January 18, 2006
Unpublished Data Clarifies SRI Proxy Voting on CSR Resolutions
by Bill Baue
In part two in this multi-part series, SocialFunds.com examines unpublished data associated with a
recent report from The Corporate Library, and finds that statistics are not always what they seem
at first glance.
The Corporate Library (TCL) report on 2004 and
2005 mutual fund proxy voting released last week defies expectation in many ways. With the
spotlight shining on proxy voting, one would anticipate mutual funds voting more independently of
management and thus increasingly supporting shareowner resolutions. The report, which is based on
45 funds' N-PX filings
with the SEC, does indeed document a 1.5 percent
increase in overall support for shareowner resolutions, from 35.8 percent in 2004 to 37.3 percent
(Read all four articles in the series here.)
However, when delving deeper
to distinguish between support for corporate governance (CG) resolutions and those on corporate
social responsibility (CSR), the report surprisingly finds support for the latter decreasing
1.3 percent, from 17.5 percent in 2004 to 16.3 percent in 2005. Even more confounding are the
results that break down between mainstream funds, which support CSR resolutions less than 10
percent of the time, and socially responsible investing (SRI) funds, which support CSR resolutions
about 60 percent of the time, according to the report.
"Mainstream funds' support for CSR
resolutions declined by around 1.4 percent, however SRI funds' support for CSR resolutions declined
by almost 3 percent," writes Jackie Cook, the senior research associate at TCL who conducted the
analysis, in the report she authored. "The decrease in SRI funds' support for CSR resolutions was
caused by decreased support by Catholic
Funds (-14.5 percent), Pax (-13.3
percent), and Parnassus (-7.5 percent)."
The finding that SRI funds decreased support for CSR
resolutions--which are predominantly filed by SRI funds and allied organizations--is quite
counterintuitive, and may result in large part from the inherent limitations in the research
"It should be noted that, since most SRI funds apply screening criteria to
their portfolios, these funds are not voting on a comparable set of resolutions to those of
mainstream funds," Ms. Cook writes in the report. "Rather, they are voting on a subset at
corporations that have already passed the various screening criteria applied by individual funds."
In other words, SRI funds are voting on a much smaller set of resolutions--not only
because of screening, but also due to the simple fact that SRI fund firms tend to be much smaller
than mainstream fund families and hold many fewer companies in their portfolios.
not any conscious decision on our part to reduce voting on CSR," said Jerry Dodson, president of
Parnassus Investments. "It's probably because the statistical sample is so small that a change in
how two or three resolutions are voted would make a major difference."
According to unpublished
data provided to SocialFunds.com by Ms. Cook, Parnassus voted on only 16 resolutions in 2005,
opposing five and abstaining from five. Of the 20 resolutions it voted on in 2004, it opposed
eleven. So the 7.5 percent decrease in Parnassus' support for CSR resolutions hinges on a single
"We don't vote for all the CSR resolutions--we do look at each one carefully, but
sometimes we don't agree with all of them," Mr. Dodson told SocialFunds.com.
vice president of social research at Pax World Funds, concurs.
"If there is a change in
our numbers, it is not due to a change in policy or voting procedure--it's more likely just a
statistical anomaly," Ms. Green told SocialFunds.com.
Of the 13 resolutions it voted on in
2004, Pax opposed only two and abstained on one, according to the data provided by Ms. Cook. Of
the 11 resolutions it voted on in 2005, Pax opposed four--as with Parnassus, the decrease resulted
from a mere vote.
Another somewhat confounding aspect of the report is how "abstain" votes
are counted, which voice neither support ("for") nor opposition ("against").
investigating Investor Responsibility Research Center (IRRC) data published in the Social Investment Forum (SIF) 2003 Trends Report a few years ago, Ms. Cook contacted IRRC to
inquire how it counts abstain votes on individual resolutions to ascertain whether they qualify for
re-submission, and discovered that IRRC leaves them out of the equation. IRRC divides the number
of votes "for" by the sum of the number of votes "for" plus the number of votes "against." IRRC
had consulted with the SEC, which confirmed that this is the correct method for tallying votes to
determine eligibility for re-submission.
For the purposes of the TCL study, which seeks to
provide a comprehensive view of mutual fund proxy voting, Ms. Cook reasoned that including the
"abstain" votes (by dividing the number of votes "for" by the sum of all votes--"for," "against,"
and "abstain") was necessary to account for all votes cast.
"Abstain votes skew the
results either way you count them," Ms. Cook told SocialFunds.com.
votes (as IRRC does) paints an accurate picture of voter will on specific resolutions, as it only
counts clear expressions of support or opposition, but it neglects to account for a portion of the
votes cast--perhaps appropriately, as these voters deferred from expressing a clear will.
Including abstain votes (as the TCL report does) accurately accounts for all votes as is
appropriate for a broad survey of proxy voting across entire mutual fund families. However, one
effect is that this method de facto lowers support by increasing the denominator--despite
the fact that these voters expressed neither support nor opposition to the resolution.
we were to apply the SEC-sanctioned IRRC method to the fund voting data, the effect would
definitely be an increase in support for CSR resolutions by SRI funds," said Ms. Cook.
Catholic Funds and Parnassus, which together account for the lion's share of the decrease in
SRI support for CSR resolutions, also account for the most abstentions on CSR resolutions--Catholic
Funds abstained on 61.4 percent of its CSR resolutions in 2005, and Parnassus on 31.3 percent.
"Our proxy guidelines include an explicit list of issues on which we abstain," Catholic
Funds President Ted Zimmer told SocialFunds.com. "We chose these because we do not claim much
expertise on these issues and/or because it may not be clear how Catholic values influence the
appropriate response to the issues."
Using the IRRC calculation method, SRI fund support
for CSR resolutions decreased a mere 0.5 percent from 2004 to 2005, and increased by 2.6
percent for all shareowner resolutions (as opposed to decreasing 0.4 percent, as the results
that count abstain votes find.)
source: Jackie Cook
Counting abstentions as the SEC instructs results in a
much different picture of SRI fund voting on shareowner resolutions, especially those focused on
Part one of this multi-part series
examines mainstream mutual fund proxy voting results from The Corporate Library report.
Part three of this
multi-part series examines unpublished data and finds decreasing support for climate change
Part four of this multi-part
series steps back to consider the implications of the findings and how to harness them to push
mutual funds to support corporate social responsibility.
|SocialFunds.com thanks Jackie Cook for generously offering access to unpublished
research data and analysis for this and subsequent articles.|