Title: How Does FDI React to Corporate Taxation? Old and new results
1How Does FDI React to Corporate Taxation?Old
and new results
- Conference on Tax competition in the EU25
Prague, December 16-17, 2004 - Agnès Bénassy-Quéré Lionel Fontagné and Amina
Lahrèche-Révil - CEPII
2Factor mobility and tax competition
- Traditional tax competition literature
- Increasing integration ? pressure on tax policies
- Small countries more prone to tax competition
- Race to the bottom ? EU enlargement context
- Imperfect competition
- Trade costs scale economies ? home bias, higher
taxes in the largest countries (Haufler Wooton,
1999) - Agglomeration economies ? agglomeration rents,
tax competition limit pricing (Baldwin
Krugman, 2004)
3Evidence on taxes and FDI
- Why shouldnt FDI react to taxes?
- Transfer pricing and intra-firm debt ? profit ?
location - Tiebout (1956) taxation and public-goods
provision - Markusen (1995) structural determinants gt
taxation - High tax high pre-tax return
- Imperfect competition taxes location rents
- Empirically MNF do react to tax incentives
- Double-taxation rules
- repatriated profits importance of tax treaties
4Focus
- ? Does FDI really react to tax incentives (WP1)?
- ? How does it react exactly (WP2) ?
- ? What happens in the enlarged EU (WP 3)
- Look at
- Tax differentials (not tax levels)
- correct for double taxation rules
- Positive / negative tax differentials
- nonlinearities
5FDI and tax variables
- Sample
- 11 OECD countries (9 EU countries Japan US) /
25 EU countries - 1984-2000 (2002 for EU) ? panel data analysis
- Constant FDI in dollars
- Tax variables
- Statutory tax rates
- Apparent tax rates receipts/taxable income
- Ex-ante measures (effective average/marginal tax
rates)
6Gravity variables
- Market potential (? FDI) ( enlarged MP)
- Size of the investing country (PPP) (? FDI)
- Distance (?)
- Fixed costs lt trade costs ? locate close to
markets, and FDI increase with distance (FDI
substitute to trade) - Fixed costs transaction and information costs ?
detrimental to FDI and trade - FDI complementary to trade ? distance detrimental
to both - Common language (? FDI)
7Additional variables
- Relative unit labor costs
- Bilateral real exchange rates (in level)
- Public expenses in the host country
- Share of public investment in total public
expenses - Fixed effects on investors and hosts
8Baseline estimates - comments
- ? market potential attracts FDI (demand effect)
- ? increases outward FDI (supply effect)
- Distance not significant / common language
significant - Higher taxes in the recipient country always
discourage FDI significantly - Tax differentials can compensate for location
- 10 disadvantage in market potential ? 5 points
lower statutory tax rate
9Baseline estimate - Results
10Checking robustness
- Potential endogeneity of effective taxation to
FDI - Apparent tax differentials instrumented using
their lagged values or statutory tax
differential. Estimates robust - Cost variables wrongly signed (appreciation /
increase in ULC ? increases FDI) - Balassa-Samuelson (RER / GDP) quality of labor
- Public expenses
- No impact of volume higher share of investment
attracts FDI taxation has less impact (higher
tax rates compensated for by public-goods
attractiveness?)
11Taxation in third countries
- MNF arbitrate between several potential locations
? the whole vector of tax differentials should
matter - weighted average of taxes in alternative
locations (using the inverse of distance) - Tax differentials in EATR not significant, but
potential collinearity with bilateral tax
differentials - Results much more significant using ex-post
taxation - Higher taxes in the host/investor discourages
FDI, but higher taxes in third countries increase
FDI to the host
12Non-linear impact of tax differentials
- Do double-taxation rules have an impact on the
sensitiveness of FDI to taxation? - Hypothesis higher impact when the exporter
applies exemption - Excess foreign tax credit not reimbursed
- ? only higher taxes abroad should discourage FDI
stemming from crediting countries - Costly relocation tax uncertainty tax
deferrals - ? Higher sensitivity to large tax differentials
13Tax schemes
- Hypothesis not confirmed
- Same impact for exemption or credit
- Hines Rice (1994) same results for investors in
the US - A possible explanation lower tax rates in
crediting countries (except Japan) - MNF mostly feel a disincentive to invest abroad
14Negative vs. positive tax differentials (1)
- Coefficient on positive tax differentials larger
- A higher tax rate in the host is more harmful to
inward FDI that a lower tax rate is attractive
for FDI - ? FDI more sensitive to tax disincentives
15Negative vs. positive tax differentials (2)
- Firms form crediting countries only react to
positive tax differentials - But same asymmetric behavior for firms
originating from exempting countries - ? No major risk of tax competition FDI does not
react to lower taxes abroad. But strong incentive
for high-tax countries to lower the tax burden to
attract FDI
16Non-linearities (1)
- Sources of non-linearities
- Complexity and instability of tax codes ?
