Consumer behavior is significantly influenced by color, reflecting psychological and emotional responses. Previous research has shown correlations between color preferences and purchasing decisions across various product categories, including vehicles and footwear, with price often reflecting these preferences. While the impact of building facade color on building popularity has been studied in different regions, research specifically examining the impact on residential purchasing behavior and price, particularly across diverse consumer groups, is lacking. This study addresses this gap by innovatively using housing prices as an evaluation criterion to explore the color preferences and consumption psychology of different consumer groups categorized by purchasing power and spatial regions in Fuzhou, China. Fuzhou's diverse population, high volume of residential transactions, and cultural similarities with other East Asian regions make it a suitable location for this study, with the potential for broader applicability.

Existing literature highlights the impact of color on consumer behavior. Studies have shown that color preferences are linked to psychological responses (Palmer and Schloss, 2010; Sable and Akcay, 2011), and that certain colors attract more attention and drive sales (Raizada, 2012). However, color preferences are also satisfiable only when the product itself is beneficial (Yu et al., 2021), and preferences can be group-specific due to shared cultural experiences, though individual variations exist (Labrecque et al., 2013; Iris et al., 2015). Previous research has examined the influence of building facade colors on popularity, using methods like experimental approaches and questionnaires (Cubukcu and Kahraman, 2008; Kaya and Crosby, 2010; Li et al., 2020; Gou and Wang, 2017; Gou et al., 2021; Li et al., 2022; Zhai et al., 2023). However, these studies often lacked a direct examination of the pricing impact of color attributes on consumer purchasing behavior and lacked focus on variations in affordability and geographic distribution among consumer groups. This research aims to bridge this gap.

This research employs a two-part framework. Part I focuses on developing research variables for building facade colors, categorizing them into color quantity (number of colors) and color feature variables (lightness, saturation, hue, color balance, and differences between primary and secondary colors). Part II analyzes consumer preferences in relation to housing prices, using cross-sectional transaction data from 906 second-hand residential communities in Fuzhou, China, in 2020. The data includes outcome variables (average transaction price), control variables (location environment, self-characteristics, facility accessibility), and research variables (facade color variables). The selection criteria for residential communities ensured consistent housing types, architectural styles, and sufficient transaction records. Data acquisition for color variables involved computer-aided street image recognition and on-site colorimetry using a colourimeter. Three regression analysis methods were used: linear regression (to analyze preferences of all consumption groups), quantile regression (to analyze preferences of groups with different consumption abilities), and spatial regression (geographically weighted regression – to analyze preferences of groups with different geographic spatial locations). The hedonic price model formed the theoretical foundation for the regression analyses.

Linear regression analysis revealed a strong fit and explanatory power (R-squared > 0.76). Control variables showed expected correlations with housing prices, consistent with previous research. The key findings regarding the research variables are as follows: **For all consumer groups:** A preference for more colors (more than one) on building facades was observed. A preference for red and blue color tendencies was also found. High contrast in color saturation and low contrast in color lightness between primary and secondary colors was favored. **For consumer groups with different purchasing power:** Lower-income groups showed a stronger preference for a greater number of colors. Higher-income groups were less concerned with the number of colors. Lower-income groups preferred brighter, less-saturated primary colors, whereas high-income groups preferred darker, high-saturation colors. Lower-income groups favored high contrast in saturation between primary and secondary colors, while high-income groups preferred high contrast in lightness. Middle- and low-consumption groups showed clear preferences regarding color contrasts, while the high-consumption group showed more diversity in their preferences. Low-consumption groups generally preferred harmony and coordination of colors, high-consumption groups preferred vibrant contrasts, and middle-consumption groups showed nuanced preferences involving color saturation. **For consumer groups with different geographic locations:** Spatial regression analysis indicated significant geographical variations in color preferences. Groups sensitive to the quantity of colors were spread out geographically, but those preferring four or more colors concentrated in eastern Fuzhou. Preferences for the red-green color tendency of the primary color were clustered in commercial areas, with a contrast between areas where red was preferred (indicating vitality) and areas where it was less favored (like the historical Three Lanes and Seven Alleys, showing preference for simplicity). Preferences for the yellow-blue color tendency showed concentric urban spatial stratification, with peripheral groups favoring yellow (potentially linked to wealth) and downtown groups disfavoring it. Groups sensitive to the differences between primary and secondary colors also displayed geographic clustering, reflecting varied priorities in harmony, contrast and color balance.

The findings highlight the complexity of color preferences, showing both commonalities and significant variations across consumer groups defined by income and location. The preference for multiple colors might reflect a connection between color quantity and construction cost (in second-hand properties, implying a premium driven by consumer preference). The preference for red and blue tendencies aligns with prior research, suggesting universal emotional associations. However, the variations across income levels and locations provide insights into how individual circumstances and cultural context shape aesthetic preferences. Higher-income groups might value uniqueness and sophistication, reflected in their preference for less conventional color schemes. The spatial distribution of preferences also indicates the importance of considering local context and the influence of neighborhood characteristics.

This research provides a novel methodology for studying color preferences in urban residential settings, utilizing housing prices as a proxy for consumer preference. The findings reveal nuanced preferences for building facade colors across various consumer groups, highlighting the importance of considering income, location, and the interaction of these factors in urban planning and housing development. Future research should consider expanding the range of color variables, exploring other influencing factors (materials, architectural style), and conducting cross-cultural comparative studies.

This research is limited in its geographical scope, focusing on Fuzhou, China, and may not be generalizable to other regions. The analysis focuses primarily on color quantity and features, and other aspects of facade design might influence preferences. The inferences regarding psychological states of consumer groups are based on interpretations of color preferences, and direct psychological measures would strengthen the analysis. Finally, cultural factors and their influence on color preferences were only implicitly considered. Future research should extend the study to a broader geographic area and incorporate more direct measures of psychological factors, socioeconomic status and culture.