Historically, scholars have researched fashion cycles (mostly in women’s clothing) in terms of their existence and styles’ recurrence (Belleau 1987; Kroeber 1919; Lowe and Lowe 1982; Richardson and Kroeber 1940; Robinson 1976; Young 1937) and comparatively between countries (Balkwell and Ho 1992; Curran 1999). Because of practicing forecasters’ assumptions about the influence of colors in one product category on another (A Source 1985; Becker 2016; Linton 1994; Verlodt 1994) and the dearth of empirical research on fashion change in color (Stansfield and Whitfield 2005), this study compared the relative lengths, patterns, and timing of cycles or trends in HF and AP colors in terms of 10 hues (regardless of saturation or value).
Results confirmed and added to the literature on the existence of measurable color cycles in AP and HF. Using the broad perspective of hue, a wide range in cycle lengths was observed and led to the authors’ concluding differentiation of three length categories: short (4 years or less), medium (5–9 years), and long (10 or more years). Thus, excluding cycles that appeared to start before 1969 or end after 2009, 70 short, 28 medium, and 11 long cycles were identified. The range of cycle lengths was similar for HF and AP, contradicting Linton’s (1994) assertion that AP cycles were shorter than HF. The time frames were similar to those found by Ulrich and Lee (2008), who reported bell-shaped cyclical patterns ranging from 7 to 21 years for different floor covering types and some shorter cycles existing within longer term trends. Our study also observed short cycles within the long cycles of some colors (e.g. red in HF and black in AP). Although conclusions on the lengths of cycles in this study cannot be broadly generalized, results added new specificity to the conceptual descriptions of short to long cycles by Brannon (2010), Sproles (1981), and Sproles and Burns (1994).
Depending on the particular color, over the course of the 40-year study period, cycle lengths increased, decreased, remained relatively consistent, or fluctuated in length. Table 2 provides key highlights of findings in AP and HF by color. A single decreasing or increasing trend across all colors could not be concluded. Thus, the idea that fashion is generally moving faster now than in the past, as illustrated by the fast fashion phenomenon (Miller 2006), was not consistently substantiated for color across our findings. Perhaps that could be shown for color fashions if the perspective on color was expanded to study more than 10 hue families, such as looking at secondary and tertiary hues, as well as variations in value and saturation. The short cycles within long ones may have reflected such variations. However, the finding that cycle lengths trended longer for some colors (yellow, purple, brown in HF and black in AP) suggests that the rhythm of fashion change is not as predictable as early cycle researchers concluded (Kroeber 1919; Young 1937) and may not be as perfectly inevitable as Robinson (1958) avowed. At any time, fashion change is subject to varied influences (Nystrom 1928) and may demonstrate unpredictable changes as a result of this influence. Becker (2016), a practicing forecaster, recently noted that many factors can interrupt color trend predictions, causing them to be neither linear nor cyclical. Although our research did not explore possible influencing factors, results coincide with Nystrom’s and Becker’s observations in that actual color adoption in HF and AP did not follow consistent and related linear or cyclical paths.
Based on his own and previous research, Oberascher (1994) concluded that colors used in architecture and interior spaces did evolve in a cyclical pattern characterized by hue, saturation, value, and combinations thereof. The methodological specifications of the study reported here, which were governed by the goal of comparing two product categories using 10 hues, predetermined that findings could not be compared to his described recurring patterns. The comparative intent and 40-year period dictated that a research question on recurrence not be asked. However, visual observation of graphed patterns hints at recurrent cycles, particularly for colors with relatively more short to medium cycle lengths, suggesting a possible contradiction of Stansfield and Whitfield’s (2005) conclusion that there was no evidence of regularly repeating cycles in color.
In HF and AP, colors having the highest overall incidence were the ones that tended towards the longest and therefore fewest completed cycles. Colors with shorter cycles typically had more of them. HF brown, AP red and black, and HF and AP blue and white all had long cycles or lengthy fluctuating continuity; their patterned behavior could exemplify classic fashions (Brannon 2010). Ulrich and Lee (2008) found one category of rug surface (average height cut pile) could be characterized as a classic in comparison to other surfaces that showed cyclical or trending patterns (shag, looped styling). Colors with overall lower incidences (e.g., orange, gray) had more frequent periods of disappearance. Although colors are not generally cited as examples of fads, the short cyclical patterns and gaps in appearance of orange and gray marked them as something similar to that.
Some colors’ cycles demonstrated a mixed pattern of bouncing back and forth between short and medium lengths. Short cycles might follow long ones, suggesting that product developers might have been ready to move away from a color before consumers were, or that there was a brief interlude until a different version of the color was featured again. This could be explored in future research by studying more color variations. Had there not been the goal of directly comparing HF and AP, the jagged peaks and valleys of the graphed patterns could have been reduced by plotting 3 year averages (Richardson and Kroeber 1940; Ulrich and Lee 2008), achieving a smoother sense of cycles’ shapes and lengths.
Overall, there was little evidence of clearly parallel patterns of change between HF and AP colors. The data as a whole also did not conclusively support the idea of a time lag between a color cycle starting in apparel and then moving to home furnishings (A Source 1985; Verlodt 1994), or vice versa. The case to be made for the appearance of a color in one category subsequently influencing its appearance in the opposite category was best implied for colors demonstrating much higher incidence in one product category than the other. Little clear or consistent evidence of a time lag in patterns between categories could be observed for half of the colors (yellow, orange, blue, gray and white).
Although we did not seek to identify different colors as being dominant in HF or AP, findings clearly showed that for some of the colors. Blue was the only color with similar incidence in both, suggesting equal importance in HF and AP. Orange and gray had low incidence for both categories. Color preference in both what one wears and one’s home environment could illustrate Nystrom’s (1928) concept of a dominating ideal held by large numbers of people and therefore influencing fashion. Certain colors could be perceived by a majority as what they want to wear on their bodies while others are what they want to view in their spaces. Based on the data (see Table 2), it can be inferred that consumers perceive black and red more as AP colors and brown and white more as HF colors.
Those wishing to understand, interpret, and apply color cycles in apparel and home furnishing products as well as those interested in exploring fashion cycle theory can take something from this research. The implication for the design and selection of products is that the length of cycles is not necessarily the same or similar with different colors or between home furnishings and apparel. Some colors seemed clearly more fashion-driven because they had shorter cycles and could disappear entirely for consecutive years. Other colors seemed to be classics because of their constant or near constant presence. The literature on fashion color cycles was expanded with a clear conclusion that they are not simple phenomena. With different colors having a greater or lesser presence and shorter or longer cycles in apparel and home furnishings, and the study limited to 10 hue families, enough parallel or similar patterns of change were not discerned to support substantive forecasting guides. However, the key findings by color highlighted in Table 2 could be insightful for product developers.
Researchers interested in the crucial component of product color could move ahead by studying the color cycles of, for example, smaller categories of products for the home such as kitchen tools, accessories, and household linens in relation to home furnishings or in comparison to clothing accessories. Although the scholarly literature on fashion cycles is already dominated by women’s dress, a potentially insightful additional study of women’s apparel in relation to color cycles would be to see if Oberascher’s (1994) cycle stages can be observed. As consumers we sense cycles of color availability, so helping forecasters make more informed decisions concerning color is worthy of additional research for the very practical reasons of satisfying consumers, reducing markdowns, and helping businesses’ bottom lines. As mass customization of products becomes more likely (Kim and Johnson 2009), research into how consumers make aesthetic decisions on color when designing products to be customized could also yield insights for product planners. Like most studies before ours, we found cyclical patterns, just not neatly regular ones. Thus we contend that this area of research should garner more empirical examination.