TLDR: from the little research there is on this (one paper seriously trying to eliminate confounders, many off-the-cuff options otherwise), there's still an unexplained "foreign fine gap" in FCPA enforcement, particularly when the SEC part of the (total) fine is included in the model. Confounders that were considered (but don't fully explain the gap) include: market capitalization of the company, bribe amount, voluntary disclosure, how poor is the country (GNI per capita) where the bribe was given, whether a subsidiary company was also involved, and whether a foreign regulator also fined or at least investigated the company in the same case.
The less difficult (statistically) analysis of FCPA enforcement over time supports the idea that political factors may have played a role in the amount of fines levied. E.g., before the OECD convention that allowed easy sharing of such bribery information between countries, the FCPA fines were mostly applied to US corporations, and these fines were quite small compared to the contemporary ones.
As far as I know it's hard to explore "differential DOJ targeting" because for that one would need to know the number of actual corruption cases as a baseline, which is mostly unknown. But I can confirm from a different source that mostly non-US corporations have been prosecuted, at least until 2014.
Figure [...] below shows the 20 highest FCPA penalties to date. Of the top 10, 8 were foreign.
On the other hand...
30% of [total] FCPA cases have been brought against non-U.S. companies. Yet these 30% of companies have paid for 67% of the total FCPA fines. Thus, foreign firms are paying more than five times the FCPA fines paid by domestic firms. [and more stats]
Perhaps there is some proxy measure for actual corruption cases that can be used as a proxy baseline... for the kind of confounder elimination you want, but insofar I haven't found a paper exploring such.
Sure, one could argue that the US is only bringing big profile cases against non-US companies, leaving the foreign small fish alone... but I don't see how you could easily test this alternative hypothesis.
That paper proposes that US prosecutors has a lot of power in bringing such cases:
As a result of the high cost of litigation and the broad interpretation and uncertainty of the FCPA, cooperation is generally the only viable option for firms who are charged with FCPA violations. As Yockey notes, “[t]he consequences of indictment, prosecution, and, ultimately, conviction are seen by many firms as being too great to consider doing anything other than settling.” Firms that are charged with FCPA violations generally take a severe reputational hit. Even the announcement that a corporation is the subject of an FCPA investigation can lead to significant reputational harm. This is why counsel generally advises firms facing FCPA related charges to settle on good terms as opposed to pursuing a trial on the merits.
For these reasons, federal prosecutors have a significant impact on interpreting the law and have substantial leverage when it comes to FCPA investigations. The DOJ essentially controls the disposition of the FCPA cases they initiate and impose their own extremely broad interpretation of the FCPA’s provisions. Settlements receive little judicial oversight. In light of this, federal prosecutors in FCPA cases yield immense power.
An argument in support of this view (of high discretionary power) is the spread over time of FCPA prosecutions and the fines imposed:
Given the robust enforcement of the FCPA today by both the DOJ
and SEC, the near-complete lack of FCPA enforcement in the statute’s
first two decades provides a striking contrast. Between 1977 and 1996,
the agencies collectively brought only 40 cases (the median year would
see two cases or fewer) and settled these charges on sympathetic terms
(the average of the ten highest fines was under $10 million). By
comparison, between 1997 and 2016, the agencies brought 428 cases
and started to collect blockbuster settlements (the average of the ten
highest fines in that period was $484 million). Even accounting for
inflation, the tenth highest fine today is more than twice the combined
penalties of the top ten fines between 1977 and 1996. The number of
cases and the size of penalties are of a different order of magnitude in
the last two decades than the first two decades. [...]
After the initial passage of the FCPA, American industry argued that
the FCPA put them at a competitive disadvantage with foreign rivals,
who would not be bound to similar anti-bribery rules. When the FCPA
was enacted, other major developed countries (such as Germany and the
United Kingdom) did not prohibit foreign bribery and even subsidized it
by making bribes tax-deductible. This made the enforcement of the
FCPA politically unviable. Various efforts were made to repeal the
statute, but these efforts were largely muted by the executive branch’s
decision simply not to dedicate resources to enforcement.
The latter article goes on to discuss how the OECD Anti-Bribery Convention and then the Sarbanes-Oxley Act multiplied prosecutions. As far as I can tell, that doesn't explain why the fines are a lot heavier (on average) against foreign corporations though; the latter article doesn't seem to touch this issue specifically. But it does highlight that differential costs to US vs foreign corporations was alway a (political) concern in FCPA enforcement.
Interestingly, if you consider the total sums of fines (not taking into account the per-case fine), the stats from first paper are somewhat more equal (2.015 billion by foreign corps and 1.545 by domestic corps.) I don't know if there was any specific goal to reach a kind of parity like that though.
