Last week, the news feeds on all my social networking sites had one major update: Marriott finally completed the acquisition of Starwood and became the world's largest hotel chain.

While the Marriott-Starwood merger is still in the spotlight, I would like to bring up another piece of news for discussion — there is evidence suggesting Airbnb hosts discriminates against African-American travelers. I expect more discrimination cases will soon emerge against every sharing economy platform.

In May, an African-American man filed a lawsuit against Airbnb for its ignorance of the hosts' race discrimination, which I believe was the first discrimination case against Airbnb or a sharing economy platform. This month, the discussion on Airbnb's racial discrimination heated up again with a new study conducted by three Harvard professors.

In their investigation, the professors first created four pseudo profiles with identical information except the user names, which were purposefully created as a signal for perceived race and gender of the traveler, namely as an African-American male, an African-American female, a white male and a white female. Then, they sent out about 6,400 inquiries to the Airbnb hosts in Baltimore, Dallas, Los Angeles, St. Louis and Washington, D.C., requesting for a stay in September 2015.

Airbnb allows hosts to decide whether they want to accept or decline a request from a traveler after they review the traveler's profile or his/her linked account on other social networking websites such as Facebook. The hosts would then process the requests made by one of those four pseudo travelers with one of the following five actions:

• "Yes"
• "Conditional Yes," meaning a pending decision where the hosts may require additional information from the traveler or needs more time to reach a decision. In my own interpretation, the hosts could probably want to wait for additional (often better) requests from other travelers before making a final decision, such as a request for a longer stay.
• "No response"
• "Conditional No" — In my own interpretation, even though the hosts want to turn down the current request while waiting for a better one, such as request for a longer stay if the rate remains the same, they may change their mind later and say "Yes" if they receive no other requests by a certain day.
• "No"

The professors performed a series of logistic regressions and made the following conclusions:

• Both African-American and white hosts discriminate against African-Americans.
• Both male and female hosts discriminate against African-Americans.
• All African-American travelers are discriminated against, regardless of whether they are male or female.
• Such discrimination persists across various listing types (entire property vs. shared property), host types (those with multiple properties vs. those with only one property; those with 10-plus reviews vs. those with fewer than 10 reviews), price point and neighborhood (diverse vs. homogeneous).

In the end, the authors suggested Airbnb should conceal travelers' demographic information until a reservation is made and expand its "Instant Book" option, an existing function that allows hosts to accept requests without screening the inquiries but is not commonly used by the hosts. Should Airbnb take their suggestions, it would turn all Airbnb listings into those typical hotel rooms on the block.

Think about it, accepting a reservation for an available room as soon as a traveler makes a request (reservation) is exactly what a regular hotel does everyday, right? Nevertheless, Airbnb is not just a typical lodging product. Very likely, Airbnb does not want to turn itself into another giant hotel chain either.

People choose Airbnb over hotels for a good reason. Many want the unique experience provided by the hosts. Some travelers will even end up becoming friends with the hosts after a trip — it happened to several of my friends who traveled to Taiwan and Europe in the past.

If we consider Airbnb as a social networking platform where hosts and travelers can connect with one another and make friends, I wonder how Airbnb as a company would have the right to prohibit the hosts from reviewing a traveler's profiles before making a decision.

Shouldn't the hosts also have the freedom to choose whom they want to welcome in their property as their guests? Especially when the hosts want to treat their guests like friends or family members, shouldn't they have the right to choose whom they want to hang out with? Or must the hosts accept any requests from anyone just because that is the "right" thing to do these days?

As what we know in stereotyping — where people tend to believe certain types of individuals would conduct some specific behaviors — it is natural for human beings to feel attracted to (or skeptical about) certain types of people.

Imagine you are an Uber driver, and you receive two requests for a pick up in the middle of the night. One is from a friendly person wearing clean and nice clothing; the other one is from a person wearing dirty clothes, possibly with weapons or some sort of tools in the pocket. Please note that I am not even using race or gender as a distinguishing attribute here.

Based on your gut or stereotyping, which guest would you pick?

Now, think twice about why you prefer one guest over the other. The fact is no matter how much we hate stereotyping, it is just buried in the bottom of our hearts, isn't it? In reality, it is possible that the friendly person might actually be a serial killer, and that the person in dirty clothes could be a nice person who just happens to carry some tools for work. Yet people often make decisions based on their past experience or the stereotypes they formed from various sources.

As a minority myself who is "supposed" to be discriminated against all the time — as a matter of fact, I have been turned down by an Airbnb host in the past — I would like to urge people not to make a rush to conclusion on Airbnb's racial discrimination case solely based on the Harvard study. There are many unanswered questions in their study that may need additional attention.

For example:

Profile detail: What other information was included in those four pseudo profiles? A complete profile with detailed descriptions of a person and a profile with one sentence or two could make a huge difference. In a case when the user name is the only information clue for race and gender, I expect stereotyping would carry more influence on the hosts' decisions.

Socioeconomic status: Because those Airbnb listings in the sample charged for a median price of $163 a night and a mean price of $295 a night, will other information about a traveler's socioeconomic status play a more significant role when the hosts made a decision to accept or decline an request? When no socioeconomic information is revealed in the pseudo profiles, will more hosts tend to decline an offer just for the sake of safety? Likewise, when the hosts feel uncertain about who is actually renting their property, should they be allowed to refuse a request or demand more information from the traveler?

Background check: Many landlords have the right to do a background check on potential tenants. When background check is not an option for Airbnb hosts, will it be fine for them to judge their guests based on some clues of information? In the cyber world, those clues of information often come from a person's profile. Then, why should Airbnb conceal travelers' information for hosts or push the "Instant Book" option, leaving no clues for the hosts to make an informed decision?

Profile picture: As suggested by the authors, having a profile picture could also be an important part of their experiment. Yet none of the four pseudo profiles included a picture. If a stranger wants to connect with you on a social networking site, but s/he does not use a real picture of herself/himself, wouldn't you demand for more information to verify who that person is? Or you may simply ignore or decline their requests.

Appearance: If we use a picture for a female African-American who looks like Beyoncé, Jennifer Hudson or Michelle Obama, a picture for a male African-American who looks like Will Smith or Barack Obama, and the pictures of the white counterparts with unpleasant appearance, I wonder if the experiment will yield the opposite result. If those good-looking African-Americans get a higher acceptance rate by the hosts than those white travelers with unpleasant looks, shall we call it a discrimination against those with a unpleasant look? Or is it stereotyping? If, however, good-looking African-Americans still get a significantly lower acceptance rate, Airbnb hosts may indeed discriminate.

At this point, I am pleased to see how Airbnb is taking a proactive approach in responding to people's claim about Airbnb hosts' racial discrimination. The company immediately rolled out a new nondiscrimination policy.

Nevertheless, Airbnb will likely continue to see discrimination claims until it is able to force all hosts to open their doors to whomever rings the bell without taking a peek on who is actually at the door. This prediction applies to all sharing economy platforms that allow users to see one another's profiles.

What will be next? Uber? Age discrimination? Sex discrimination?