Back in 1991, when the Internet began to emerge properly, there was a large quantity of computer software available on FTP servers, but few documents. Why? If the academic networks could carry free software, why not free research documents? The economics seemed nonsensical. Why would academics work to produce research without being paid, submit that research to a publisher (often involving payment of a submission fee), referee the work of other academics for free and pay a heavy price to buy the output back from the publishers?
In Economics a printed free publishing system has existed over a long period of time in the form of working papers - early accounts of research findings. They are usually published by the department with whom their author is affiliated. Many university Economics departments have series of working papers which are circulated to other universities on an exchange basis, or at a nominal charge. Wider use of working papers has however been hampered by problems of current awareness and the handling costs associated with holding stocks of papers.
To realise the vision of a free electronic publishing environment in Economics, we need to take the existing free publication system off the paper cart and put it into an electronic Formula-1 car. There are many reasons why an electronic working paper collection is superior. First, it is less costly to its publisher, because the author supplies the first copy at no cost and the marginal cost of an additional copy is close to zero. Second, if a good current awareness database system for working papers is available it can break the 'insider problem' of papers only circulating among a small group of researchers who thereby gain an advantage. If the insider circle is based in the United States (as it often is) British research is at a considerable disadvantage. Third, it allows poor institutions with Internet access to obtain research results at low cost.
WoPEc (Working Papers in Economics)  was the first collection of electronic working papers to open in Economics. It received its first contribution in April 1993, and continues to store working papers, though its main activity now is to collect metadata on electronic working papers stored at other institutions.
Collecting bibliographic references to documents rather than collecting the documents themselves may seem odd at first, since most pre-print services hold documents and metadata on the same site. A central Economics pre-print archive was indeed opened in the US in 1993, the Electronic Working Paper Archive, based upon the Ginsparg software used in the Los Alamos National Laboratory Electronic Preprint Archive . It has not been overwhelmingly successful. Other 'local' archives continue to proliferate. These may be held on institutional or departmental servers, or even collections of documents in an individual's Web home page. Our collection of bibliographic information from these sites now runs to well in excess of 3000 documents from around 300 series. The exposure that papers on the system enjoy is quite impressive. WoPEc ran unfunded from April 1993 to August 1996, and was developed by the author. In the second year the project gained José Manuel Barrueco Cruz of Universitat de València. ELib funding was awarded in May 1996. A whois++ based search interface has now been implemented.