On the other hand, the 2nd paper notes:
Germany, Switzerland, and the United Kingdom have revised
their anti-bribery laws and now regularly bring foreign corruption cases
against domestic firms. Other OECD nations, including France and
Canada, are reforming their anti-bribery legislation and possibly could
become more active enforcers. Nevertheless, the vast majority of
OECD states have limited to no enforcement of their anti-bribery
laws.
This is quite interesting, because the heaviest US fines were exactly against companies from countries that now seriously enforce such anti-bribery laws domestically; going back to the first paper:
Figure [...] below shows the nationality of the foreign corporations most heavily targeted in FCPA enforcement actions. The United Kingdom, Germany, Switzerland, and France have had the most number of corporations targeted in an FCPA enforcement action.
So, possibly, just possibly, the US is riding the coattails (of such domestic anti-bribery investigation abroad) with their own FCPA fines. This doesn't explain the magnitude of the fines though.
Here's an updated (to 2016) fines top also contrasted with the earlier, pre-OECD convention fines (from the 2nd paper); there now one more heavy fine against a US corp, shifting that tally to 7:3.
And for (my) "riding the coattails" theory, it's somewhat backed by this 2nd paper:
The OECD Convention has promoted information sharing both
formally and informally. On a formal level, the treaty committed
governments to providing mutual legal assistance in gathering evidence
and sharing information. Prosecutors can make requests to their
overseas counterparts for documents or to find individuals. This formal
legal assistance is frequently acknowledged by the DOJ in their
settlements. Aid in evidence gathering also often occurs in a less
formal and less centralized manner. Investigators and prosecutors are
able to reach out to foreign counterparts without necessarily going
through their national governments. As a result, cooperative
relationships can form even without the encouragement (or even with
the discouragement) of high-level political (and often elected) leaders.
In addition, the OECD’s requirement that countries enact domestic
laws to criminalize bribery has created government offices whose
regulators are responsible for investigating claims of bribery. Even if
foreign governments do not prosecute many cases themselves, the fact
that all OECD states have government offices with jurisdiction over
foreign corrupt practices provides American officials with a host of
foreign regulators who share their mandate. These foreign
investigations can be fertile ground for evidence sharing. For instance,
the French government’s investigation into Technip’s bribery of
Nigerian officials resulted in tips to American officials regarding
Technip’s and Halliburton’s activities in Nigeria. Similarly, the
American case against BAE Systems (“BAE”) was built on the British
investigations into the company, an investigation that was shut down for
political reasons in the United Kingdom but later resulted in a joint U.S.-
U.K. settlement.
Through formal and informal channels, the OECD Convention has
resulted in much more transnational cooperation in evidence
gathering. Both the treaty’s legal obligations and greater foreign law
enforcement interest in corruption has resulted in more investigations
and greater willingness to share evidence. This expanded cooperation
has been critical to a spike in successful American prosecutions.
Like I said, this still doesn't get to the magnitude of the fines on foreign corps though. And of course opinions differ on that
This Table (while suggestive) does not control for the size of the
bribe, the level of corporate benefit from the bribe, the degree to which
top management was involved in the bribery scheme, the level of
cooperation with U.S. officials, or the quality of evidence; so it is not
evidence that the DOJ and SEC are making sure that as many foreign
firms face prosecution as American ones. However, some commentators
have maintained that the U.S. policy of seeking high fines against
foreign firms is discriminatory and violates general international legal
principles of equal treatment before the law. While most
commentators (including the author) would strongly resist the idea that
the DOJ or SEC are purposefully discriminatory toward foreign firms,
one study has found that foreign firms face higher FCPA fines than
American ones. Professors Stephen Choi and Kevin Davis find that
the DOJ assesses greater FCPA penalties against foreign firms than
domestic ones, even accounting for the size of the bribe and whether the
firm voluntarily disclosed the illegal activity, although not controlling
for the quality of the evidence, the participation of senior executives, or
the level of cooperation.
Note that this "one study" is a different one from the one I started my answer with. So the opinion may be more widely shared that this (undocumented) "counter-consensus" claim. The abstract of the latter
We examine factors that might explain how sanctions imposed in Foreign Corrupt Practices Act (FCPA) enforcement actions vary across the firms and countries implicated using a dataset of FCPA actions resolved from 2004 to 2011. We find evidence that the sanctions in an individual FCPA action are positively correlated with the egregiousness and extensiveness of the bribe. The sanctions also increase if the ultimate parent company of entities involved in the FCPA violation is foreign and if foreign regulators are involved in the action. At the country level we report evidence that the SEC and DOJ impose greater aggregate sanctions for violations in countries with a lower GNI per capita and weaker local anti-bribery institutions. The SEC and DOJ also impose disproportionately greater aggregate sanctions for violations where the home country of the ultimate parent company of FCPA defendants has a greater GNI per capita, stronger anti-bribery institutions, and a cooperation agreement with U.S. regulators. Overall, these findings suggest that factors besides those deemed relevant by U.S. and international law influence enforcement of the FCPA.
