The factors affecting people's willingness in participating in and paying for upgrading their seats from the economy class to the premium class

The study investigates whether and how passengers in Taiwan would participate in airline seat upgrade bidding and how much they would pay, motivated by airlines’ need to monetize unsold premium inventory and passengers’ desire for better service at discounted prices. The context is increasing competitive pressure and low margins in the airline industry, alongside the emergence of upgrade bidding programs worldwide. Taiwan had not implemented such programs at the time of study, creating a need to understand potential passenger behavior. The research aims to identify key determinants of participation and WTP, assess differences between long- and short-haul scenarios, and provide actionable guidance for airlines’ marketing and pricing strategies.

Seat bidding has evolved since JetBlue’s early adoption (2008), with platforms like Plusgrade enabling over 50 airlines (e.g., Air Canada, Air China, Cathay Pacific, Etihad, Lufthansa, Singapore Airlines) to monetize upgrades through mostly closed, one-time bid systems; some open and multi-bid systems (e.g., Virgin America, SeatBoost) also exist, typically with restrictions (e.g., no group or concession fares). Upgrade bid methods vary by route, season, and airline, with one-time bids most common. The double-hurdle model literature shows broad application to consumer demand and WTP problems: cigarette and food consumption, fuel and energy choices, crop insurance WTP, parking, and transport services. Prior work demonstrates its suitability to separately model participation and expenditure decisions, justifying its use for seat bidding participation and WTP analysis.

Design: A stated-preference, face-to-face survey was conducted at Taiwan Taoyuan International Airport’s immigration reception hall on Aug 14–15, 2019, using convenience sampling. Investigators were trained and instructed to maintain neutrality. A total of 400 questionnaires were collected; all were valid (100% valid rate). Sample adequacy was checked with Cochran’s formula (n≈385 at 95% CL, ±5% precision), indicating sufficiency. Questionnaire: Three parts captured (1) socio-economic and travel characteristics (12 variables including gender, age, education, marital status, occupation, income, recent flight frequency, purpose, typical cabin, loyalty membership), (2) experience/awareness of seat bidding, and (3) WTP for upgrades under two hypothetical scenarios: long-haul (Taipei–Los Angeles, upgrade from economy to premium economy) and short-haul (Taipei–Tokyo, upgrade from economy to business). Respondents were briefed on seat bidding operations before answering. A one-time bid framework was assumed (no real-time competitive feedback), and respondents considered only their own participation. WTP Elicitation: A triple-bounded contingent valuation method (CVM) was used. For short-haul, the starting bid was set at 75% of the reference price difference between business and economy; respondents answered a sequence across three layers and then provided a final WTP to mitigate anchoring. Reference price differences noted: short-haul business–economy about NTD 7000 (also noted 5200 in a separate note), long-haul premium economy–economy about NTD 12,500 (also 9500 in a separate note); exchange rate 1 USD ≈ 30 NTD. Modeling: The double-hurdle model (Cragg, 1971) was employed to jointly model (1) participation (Probit) and (2) positive WTP (Truncated) decisions, allowing different determinants and addressing zero outcomes. Likelihood, elasticities, and marginal effects were derived per Gao et al. (1995), Yen and Su (1995), and Amemiya (1985). Comparative estimations included two-level Probit and Truncated models for WTP conditional on participation. Explanatory variables encompassed socio-demographics, travel behavior, prior awareness/experience with seat bidding, and interaction terms (e.g., purpose × income).

