* Private: `r round(res20priv)`
* Owned: `r round(res20own)`
Note that this is only a rough guide since it assumes that LSOAs have exactly the same number of addresses/households in them.
Note that these values have not been uprated to correct for missing EPCs and that they are only a rough guide since they assume that LSOAs have exactly the same number of addresses/households in them.