Model 4 supports the hypothesis that U.S. companies are fined disproportionately
less compared with foreign companies [...]. The coefficient on
US Company is negative and significant at the 1% level. [...]
To test all the hypotheses in one
model, we re-estimated Model 1 of Table 3, adding Vol. Disclosure, GNI Per Capita, US
Company, and Foreign Regulators as independent variables in the same model. Model 6
of Table 3 reports the results. [...] We test the Self-Interest hypothesis in Model 6 with the inclusion of US Company
as an independent variable. The coefficient on US Company is negative and significant
at the 5% level, indicating that all other things being equal the SEC and DOJ impose lower FCPA penalties on U.S. companies, consistent with the Self-Interest hypothesis. [...]
They did a whole bunch of robustness checks e.g.
First, we reestimate
Model 6 of Table 3 with the log of 1 + the total DOJ monetary penalty for a
particular FCPA action as the dependent variable (excluding SEC monetary penalties,
which include disgorgement). We report the results as Model 1 of Table 4 with robust standard errors in parenthesis. Due to the small sample size we also report bootstrapped
standard errors in brackets for Model 1 and all the other models in Table 4. The
significance levels of the coefficients are similar. We obtain similar qualitative results as in Model 6 of Table 3 with two exceptions. The
coefficient on US Company is no longer significantly different from zero. Note,
however, that the coefficient on US Company remains significantly different from zero at
the 10% confidence level using bootstrapped standard errors.
[...] to control for possible time trends in FCPA enforcement, as a robustness
test we added indicator variables for the resolution year of the FCPA enforcement action to Model 6 of Table 3. [But] we obtain the same qualitative results as in Model 6 of Table 3.
[...] since the monetary penalties cannot take negative numbers, the dependent
variable in Model 6 of Table 3 is constrained to be zero or above. As a robustness test, we re-estimate Model 6 of Table 3 using a tobit model [and]
we obtained the same qualitative results as in Model 6 of Table 3.
In most of these the US vs. foreign indicator remained significant. It became less so when considering only DOJ (and not also SEC) action and/or considering the fine as part of structured penalty (an ordered logit model with 6 coded outcomes), but the latter hardly had any significant explanatory variables, e.g. not even the bribe amount was explanatory anymore for anything in this ordered logit. (I think this model didn't have enough statistical power, i.e. not enough data given the amount of stuff [dependent outcomes] they tried to simultaneously explain with it.)
So at least when controlling from some potential confounders, there's still a "foreign fine gap" in FCPA. Also, there's some support of the "riding the coattails" idea. A with all such (observational) research, you can never quite eliminate all confounders:
Although our models control for Bribe Amount,
they may omit factors correlated with both egregiousness of the violation and variables
such as US Company, Foreign Regulator or Foreign Reg. Sanction. For instance, it could
be that the most egregious misconduct tends to attract attention from both U.S. and
foreign regulators. This would explain the positive relationship between Foreign
Regulator and Foreign Reg. Sanction and the total monetary penalty. Similarly, it could
be that U.S. firms tend to be more skilled at demonstrating unreported forms of
mitigation. This would explain why they receive lighter penalties. Our models also do
not take into account all factors bearing on the ease of detecting the defendants’
misconduct. Those factors might be relevant if regulators employ sophisticated
deterrence-oriented enforcement strategies.
For (a converse) perspective, the EU has levelled some big fines on US companies, e.g. on Google, on antitrust grounds, a measure that Obama himself called protectionism. How do you disprove that claim? The Obama administration had a very close relationship with Google; see article for details. But Trump said the same thing even though he might not have the same close relationship with Google, although he may have other motives to do make that kind of claim. Etc. The counter-explanation for that is "they are likely a result of the Commission’s decision to focus significant attention on the tech sector. Because the vast majority of large tech firms are US-based, all else equal, it is to be expected that the majority of investigations and enforcement actions would involve US firms." But of course, that doesn't fully explain why is the Commision focusing on the tech sector to begin with, etc. "The decision to prioritize enforcement in the tech sector is not taken in a vacuum. Whether this policy preference is down to legitimate concerns about high-tech markets or to (potentially unconscious) protectionism is almost impossible to tell."
