Here is an end-of-summer thought for you: ''The ability to vote on public policy while snuggled under an electric blanket or munching on corn chips does not demean the system. On the contrary, the system can be enhanced and the American tradition honored.''

The source of that remarkable thought is Richard S. Hollander, a Baltimore television broadcaster and author of the newly published ''Video Democracy'' (Lomond Publications, Inc.). He foresees -- and embraces -- a future in which local governments have been abolished. Citizens use two-way communication channels, through cable television linked to computers in their homes, to rule themselves.

In Mediaville (his imaginary, renamed Middletown), the city council was abolished and its ''chamber converted to a TV studio . . . when the entire town had been wired for two-way cable. Each home had a computer which interacted with the cable TV system.''

And in Hollander's example, Joe Citizen, bored with TV baseball and too broke for video gambling, punches the ''public and civic affairs'' button on his cable TV and finds the director of public works on his screen. On impulse, Joe walks to his computer and messages in a motion to require that all the potholes on Main Street be filled within 30 days. The proposal is put up for debate and quickly brought to a vote. Despite the grumbling of the department head, the watching citizens approve the instant referendum, 1,567 to 985.

Arguing that the needed technology is available, Hollander says the concept ''is utterly practical and eminently democratic.'' State and national governments are still needed to deal with complex policy choices, he concedes, but small- and medium-size cities can adopt ''direct democracy'' and thereby cure official corruption, end public cynicism and gain ''better, more efficient local government.''

Well, it's a wonderful dream, but, as always, there's someone around to wake you to reality. The skeptic in this case is F. Christopher Arterton, dean of the Graduate School of Political Management in New York. His book, ''Teledemocracy: Can Technology Protect Democracy?,'' was published last month for the Roosevelt Center for American Policy Studies by the Sage Library of Social Research.

He studied 13 local experiments in the use of interactive communications technology (including all the examples cited by Hollander) and came to a far more cautious conclusion: the new technology can improve citizen access to decision-makers and broaden participation in public-policy decisions. But it cannot bypass government to achieve direct democracy, Arterton says.

The experiments, held in various locations from Hawaii and Alaska to Reading, Pa., were interesting and, on their own terms, heartening. Whether they took the form of electronic town meetings, newspaper-ballot referenda after televised discussions of policy problems, teleconferences of legislative hearings or regional forums on health policy and planning issues, these ventures broadened participation, helped spread information and gave government officials a clearer sense of public attitudes.

But Arterton finds two big problems with all of them. It takes a lot of work (and cost) to bring government decision-making into living rooms or scattered meeting halls through cable TV or leased phone lines. The people who pay the costs or make the arrangements naturally want to have a big say in setting the agenda. Net result: what purports to be an exercise in democracy ends up being an operation manipulated by some elite.

''All of these,'' he notes, ''are political . . . problems of plebiscitary teledemocracy, but I found no instances that offered hope that they could be mitigated by technology.''

The second problem is apathy. A lot of us simply don't want to shoulder the daily burdens of making public policy from our homes. Reviewing the local experiments, Arterton says, ''I found little support for the notion that citizens have the interest necessary to sustain near universal participation; in practice, too few are interested enough in politics to make plebiscites a feasible means of policy-making.. . . Most citizens, probably around two-thirds, will not participate.''

Hollander acknowledges the problem. ''The bottom-line question in a direct democracy concerns participation,'' he says. ''Obviously, government cannot be effectively run by a small cadre of dedicated activists. It demands mass popular support. In many communities, citizen participation in public affairs is an embarrassment.''

He holds out the hope ''that people do care and that they want to participate in decision-making that affects their lives.'' But in the mythical Mediaville referendum, he has only 2,552 of the city's 50,000 residents voting. Some participatory democracy. Where were all the others? Probably snuggling under their electric blankets or munching corn chips.