As the climate crisis worsens, reducing carbon emissions has become one of the most pressing priorities to mitigate its effects. In the United States in 2020, the residential sector was responsible for approximately 20 per cent of total carbon emissions from energy consumption. Targeted housing policies that increase residential energy efficiency, such as retrofitting, are therefore an important step to reduce overall carbon emissions.  

Researchers had already observed a positive association between income levels and carbon emissions. However, no study had looked at the relationship between race and carbon emissions in residential energy consumption. To address this knowledge gap, Benjamin Goldstein, an assistant professor in McGill’s Department of Bioresource Engineering, published a study along with Tony Reames and Joshua Newell of the University of Michigan that examined whether energy efficiency leads to low-carbon households in the U.S., and if these carbon emissions vary by race and ethnicity. 

Goldstein and his colleagues collected data on housing property attributes such as heating systems and floor area from CoreLogic, a privately owned database that contains property information from across the U.S. They also obtained data on energy use by source from the 2015 U.S. Residential Energy Consumption Survey (RECS), and estimated carbon emissions according to the carbon intensities of different fuel types. They classified each zip code area by the majority race of the residents in that area, and estimated the energy use and carbon emissions of these areas using regression models. 

The study found a significant negative association between energy use intensity and housing quality. Poor housing quality, which on average disproportionately affects predominantly Black neighbourhoods due to historically racist policies like redlining, was associated with higher energy use. 

It also found a significant positive association between per capita carbon emissions and per capita floor area. On average, the floor area per capita is 41 per cent lower in majority Latinx and 23 per cent lower in majority Black neighbourhoods than majority white neighbourhoods. Although Black neighbourhoods have more energy inefficient homes than white neighbourhoods, they emit less carbon on average. 

Goldstein explained that racist policies like redlining account for part of this discrepancy. Introduced in the 1930s, the Federal Housing Administration’s redlining policy marked Black, immigrant, and other racialized neighbourhoods as undesirable, making it difficult or nearly impossible for those communities to access mortgage loans for home investments. The effects of this decades-long policy that stretched into the mid-20th century are still reverberating across America.

“​​The poor energy efficiency [the study identified] in these African-American neighbourhoods can be traced back to the legacy of this discriminatory housing policy,” Goldstein said in an interview with The McGill Tribune.  

Due to the historical barriers impeding Black homeownership, renting is more common among Black neighbourhoods. The study found a significant association between rental status and energy use intensity. Due in part to the little motivation for landlords to install retrofits, neighbourhoods with higher rental rates have more energy inefficient homes, exacerbating the racial wealth gap and raising electricity bills. 

The floor area per capita in formerly redlined neighbourhoods is also 19 per cent less than in non-redlined neighbourhoods, outweighing the effect of poor energy efficiency on overall energy consumption.

“Affluent neighbourhoods have much higher emissions. This is driven primarily by household size,” Goldstein said. “The square footage of the homes, no matter how efficient your home is, will outpace those savings in energy and still make you a monster emitter.” 

Looking to the future, Goldstein wants to recreate the study in a Canadian context, but acknowledges that the methodology would likely have to be drastically different due to a lack of data availability. Canada does not have an equivalent of the American residential energy use survey that was used as a data source for this study, nor is real-estate information as easily accessible.

“I would love to do it in Canada, [but] we simply do not have the data,” Goldstein said. “I would like to challenge our government to actually collect useful data for researchers to use.”