Data Bias in Architectural Sources

Lingbo Li

Abstract

This study explores potential biases within architectural design practices, focusing on how architects cater to various income groups, particularly low-income populations, and the challenges associated with designing small living spaces. Utilizing ArchDaily, a comprehensive repository of global architectural projects, the study examines 1,423 housing projects in the United States, categorized into houses, apartments, and social housing. The research aims to identify and analyze biases in the representation of these housing types, hypothesizing a preference for high-income residential projects. The findings indicate a significant bias towards high-income projects, both in terms of number and scale, with a predominance of larger, luxurious properties. This bias is evident in the broader range of sizes and greater design complexity seen in high-income housing. Additionally, the study reveals an underrepresentation of social housing, highlighting a skewed architectural narrative that may limit resources and inspiration for affordable housing solutions. The study underscores the need for a more balanced approach in showcasing diverse housing solutions to address the growing housing affordability crisis.

Introduction

Objective: To explore biases in architectural designs, focusing on how these designs cater to various income groups, especially low-income populations, and the challenges associated with small living spaces.

Primary Source: ArchDaily, known for its extensive database of architectural projects and influence in the architecture community.

Research Question

Main Question: To identify and analyze biases in architectural designs on ArchDaily, specifically examining if there is a preference for high-income residential projects over those aimed at lower-income groups or smaller housing options.

Methodology

Dataset: 1,423 housing projects from ArchDaily, categorized into houses, apartments, and social housing.

Data Collected: Basic information (year of completion, location, size), visual documentation (photos showcasing architectural style, materials, and layout).

Tools Used: Excel for data sorting, R for statistical analysis, Adobe Illustrator for visual presentation.

Hypotheses

High-Income Bias: Hypothesizes a disproportionate representation of high-income housing projects in terms of number and scale.

Visual Representation Bias: Higher-income projects are expected to exhibit greater design complexity and innovation.

Resource Allocation Bias: Architectural resources are allocated more generously to high-income projects.

Trend Analysis: Bias towards projects from wealthier, more influential cities.

Data Visualization and Analysis

Housing Crisis Metrics: Housing prices to per capita income ratios and the Housing Affordability Index show significant declines in affordability from 2020 to 2022.

Distribution of Projects: House projects dominate the dataset (85%), with apartments (14%) and social housing (<1%) being significantly less represented.

Area Size Distribution: Houses show a wide range of sizes, apartments have smaller, more uniform sizes, and social housing projects are consistently small.

Geographic Distribution

House Projects: Concentrated in cities like Los Angeles, Austin, San Francisco, Bend, and Boulder.

Apartment Projects: Predominantly in New York, San Francisco, and Los Angeles.

Conclusion

Bias in Project Representation: There is a notable bias towards high-income residential projects, with a focus on large, luxurious properties.

Impact on Architectural Narrative: The underrepresentation of social housing projects suggests a skewed architectural narrative, potentially limiting resources and inspiration for affordable housing solutions.

Discussion

Editorial Choices and Bias: ArchDaily's project selection appears to favor high-income, high-quality projects often found in affluent regions. This bias is concerning given the lack of representation for social housing, which could influence the availability of resources and inspiration for architects working on affordable housing solutions.

In summary, the study uncovers significant biases in architectural design representation on ArchDaily, favoring high-income projects and underrepresenting social housing. This skewed representation can impact the architectural community's focus and resource allocation, highlighting the need for a more balanced approach to showcasing diverse housing solutions.