Mee-Eun Kang and I have a chapter about push polls in a new edited volume
that Paul Lavrakas and I have coming out in a few weeks.  It is titled
"Election Polls, the News Media, and Democracy" and will be published by
Chatham House.  I will produce a couple of paragraphs ...

What some consider to be a true "push poll" is run out of a phone bank.
Thousands of calls are made, and no data are usually recorded.  This is
"negative persuasion calling" rather than a poll.  This is the kind of "push
poll" that was the focus of the joint statement by AAPOR, NCPP, and the
American Association of Political Consultants.  In a poll that is used to
evaluate strategies that might work in a campaign, several positive and
negative themes might be evaluated.  But the company/consultant is
interested in collecting and analyzing the data to see what works and how.
It is worth noting that you live in NY and the data were being collected on
the New York race.

It is difficult for respondents to understand the difference between the two
techniques, of course, and the resulting negative experience can have a
detrimental consequence for all who conduct polls.

Here's the text from the introduction:
Push polling is a relatively new kind of campaign technique that is designed
to move the support of voters away from one candidate and toward another.
It has been adopted by candidates, political parties supporting a candidate,
and organized interest groups supporting a candidate or an issue.  Initially
developed and employed with some success in presidential campaigns,
especially in both the 1996 primaries and general elections, it has
increasingly been used in contests  for smaller constituencies and for many
different kinds of contests, now including referenda and initiatives.  The
technique has raised alarms among advocates of good government and fair
campaign practices as well as in the polling and survey research industry.
Push polls simulate an interview on the telephone, but they often do not
involve data collection or analysis.  As a result, they have been labeled
“pseudo polls” (Traugott and Lavrakas, 1996).  The form of questioning can
offend people who are subjected to it, and the fear of the polling business
is that the technique will contribute further to already declining response
rates and public trust in polls.

Many state legislatures have responded to the rise of push polling
by drafting legislation to outlaw it, and a similar bill was introduced in
the U.S. House of Representatives in 1997.  Such legislation has proved
problematical because many of these laws fly in the face of protected forms
of political speech under the First Amendment to the U.S. Constitution.
The key issue for legislators is defining an unacceptable practice with
sufficient precision that the proposed "illegal" behavior does not include
protected speech.  In this chapter, we review the rise of push polling,
paying attention to the distinctions between “negative persuasion
telephoning” and strategic polling designed to assess the potential
effectiveness of alternative campaign themes. We employ a systematic search
of reported occurrences of push polls in the last few election cycles in
order to develop a conceptual framework that describes who is using them and
under what electoral circumstances.  We then review current attempts at the
development of legislation to regulate the technique, with an emphasis on
the level of specificity and targeting of unethical practices.