ACCHANGE. Union Bargaining model in ATC. Stef Proost (KULeuven+TML) joint work with Amihai Glazer (UCIrvine) Thomas Blondiau (TML)

Transcription

1 ACCHANGE Union Bargaining model in ATC Stef Proost (KULeuven+TML) joint work with Amihai Glazer (UCIrvine) Thomas Blondiau (TML)

2 Motivation 1 Air traffic control in Europe is presently organized by 37 national or regional Air Traffic Control centers that organize flight routes from airport to airport. There is room for improvement in ATC in Europe: Need for better technologies; lower costs; mergers of too small ATC centers etc. EC and Eurocontrol try to incentivize these ATC s to become more efficient (monitoring and price-cap instead of cost+), push (subsidy and regulation) to implement regional cooperation, better technologies etc. But progress is very slow ACCHANGE project 2

3 Motivation 2 Need to understand the incentives of the ANSP s (air navigation service provider) to know what policies can work 3 different economic approaches: Finetune incentives of the firm (price-cap rather than cost+) Interest group theory (home bias in selecting equipment suppliers, unions etc.) Union Bargaining model Stylised model for only one ANSP (air navigation service provision) organisation ACCHANGE project 3

4 Outline Model set-up Current equilibrium Effects of a price cap Adoption of simple innovations Adoption of drastic innovations Pricing of congestion Empirical estimates of union power Future research ACCHANGE 4

5 Model set-up 1 One ATC zone Service supplied to standardized number of flights D run by competitive airlines Max Willingness To Pay (WTP) for ATC services of flight is p max Technology k=0 requires minimum 1 ATController unit per flight Other costs of ATC = OCC per flight ATC Unions negotiate on wage and employment Labor cost of 1 ATC = w competitive wage w ACCHANGE 5

6 Model set-up 2 Regulator (ATC controlled by government) maximizes Consumer Surplus of Flights minimize costs of ATC ATC Unions negotatiate on wage w and on employment and have utility or objective function (α>0) w = competitive wage per ATC controller unit L = minimum number of controllers per flight for technology k=0 w>w means better salary than market for same qualification L>L means more relaxed work conditions, less working hours per week etc. ACCHANGE 6

7 wl=constant=p*d-occ WAGE For given technology k=0 A w B Solution chosen by union maximizes Utility subject to wl = constant in zone AB L EMPLOYMENT ATC

8 Model set-up 3 Equilibrium of bargaining between ATC union and regulator (or ATC management) is an assymetric Nash bargaining solution Nash bargaining solution= divides total gain as a function of their fallback position and as a function of their bargaining power (strikes, lock-out option,..) Regulator surplus Union surplus ACCHANGE 8

9 Price ATC Wage Surplus Consumers P max P realized= average cost of ANSP OCC w1=(p max OCC-Surplus Consumers)/L w2=(p max OCC-Surplus Consumers)/L2 w L L1 L2 Total Flights, total employed

10 Nash Bargaining solution No union power (δ=1),p*=(p -OCC-w L )/D If Union has all the power ((δ=0), high wages and/or high employment α=1, only high wages count, w is maximized α=0, only high employment counts, w=w,l>l ACCHANGE 10

11 Outline Model set-up Current equilibrium Effects of a price cap Adoption of simple innovations Adoption of drastic innovations Pricing of congestion Empirical estimates of union power Future research ACCHANGE 11

12 Explaining inefficiency in cost+ regime Inefficiency (high w, high L) results from strike power of the ATC unions Assume there are domestic and foreign (or transit) users of air space, domestic regulator is more attentive to Cons Surplus of domestic users Results in more surplus for unions or higher profits for ATC As profits for ATC are ruled out by international conventions, one ends up with high costs and prices ACCHANGE 12