Sample profile: 46% male; 55% aged 18–34; 68% university educated; 56% married; most in service industry; individual monthly income commonly NTD 23,001–50,000; 47% flew 3–4 times over the past 3 years; 76% flew for sightseeing; 91% usually flew economy; 18% were airline loyalty members. Awareness was very low: 89.75% had never heard of seat bidding; 0.75% had ever participated (3 persons). Factor importance ratings (1–5): bid price (mean 4.30, rank 1), flight distance (4.24, rank 2), purpose (3.07), accompanying number (2.98). Determinants (Long-haul): Higher participation for respondents under 45; higher household income above average; those usually flying economy or premium economy; and those having heard of seat bidding. Tourism purpose combined with higher individual income increased participation and WTP. Prior premium economy travelers showed the strongest effects (participation and WTP). Some respondents indicated they would directly purchase business class on long-haul rather than bid, and higher starting price differences dampened some participation. Determinants (Short-haul): Awareness (having heard of bidding) significantly increased participation and WTP; respondents under 45 with higher education were more willing to participate and pay; flight distance was cited as important. WTP amounts: Estimated WTP for upgrade price differences were NTD 6917 for long-haul (≈55% of the reference NTD 12,500) and NTD 1684 for short-haul (≈24% of the reference NTD 7000). These align with anecdotal ranges (20–40%) for short-haul; long-haul WTP was higher than that range. Marginal effects (examples): Long-haul—under 45: +1.26% participation, +1.99% expenditure probability, +NTD 564 WTP; higher-than-average household income: +1.94%, +4.16%, +NTD 1175; usually flew economy: +2.98%, +6.80%, +NTD 1921; usually flew premium economy: +4.44%, +19.24%, +NTD 5439; having heard of bidding: +1.91%, +4.94%, +NTD 1397. Short-haul—having heard of bidding: +0.03%, +0.09%, +NTD 547; under 45 and highly educated: +0.04%, +0.08%, +NTD 487; flight distance important: +0.04%, +0.05%, +NTD 275. Overall: Flight distance matters; awareness strongly boosts willingness; younger and higher-income travelers are more receptive; prior experience with premium economy is a strong predictor of higher WTP, especially for long-haul upgrades.

The study addresses whether Taiwanese passengers would engage in seat upgrade bidding and what they would pay. Results show that willingness is segment-specific: younger passengers and those familiar with seat bidding are more likely to participate and to bid higher amounts. Flight distance shapes both participation and WTP, with long-haul contexts eliciting higher WTP shares of the reference price difference (perceived value of comfort over long durations), while short-haul scenarios benefit from lower entry thresholds and simpler decisions. Prior premium economy flyers show particularly strong marginal increases in both participation and WTP for long-haul, implying airlines should target them for business-class upgrade bidding. Awareness has sizeable monetary effects, suggesting education and marketing campaigns could significantly shift uptake and revenue. The quantified WTP levels (55% long-haul; 24% short-haul) provide practical guidance for setting bid floors, default ranges, and targeted offers, and corroborate the notion that optimal pricing should differ by route length and passenger segment.

This paper applies a double-hurdle framework to quantify determinants of participation and WTP in airline seat upgrade bidding among Taiwanese passengers, across short- and long-haul scenarios. Key contributions include: (1) identifying core predictors (younger age, prior awareness, higher income, prior premium economy use, and tourism purpose) and demonstrating that flight distance is a significant moderator; (2) providing WTP benchmarks—NTD 6917 for long-haul (≈55% of reference difference) and NTD 1684 for short-haul (≈24%); (3) delivering marginal effects quantifying how awareness and traveler characteristics shift participation and spending. For practice, airlines should educate the market about bidding, target younger travelers and frequent premium economy users (especially on long-haul), and differentiate bid ranges by distance. Future research should leverage revealed-preference data once such programs launch in Taiwan and examine seasonality, membership tiers, party size, and dynamic multi-bid behaviors.

The study uses a stated-preference, cross-sectional, convenience sample from a single airport over two days (Aug 2019), which may limit external validity and introduce selection bias. Awareness of bidding was very low, possibly affecting hypothetical responses. The survey implemented a one-time bid structure and could not capture dynamic competitive bidding behavior. Seasonal effects (peak vs. off-peak, holidays), loyalty status effects, and group travel dynamics were not analyzed. Scenario price references may vary in practice, and some inconsistencies in reference differences are noted. Findings precede actual implementation in Taiwan, so revealed-preference validation is needed.