Bilateral Implicit Taxes and

1
Bilateral Implicit Taxes and Anti-Competitive
Banking Regulation David G. Harris Syracuse
University Lubin School of Accounting Whitman
School of Management Syracuse, NY 13244 (315)
443-3362 Emre Kilic University of Houston C.T.
Bauer College of Business Houston, TX
2

• I. Conceptual framework, institutional setting,
  and hypotheses
• A. The party with greater contracting power
  bears less of a tax increase
• 1. Harberger (1962) demonstrates that
  elasticity determines tax cost sharing
• 2. Generally, elasticity of supply and demand,
  information asymmetry and oligopoly drive
  contracting power
• B. Banking law substantially restricted banks'
  operations to a single state until after 1993
• This could have increased banks contracting
  power by reducing the supply of banks
• C. The mortgage market being inherently local
  suggests banks have greater contracting power
• 1. Buyers are geographically restricted
• 2. Banks have wide investment opportunity set
  other personal loans, business loans,
  investment assets, etc.
• 3. Local knowledge is especially valuable,
  e.g., pricing of comparable properties,
  planned development, state and local laws and
  regulations, school districts, etc.

3

• D. Home mortgage interest rates and taxes,
• 1. Banks pay taxes, which they shift onto
  borrowers to the extent possible
• 2. Mortgagors deduct mortgage interest, which
  banks capture to the extent possible
• 3. If regulation increased banks' contracting
  power, its repeal reduced tax- shifting
• E. General Economic findings on tax shifting and
  competitive markets
• 1. In non-competitive markets firms pass taxes
  onto customers
• Some research suggests more than a 100 shift
• 2. Consistent with general economic notion of
  firms earning only economic rents in fully
  competitive markets, i.e. no shifting.

4

• Hypotheses
• 1. Mortgagors bear implicit tax costs from
  banks taxes.
• 2. Mortgagors bear implicit tax costs from their
  mortgage interest tax deductions.
• 3. Relaxing state regulation of interstate
  banking increases banking industry
  competitiveness.
• 4. Mortgagors implicit tax costs decline with
  banking industry competitiveness.

5

• II. Findings summary, the results are consistent
  with
• A. Banks shifting their taxes onto mortgagors,
  increasing mortgage rates by 10 basis points
• B. Banks capturing the value of mortgagors
  interest tax deductions, 11 basis points
• C. After interstate banking was allowed, a
  decline of 8 basis points, or 40, in implicit
  taxes
• D. Economic intuition
• 1. Dollar cost for a typical loan is small
• a. For a 30-year, 100,000, 6 mortgage, only
  13 per month
• b. Only 4,651 over the life of the loan
• 2. Market-wide, dollar amounts are not small
• a. 11 trillion market
• b. Implicit taxes of 21.7 billion per year
• c. About 30 of the value of homeowners
  mortgage interest deduction

6

• III. Methodology
• A. Data
• 1. Detailed, bank-specific, financial data from
  1977-2004
• a. Proxy for bank tax rates with the median,
  single-state bank, by year, by state
• b. Both State and Federal are considered
• 2. Detailed, individual-loan-level mortgage
  origination data
• a. Loan-specific control variables
• b. Loan-specific interest rates
• c. Take the median interest rate loan, by
  state, by year
• 3. Estimates of the average marginal value of
  mortgage interest deduction
• a. NBER tax calculator estimates this by
  state, by year
• b. Calculations based on actual tax-return
  numbers
• c. Both State and Federal are considered
• 4. State macro economic data are used as
  control variables

7

• B. Analyses
• 1. We estimate equation (1)
• 2. Where, all variables are taken as their
  annual first-differences, and
• a. For the median loan in state i in year t
• (1) RATEit is loan contract interest rate
• (2) FEESit is initial fees and charges as a
  percentage of loan amount
• (3) SIZEit is natural logarithm of loan size
• (4) TERMit is natural logarithm of loan
  maturity in years
• (5) LPRit is loan-to-price ratio.
• b. For state i in year t
• (1) UNEMPit is average unemployment rate
• (2) GDPit is natural logarithm of gross
  domestic product per capita

8

• c. TBONDit is yield on a treasury bond with the
  same maturity as the median loan
• d. For commercial banks in state i in year t
• (1) BTAXit is the median ratio of the sum of
  federal, state and local taxes to income
  before taxes
• e. For individuals in state i in year t
• (1) ITAXit is average marginal combined state
  and federal mortgage interest tax deduction.

9

• 3. We estimate equation (2)
• 4. Where the variables are as defined above, and
  D equals 1 in 1994 and after, and 0 otherwise.
• 7. Equations (1) and (2) are also estimated with
  loan specific variables for the maximum loan in
  each state-year, and with similarly determined
  tax variables.

10

• Robustness Tests
• A. Full, 4,851, 655, sample of loans. Find
  strongly confirming results. But, cannot do
  first-differences.
• B. Combined median and maximum samples, with
  dummies on maximums. Also, confirming
• C. Vary the median loan by three observations
  plus or minus
• D. Select random time periods to test for
  coefficient variations.
• E. Also control for lender ROA
• F. We use publicly available, aggregated, loan
  data, and find similar results, though State and
  Federal cannot be separated. Control variables
  are inconsistent because of inability to match
  controls and interest rates with averages.

11
(No Transcript)
12
(No Transcript)
13
(No Transcript)
14
(No Transcript)
15
(No Transcript)
16
(No Transcript)
17
(No Transcript)
18
(No Transcript)
19
(No Transcript)