I think your analysis is correct. The plain language of the criterion suggest to me that we are to focus on who bears the effect. The criterion does not ask us to view the predominate effect on a minority or low-income population in relation to non-minority or low-income populations. So, the fact that effects of the project are borne by any minority or low-income population as compared to the effects borne by all other groups in the project area, corroborates your analysis. Moreover, to suggest otherwise would result in ignoring the effects of projects on minority populations when only they would be affected by it. For example, your second scenario would not even need to be addressed because the project only affects one community. I would even go a step further to suggest that the analysis would not be necessary when both communities are minority or low-income, if the focus were on the effect in relation to non-minorities or non low-income populations. Such a conclusion would not
seem to align with the purpose and intent of the EJ.

It sounds to me like the issue revolves on how the project area is defined. If a project is proposed to go through two low-income or minority communities and they would be the only beneficiaries of it, you might be able to say that the impacts were not being borne dispropotionately by those groups. But as in most cases, the project benefits a larger community and that community is not low-income or minority, then it seems to suggest that the "project area" needs to acknowledge that larger context. A narrow definition of the project area would defeat the intent of the EJ analysis.

So you suggests that we only compare within the "project area" and that the "project area" has to be big enough to include all the beneficiaries of the project. Normally, a transportation project benefits almost everybody or at least the whole county/region. In that case, we would have to include the whole county/region in the project area? If so, it would be impossible to analyze impacts to other resources (traffic,noise, bio.... ) for the such a big area. Or do you suggest that the "project area" for the E.J analysis could be way larger than "project are" defined for other analysis? If this is the case, we would always have a "disproportionate high and adverse effect" (and thus an E.J issue) whenever there is a transportation project significantly affecting a minority and low-income community. Is this correct?

Thanks for all your responses

Some people argue that if all communities in the project area are minority/low-income then there is no disproportionately effects and thus there is no EJ issue. All impacts should be addressed as "community impacts". However, I'm not 100% sure if this is true because the criterion doesn't say "predominately" here is to compare within the project area or outside project area so people can interpret either way.

EJ methodology is a tricky thing mostly because to a degree some of the determination is based on the analyst's judgement.
In our state, we do not "identify" a community as a low-income or minority community since we feel that was not part of the intent of the executive order nor of Title VI. It is to help prevent low-income and minority populations from bearing the burden (in this case of transportation projects). So we run demographics and set a study area. Our study area is generally 1/2 mile from center line adjusting for census block group boundaries and neighborhood boundaries. In other words it is not a hard and fast 1/2 mile from center line but rather where will the project's direct and indirect effects be felt? If we need to adjust the boundary due to natural features or boundaries than we do so. If there are minorities or low-income within that study area, we then need to look at the effects that could be felt disproportionately. We look at noise effects, relocations, air toxics etc. To answer the question of whether those populations are being affected disproportionately or not, we look
at who is not being affected. The example I always give is a highway widening. Is the widening and relocations etc. happening on the side where there is a higher percentage of low-income or minorities? This happened historically. It was cheaper land and those folks didn't have political power.
Also if there are relocations and the study area is, let's say 20% EJ, then look at the surrounding areas where there are no relocations. If it is high or higher percent, it is not disproportionate. If it is lower, especially if it is considerably lower, it appears disproportionate. At each step, the analyst must decide whether to ask more questions and investigate further or not.
As you can see, I am disagreeing with a lot of what has been said here. Remember, the point is to determine if there is discrimination, whether it is on purpose or not. Are the people who are bearing the burden bearing it to a greater degree than those who are non-minority and higher income? Going through a community that has a high proportion of minority and/or low-income throughout the project area does not mean those are just community effects especially if that road could have been built elsewhere. It still needs to be examined.
Cumulative effects need to be taken into consideration as well. If a highway divided a low-income community years ago and now that highway is being widened, those previous effects should be considered along with the current project effects along with the, let's say, sewage treatment plant that the city is going to build adjacent to that community.
EJ does tend to be more of an art than a science, which is why it is confusing and hard to measure. After all, how high is high if it is "high and disproportionately adverse"? When it drives you crazy, remember the goal is to prevent discrimination and disclose our effects.

Good summary! It is important to think through your project to determine the EJ analysis needed. We are working on a project on tribal lands which will require the relocation of some residences. However, since the project itself is intended to provide access to the tribal community, we have taken the position that we are not having disproportionate effects. Yes, we will affect a minority population, but since we have no alternatives that do not affect the population, there is no way the effects are disproportionate. Therefore it is a community impact rather than EJ. If, on the other hand, we were putting a road through a tribal community because land there was cheaper than in a nearby non-minority community, we would need to consider EJ.

Thanks!
I disagree with your assumption that because there are no alternatives that do not affect the population, that the effects are not disproportionate and therefore they are community effects not EJ effects. The conclusion that effects are disproportionate is not changed by whether effects can or can't be avoided.
So often people seem to think they cannot conclude that a project has high and disproportionate adverse impacts. That is not true. You can. However, if that is your conclusion, you do need to look at avoidance, minimization and mitigation. If that is the only possible alternative, then that should be documented. Or let's say another alternative would not produce high and adverse disproportionate effects but would eliminate an important wetland. If so, explain that. Also, discuss whatever has been done to avoid and minimize the effects. Then you need to look at how you can mitigate those effects.

A good example is a project that is a bridge replacement. The alternative bridge choices will have an impact on a community that has a high Hispanic population. All bridge replacements will have an adverse effect on the community since the new footprint will have to be different than the old. However, since the footprint varies in size for each alternative, the one that would have the least impact on the community will hopefully be chosen. Fewer businesses and homes will be affected by that one. That choice happens to be a replacement drawbridge. The DOT does not want to build another drawbridge due to maintenance costs. A choice will have to be made between the higher costs for the drawbridge and less impact on the community. If the DOT decides to go with a longer footprint, it would be a high, disproportionate adverse impact on the population since most of their small downtown would be wiped out. The lead agency will need to document the reasons for the choice and look at any
mitigation that can be provided such as possible relocation of some of the businesses within the community.

In your case, you could conclude that the benefits received particularly by the EJ population offset the disproportionate adverse impact. USDOT/FHWA guidance clearly allows for that.