How to deal with obtaining optimal prices of services Исполнитель

How to deal with obtaining optimal prices of services

One of the main successes of the internet market is to achieve optimal prices which are profitable for both companies and also internet users.

So far I have described the imminent need for introducing a traffic-sensitive pricing scheme atall levels of Internet connectivity. What needs now to be specified is the type of trafficsensitivescheme that should be implemented. This section attempts to clarify this aspect.

An optimal pricing mechanism should maximize the number of voluntary market exchangesbetween suppliers and demanders, because such exchanges allow individuals to maximize theirindividual preferences. An optimum situation (Pareto optimum) is one where individualshave maximized those voluntary exchanges by pricing goods and services at their marginalcost.

Hence, an optimal pricing scheme should target both demand and supply of the Internetmarket. On the supply side, prices should compensate suppliers (ISPs) of scarce Internetresources. On the demand side, prices should allocate resources efficiently by considering thevalue each user places on Internet demand: Sending signals to originators of Internet demand(end-users and institutions) and to intermediaries of IP transport (ISPs), so they can makeinformed decisions about Internet demand. I will focus mainly on the effect of pricing on thedemand side. A multi-part tariff pricing scheme, can help attain both of these goals.

Economic theory suggests that demand for access should be priced both with fixed recurringand non-recurring fees to recover costs and that demand for usage is charged with anappropriate traffic-sensitive scheme.

The optimal pricing tariff should be a multi-part tariff: A flat fee and a varying fee. The flatfee should be directed to cover costs such as monthly customer support, equipmentmaintenance, billing and accounting. The charging price for the flat-fee portion of a tariffshould vary depending on factors such as, the maximum bandwidth (capacity) of theconnection and the air mileage from user’s home to the ISP’s network. The variable portion should be targeted to recover usage costs of the network and should vary depending on:

Current network load; rate of transfer requested, and the volume of the transfersession (number of packets transferred—sent/received). The expensive investments innetwork infrastructure comprise the majority of the costs of the Internet service provision.

When the network is not congested the marginal cost of transporting an extra packet isessentially zero. When congestion occurs, traffic-based pricing should be used and applieddepending on the three factors just mentioned, and if it is heavy-bandwidth traffic that isrequested then a higher price per unit should be charged than when the request is for lightbandwidthtraffic. In this way, only the users that request service during congestion timeswill pay to expand capacity. Those who are willing to wait will be rewarded with lowerprices or no charge at all for usage.

A pricing mechanism with these features would also solve the problem of worrying abouttime-of-day pricing, as the proposed tariff would depend on the current load of the network.

Notice that this tariff could be applied to end-users and institutions as well as to the differenttypes of ISPs. The latter are just intermediaries in the transfer of data from one end-user toanother. In other words, just as in other industries, corporation’s customers may face higherrates than residential users.It must also be observed that while the goal of any Internet pricing mechanism should be tooptimize the number of economically efficient connections, it is important to distinguish itfrom the goal of Universal Service. Though highly unlikely, it may be possible to have fullInternet connectivity for all users. However, this goal would not be economically efficient, asthere are clear substitutes for Internet communications that may be more efficient for someusers. For example, it may be more efficient for an individual to make a telephone call, send afax, or go to the library than to use the Internet for equal purposes.

The issues of economic efficiency versus equity have been endlessly debated for otherindustries too, such as, telephony, cable, electricity, and water. As it is the case in theseindustries, no clear answer has been identified to solve the efficiency versus equity trade-offon the Internet either. What it is clear, however, is that Universal Service for the Internet isrelated to the way income is distributed in society rather than to how the Internet is priced.

I have suggested that the Internet is a value-added network and service. As it resembles thetelephony network in some way, the latter may be used as a guide to understand theInternet’s network. I have also identified the type of costs that an ISP (Internet provider)incurs in such a provision. The different types of costs to be considered for Internet pricingare:

Access costs:

· Access costs are the fixed costs of providing the network infrastructure(installation costs) and the fixed costs of connecting a customer to the ISPnetwork (activation costs). The former include ISP’s payment for the leased line,cost of the routers. The latter are costs incurred by the customer when connectingto the network. Apart from his computer equipment, a customer will need to payfor obtaining Internet connectivity to the ISP network.

  Usage costs:

· There are two types of usage costs: Network load costs and maintenance costs.While maintenance costs are period fixed charged, network load costs shoulddepend on the current status of the network—congestion.

· The marginal costs of sending an extra packet. This cost is zero when the networkis not highly congested.

· The social costs of delaying other users’ packets when the network is congested.

This cost is translated through delay and dropped packets.

· The marginal costs of connecting users to the network. Each new subscriberimplies costs for access lines. The costs are fixed only when excess networkcapacity (not necessarily bandwidth) is present.

· The costs to expand network capacity: Adding new routers, more bandwidth,and new staff. These costs vary in a lumpy manner—with the increase in number ofusers connected a larger network is needed.

ISPs run a network service that connects users to the Internet. In the short-run, costs in anISP network are fixed except for congestion cost. That is, when a user on the Internet sendsout a file, he is using up bandwidth that could otherwise be used by other users in thenetwork. All the costs associated with number of customers, electronics, leased lines, etc.have been paid for already and are seen as fixed costs. In the long-run, however, all costsvary. The ISP will need to invest in a larger network as the number of customer increases andas more usage (from the highly appealing intensive-bandwidth applications) is demanded.

There are two basic types of costs that ISPs need to recover: Access costs (recurring andnon-recurring) and usage costs. Internet access costs should account for two factors: Thesetting up of the ISP network and the costs of connecting the customer to the network. Nonrecurring-access costs refer to those of recovering the investments of building the network.They should be charged to the customer when he connects to the ISP. Recurring access costs(or customer-fixed costs) do not vary with usage and refer to those of equipmentmaintenance, customer support, leases, monthly billing and accounting, etc. Usage costsshould vary with the customer’s usage and the current state of network congestion. Theyshould reflect the cost each end-user imposes on the network and on other users when hedemands service.

I have also stated that the main reason for some sort of a traffic-based pricing is to controlcongestion, as current pricing mechanisms are not efficient in such a practice. Controllingcongestion is crucial, because it is the most important economic problem of today’s Internet.Based on these grounds, I have proposed an optimal pricing mechanism. The first and mostimportant characteristic of this pricing scheme is that is traffic based. It should have a fixedportion varying positively with the maximum bandwidth of a connection and a variableportion varying positively with: (1) the user’s rate of transfer; (2) the current load on the ISPnetwork; and (3) with the size of the transfer request. I have also shown the twofundamental elements needed for the functioning of the rate-of-transfer pricing: A toSsensitive protocol and a computer interface allowing the involvement of the user in thedecision making process. Furthermore, I have not only illustrated the inefficiencies createdwhen prices are not optimal but also explored a solution for when optimal prices cannotrecover costs of provision.