13 Will price-cap improve efficiency? EU regulator wanted more surplus for the transit and domestic users EU is less influenced by national strike actions as there is only one out of 28 that protests things would be different if the ATC unions became more European Price cap forces the national regulator (or ATC manager) to lower total costs (OCC+wL) 3 Possible outcomes: ++ Lower Other Costs and lower wl - -Price cap is not observed (depends on sanction,..) - -National governments start to subsidize the ATC operations when unions have a lot of power (cfr. Public transport subsidies,..) ACCHANGE 13

14 Price Wage Surplus Consumers OCC P max P realized Price cap Average cost subsidy w1=p max OCC-Surplus Consumers w2=(p max OCC-Surplus Consumers)/L2 w L L1 L2 Total Flights, total employed

15 Small innovation or better procurement? Will be accepted by unions as this increases the total surplus to be distributed But unions will be afraid of technologies that cut costs by standardization as this may mean they are more easily replaced by training and hiring abroad (this would reduce their bargaining power (1-δ) Risk of natural coalition of home equipment manufacturers and local unions to put pressure on home ATC regulators ACCHANGE 15

16 Drastic innovation ( virtual center ) A virtual center allows a foreign ATC to take over a large part of the ATC operations Preferable to have at least two virtual center suppliers Not their actual use but the threat of their existence may be sufficient The virtual center supply puts a bound on wl+occ (there is a cheaper alternative) Maximum WTP for local ATC operations decreases from p max D to cost of virtual center VCC and wl+occ has to be cheaper than VCC ACCHANGE 16

17 Price Wage Surplus Consumers P max P realized OCC w1=p max OCC-Surplus Consumers w L L1 L2 Total Flights, total employed

18 Price Wage Surplus Consumers P max P realized OCC PRICE OF VCC w1=p max OCC-Surplus Consumers w L L1 L2 Total Flights, total employed

19 Incentives for ATC to adopt time of day pricing? Two impediments for adopting fine (tuned) charging In a no-union world, one reduces the user cost (including queuing costs) but increases the money cost to the user this may not fit the price cap regulation if it does not correct for quality of service In a unionized ATC world, there are two opposing factors There are more revenues (waiting time is converted into toll revenues) that can be used to increase wages and/or profits When queues go down, one needs less ATControllers ACCHANGE 19

20 Cumulative departures and arrivals Speed-flow Bottleneck model Fixed Demand Variable Demand Fine Toll Uniform Toll Coarse Toll Bottleneck equilibrium without time dependent (peak load) pricing Max queue determines Number of ATCcontrollers D Waiting time Queue length A t t first t t tilde departure time to arrive just in time t* t last Real time

21 Empirical illustration Data for 31 ATC with output, Value added, wages, labor productivity Union power is difficult to measure directly as strikes are not a good indicator, it is the threat that counts Wage mark up (w-w ) can be estimated by comparison with wages in other sectors with same qualification Excessive Employment (L-L ) is estimated by comparing with the most efficient ANSP (Ireland) Maximum wilingness to pay for flight control services is estimated by using the highest price in Europe per composite flight hour (Belgium) Generates relative bargaining power of government and preferences of the unions ACCHANGE 21

22 Bargaining power and union preferences ACCHANGE 22

23 Results Large differences across countries Union power High union bargaining power in Belgium, NL, Spain Low union bargaining power in Sweden, Denmark, Ireland Union power is used differently: High wages in Ireland, Cyprus, Portugal, Spain High employment in Belgium, Germany, France. ACCHANGE 23

24 Future research Paper shows that union power can explain some inefficiencies of the sector Further Empirical validation: Do countries with more union power have less standardization, easier subsidies etc. But what determines union power? Social institutions, rights to strike, substitutability ACCHANGE 24

25 Summing up Powerfull ATC unions is one of the factors behind the inefficiency of the cost+ equilibrium Potential implications: price-cap could lead to state subsidies Home bias in equipment and non-standardized procedures may be preferred by present ATC s because it protects the union power Drastic innovations (virtual centers) may be efficient to limit ATC unions power Unions may not be in favor of time differentiated charges as this will reduce the number of controllers needed ACCHANGE 25