Why People Don't Open Their Windows, and Why That Matters for Social Housing

What the research is actually about 

Marzieh's focus is on residents living in naturally ventilated flats, in London, many without mechanical ventilation or any meaningful cross-ventilation. For these residents, opening a window is often the only practical way to manage heat and air quality during warm weather. And yet, as Marzieh puts it, we still know surprisingly little about how people use their windows, what motivates them to open or close them, and what the consequences are for overheating and indoor air quality. 

Over two consecutive summers in 2024 and 2025, she monitored around 55 apartments, collecting indoor and outdoor environmental data, building characteristics and neighbourhood surveys, and crucially, window opening behaviour itself: not just whether windows were open or closed, but the angle at which they were opened. She also gathered responses from residents about the drivers behind their behaviour, asking directly what motivated their decisions. 

What the data is already showing 

Even before her analysis is complete in 2027, there are some early findings she has shared. In some of the bedrooms she monitored, night-time temperatures reached 30 to 32 degrees Celsius.  That is the lived experience of people sleeping in dense urban areas where the urban heat island effect compounds already high ambient temperatures. 

We know from existing research that 26 degrees is roughly the threshold beyond which sleep becomes significantly disturbed. The current guidance in TM59 (the industry standard methodology used to predict and assess overheating in residential buildings) assumes that at that point, residents will open their windows. What Marzieh's data is showing is that, for many people, that simply does not happen. 

The reasons are more varied and more human than the standard models assume. Noise pollution and air quality are the most cited factors, with residents weighing up whether the benefit of cooler air is worth the cost of traffic noise or pollution coming in from outside. Security concerns matter too, particularly in dense neighbourhoods where opening a window at night can feel unsafe. And for a small number of residents, there are personal factors that also influence the decision, though Marzieh is careful to note that this requires further investigation before drawing firm conclusions. 

The core finding so far is clear: the generalised, simplified assumptions embedded in Part O and TM59 do not reflect what real people actually do. Any overheating risk assessment that relies on those assumptions without accounting for real occupant behaviour is likely to underestimate risk for the people who need protection the most. 

Why this matters for social landlords 

The residents in Marzieh's study are small flats in dense urban environments with limited ventilation options. Cost challenges mean air conditioning is rarely a realistic alternative. These are the households most exposed to the health consequences of overheating, and they are the households whose behaviour is least likely to match the neat assumptions in our regulatory frameworks. 

As social landlords think about retrofit, about new build standards, about asset investment decisions, understanding that window opening behaviour varies significantly based on neighbourhood, orientation, social background, and individual circumstance is not a minor technical detail. It goes to the heart of whether the homes we provide are genuinely safe in a warming climate. 

Marzieh is also exploring temporal patterns in behaviour, separating night-time from daytime use, and weekday from weekend patterns, alongside residents' working patterns, including shift work and non-standard hours. This granularity will be important as we try to build regulation that actually reflects the diversity of how people live. 

One year to go 

Marzieh has approximately one more year left on her PhD, and she is currently in the data integration phase, using data mining approaches to identify the most influential parameters and classify behaviour patterns. When she completes her research, the implications for overheating standards, ventilation guidance, and how we assess risk in social housing will be significant. 

I am looking forward to sharing more from her when the work is done. In the meantime, I wanted to flag this research to colleagues across the sector because it is a reminder that the gap between what our models assume and what our residents actually experience is sometimes very different.