information about taxation highly imperfect - Costly relocation
- Tax deferrals, partial tax avoidance
- ? Large differentials should matter more
- 3-steps estimations
17Non-linearities (2)
18Non-linearities (3)
- Tax differentials and cubic tax differentials
negative - Impact stronger with larger tax differentials
- Investors react linearly when exemption react to
large tax differentials when credit - Only positive tax differentials matter
- ? Tax disincentives more powerful than tax
incentives - ? Convergence in tax rates when investors in
exemption
19Quantification
- Market potential versus tax differential
- Credit vs. exemption
- 1 sd increase in market potential ? 31 increase
in FDI inflows - investor in credit ? 7 points tax rent for
recipient - Investor in exemption ? 10 points tax rent for
recipient - Impact of nonlinearities
- Investor in exemption ? 7 points tax rent for
recipient - investor in credit ? 24 points tax rent for
recipient
20Conclusion
- NEG tax competition ? zero taxation (taxable
rents) - Our findings FDI does react to tax differentials
- Conclusion
- No zero taxation structural determinants matter
incentive to cut taxes falls on high-tax
countries - Asymmetry between exempting/crediting countries
incentive for tax competition depends on the
geographical composition of FDI - EMU mostly exemption rules
- ? convergence of tax rates to the lowest level?
21Tax competition in an enlarged EU
- Similar Focus
- Tax differentials
- Positive versus negative tax differentials
- Third countries?
- Sample heterogeneity? ? EU15 versus NEM
22Stylized facts
- Decreasing corporate taxation in the EU25,
especially for statutory tax rates
23Estimation strategy
- Bilateral, gravitational setting, market
potential - 1990-2002, annual
- Tax measurement statutory ex-post taxation
(GDP/VA) - Results
- Gravity significant, distance lt 0
- Statutory not significant, ex-post significant
and lt 0
24Cost and competitiveness (1)
- ULC differential /or bilateral RER
- ?7 positive ? higher costs attract more FDI
(labor quality?) - ?8 positive ? improved competitiveness attract
FDI - ?7 ?8 ? no sign change
- Tax variables
- not very robust
- Tax lt 0 with competitiveness
25Cost and competitiveness (2)
- Geographic dummy
- When d ? EU15 ? relative ULC significant,
positive - When d ? NEM ? bilateral real exchange rate,
positive - With both variables, relative ULC only for EU15.
RER for both but elasticity higher for NEM. - Taxation
- Statutory taxation not significant
- Ex-post taxation significant.
26ULC mostly affect EU15, RER the NEM
27Taxation only impact FDI in the EU15
28 but things are not so simple
? Higher taxes in the recipient divert FDI
29Further problems competitors taxes (1)
30Competitors taxes (2)
FDI is diverted by higher taxes in the recipient,
but attracted by lower taxes in the recipient,
compared to its (distance-weighted) competitors
31Conclusion
- EU25 is different
- Distance
- But higher taxes divert FDI in the NEM, as in
OECD - Orders of magnitude tax competition and
geography - 1 point change in competitors tax differential
must be compensated for by a 1.4 point change
in the opposite direction in the recipient
country ? sizeable - a 1 sd increase in the market potential can be
compensated for by a 3.1 points increase in the
apparent tax differential in the recipient
country ? even more sizeable