Continuing to examine Donald Trump's interview with the Washington Post earlier this week reveals a candidate making some wise decisions on-the-fly.

The Post asked Trump the following questions regarding purported racial discrimination in the American law enforcement system:

HIATT: Do you see any racial disparities in law enforcement -- I mean, what set it off was the Freddie Gray killing, as you know. Is that an issue that concerns you?

TRUMP: Well, look, I mean, I have to see what happens with the trial. I --

HIATT: Well, forget Freddie Gray, but in general, do you believe there are disparities in law enforcement?

TRUMP: I've read where there are and I've read where there aren't. I mean, I've read both. And, you know, I have no opinion on that. Because frankly, what I'm saying is you know we have to create incentives for people to go back and to reinvigorate the areas and to put people to work ...

RUTH MARCUS, COLUMNIST: But Mr. Trump, if I could just follow up on Fred's question. I think that what he was trying to get at was the anger in the African American community that held some of the riots and disturbances this summer about disparate treatment and about ... clearly you say you've read where there is disparate treatment. But it is pretty undeniable that there is disproportionate incarceration of African Americans vs. whites. What would you -- is that something that concerns you?

TRUMP: That would concern me, Ruth. It would concern me. But at the same time it can be solved to a large extant with jobs ...

It is clear that Hiatt and Marcus were trying to bait Trump over the racial differences in incarceration, and that Trump was having none of it, with Trump responding that "I've read where there are and I've read where there aren't."

He is right, the data is most certainly not clear. The conventional narrative that there is racial discrimination in favor of non-Hispanic whites evident in the U.S. incarceration data falls apart upon serious examination.

One of the best data sets to prove the point was recently published by the Police Foundation, and it shows the "percentage of males 18 to 39 incarcerated in the United States, 2000, by nativity and level of education, in rank order by ethnicity."

The argument is that if you are not white (i.e., non-Hispanic white), you are far more likely to be incarcerated in the U.S., and this difference is substantially -- if not dominantly -- due to systemic racism against non-whites in the American justice system.

If you just look at the macro-groupings of non-Hispanic whites, Latin Americans, and blacks, the systemic racism theory that the Washington Post editorial board was getting at appears to be supported. The incarceration rate of non-Hispanic white males 18 to 39 is just 1.66%, versus 3.26% for Latin Americans, and 10.87% for blacks. In other words, the darker your skin color, the far more likely you are to be incarcerated.

And these are often the only three macro-groups those promoting the racism narrative discuss.

But what about Asians? Their incarceration rate is only 0.62%, or about one-third the rate of non-Hispanic whites? This does not fit the theory. In fact, this data -- on its own -- proves the racism argument is nonsense.

If your theory is that darker skin tone equals higher incarceration rates due to systemic racism, then all -- not just some -- of the data must fit the theory.

If incarceration rates could be explained by skin color, then we must conclude that the justice system discriminates against blacks when compared to Latin Americans because of racism, and discriminates against both Latin Americans and blacks compared to whites because of racism, and discriminates against whites compared to Asians because of racism. Wait a minute, that last point is inconvenient. The justice system really discriminates against whites in favor of Asians?

Of course not, but that is where the race-baiters' theory inevitably leads if all the data is considered, not just cherry-picked data that suits the racism narrative.

Just to be clear, we are also not dealing with small numbers, such that some could claim that the difference in incarceration rates between these groups is not significant. For incarcerated males 18 to 39, there were 29 million whites, 5.5 million blacks, 7.5 million Latin Americans, and 1.9 million Asians. The size of these groups is each large enough that the differences between them is not due to insufficient "sampling size"; rather, the differences are very real.

The situation gets even worse for the systemic racism proponents if we dig deeper into the data and consider variation within each macro-group. Again, the size of these groups is also large enough (i.e., hundreds of thousands to millions) that the differences are real, and not just artifacts because we have too few of one or more samples to assess what is really going on.

So if racism is the dominant explanation between incarceration rates of various ethnicities, how is it that individuals from India (in the Asian macro-group) have an extremely low incarceration rate of just 0.22% (almost eight-fold lower than non-Hispanic whites at 1.66%), whereas Laotians and Cambodians (also in the Asian macro-group) have a much higher incarceration rate of 1.65%? Is the justice system discriminating against both non-Hispanic whites and Laotians/Cambodians in favor of Indians, with no discrimination between non-Hispanic whites and Laotians/Cambodians? That doesn't make any sense.

Well, none of the race-based arguments make any sense, but soldier on we must.

Among Asians, the incarceration rates are as follows: Indian, 0.22%; Chinese/Taiwanese, 0.28%; Korean, 0.38%; Filipino, 0.64%; Vietnamese, 0.89%; and Laotian/Cambodian, 1.65%.

Once again, the size of each of these groups is large enough that the incarceration rate differences are real. Is the justice system actually discriminating against Vietnamese in favor of Koreans? Keep in mind that the incarceration rate ratio of Vietnamese:Koreans is 2.3, whereas the corresponding ratio for Latin Americans:non-Hispanic whites is just 2.0. In other words, using the reasoning of the racism theorists, there is more statistical evidence for systemic racism against Vietnamese in favor of Koreans than there is for systemic racism against Latin Americans in favor of non-Hispanic whites.

Add to this the undeniable fact that the incarceration ratio of Laotians/Cambodians:Indians is 7.5, whereas that of blacks:non-Hispanic whites is just 6.5, and we have more statistical evidence for the systematic racism against Laotians/Cambodians compared to Indians than we have for blacks compared to whites.

The situation becomes even more tenuous for the race-based theory when we go on to look at the incarceration rates among the various Latin American sub-groups: Salvadoran/Guatemalan, 0.68%; Colombian/Ecuadorian/Peruvian, 1.07%; Mexican, 2.71%; Dominican, 2.76%; Cuban, 3.01%; and Puerto Rican, 5.06%.

And yet again, apparently more statistical evidence for systemic racism against Puerto Ricans compared to Salvadorans/Guatemalans than exists for blacks versus whites.

It is obvious that the justice system is highly unlikely to be systematically discriminating against Puerto Ricans in favor of Salvadorans/Guatemalans, against Laotians/Cambodians in favor of Indians, and against non-Hispanic whites in favor of Asians -- but that is exactly what the race-based theory for incarceration differences tells us must be occurring when all of the racial data is considered, as opposed to just a subset of the data that conveniently suits the racism narrative.

The rational conclusion, which Trump appears to have taken, is that the data -- when not cherry-picked to arrive at pre-determined findings -- does not show any clear systemic racism in the criminal justice system. Any good theory in the social or natural sciences must explain all available data, not just some of it. Based on this criteria, the racism-based incarceration rate theory fails the test and must be rejected.