The Tax Reform Act of 1986 reduced individual income tax rates below t
hat of corporations. The fear that this would lead to a systematic dis
incorporation has not apparently materialized since the major stock ex
changes report that only about 100 partnerships were traded during the
next five years.1 Scholes and Wolfson (1992) suggest that this result
is predictable because of the additional nontax costs of operating as
a partnership, which include increased transactions costs, more restr
icted access to capital markets, and less control over management. Gue
nther (1992) and Terando and Omer (1992) provide evidence that firms m
ust have considered both tax and nontax costs when choosing organizati
onal form. This study examines whether the factors taken into consider
ation in organizational form choice also affected the market value of
a sample of firms in the oil and gas industry during the period 1985-1
988. We chose for our tests a valuation model used by Harris and Ohlso
n (1987) and others2 since many of the publicly traded (master limited
) partnerships (MLPs) were created in the oil industry. The studies by
Guenther (1992) and Terando and Omer (1992) show that MLPs generally
had only one line of business, less debt, and higher dividend yields t
han their corporate counterparts. These differences are predictable in
view of the tax consequences of operating in the partnership form. Gu
enther also finds MLPs to be less profitable, which is consistent with
higher nontax costs of the partnership form. We show that MLPs invest
ed significantly less in exploration for new deposits. The combination
of the lower exploration expenditures, higher dividend yields, and po
or financial performance suggests that MLPs may have been set up as li
mited-life entities to distribute assets to their unit holders in a ta
x-efficient manner.3 We extend the Harris and Ohlson (1987) valuation
model by adding dividends and exploration levels. Exploration levels a
re found significant for both MLPs and corporations, but dividends are
relevant only in the MLP model. We also find that dividend levels are
significantly explained by asset values only for the MLPs. These valu
ation differences are consistent with tax-motivated organizational for
m choice and the perceived passive nature of the MLPs.