Given that Yannick and I stayed in graduate school for such a long time, we made many friends who got PhD’s in various fields–economics, business, engineering, science… Some went for academic career and are already titled professors; some are still struggling for their tenure; some went to industry and become millionaires from IPOs; some got laid off and came back to school. And those fresh ones, seeing footsteps from their senior fellows, feel confused. They seek help, from advisors and professors, from fellow PhD students, from friends, from families. And the truth is: the more people they ask, the more information they gather, the more difficult the decision becomes.
After our discussion with mOOm about the value of a graduate degree and seeing the struggle many of our friends are facing, I decided to draw a decision tree to help people go through the process. This is, by no means, a replacement for advice from an academic advisor, or your career development help center. This is just my little game to find out whether what I learned in textbook can really find some use in real life.
I am not an engineering and science major, but I find PhDs in engineering and science often face a more complicated choices. They can choose to do a post-doctoral fellowship, just like residency for MD. People from economics or business school normally do not have to go through this. So I decided to draw a decision tree for Sisi, a fresh PhD from science and engineering.
Sisi got her PhD in applied math from a top school. As a fresh PhD, Sisi does have some options, and that’s exactly what she and her family are struggling with.
(1) She can get an assistant professor position in a 2nd or 3rd tier university right now, which pays about $60-80k.
(2) She can work in the industry, which pays six figures right away.
(3) She can wait for another year or two as a post-doc fellow, and shoot for a faculty position in a1st tier university.
Here is a very Simplified version of a pseudo-decision-tree.
Two main reasons make this tree “pseudo”. First, Sisi is struggling with assigning probabilities to each branch. What’s the probability of getting unemployed in 2 years? Sisi had a friend who joined a very big and profitable company 2 years ago, and come back to school now due to company re-structure. What’s probability of building a successful business of her own? Go to any bar in Silicon valley any day, Sisi will find 1/2 of people there dreaming of building another google, or at least a youtube. But we had only one Google, one youtube. Sisi believes that she has a technical edge to build his own business, but how can she assign a probability on this?
Second, Sisi puts only monetary value in stead of utility on each node. What makes this decision even more complicated is that Sisi and her family assign different utility to each outcome. Sisi’s husband wants her to go to a 2nd -3rd tier schools, so she can has less pressure and get a baby soon. He also puts higher utility on going to industry, since he has the same dream as a valley girl and wants Sisi to join him.
However, as a female who window-shops designer clothing from time to time, Sisi puts higher utility on “brand names”. Practically speaking, option-3 is the least ‘economical one. Doing a post-doc costs another 2-3 years with minimum pay. After that, if she gets to a 1st tier university, the pay is generally lower than 2nd tier university. Top universities always have higher bargaining power, in accepting students, in recruiting faculty. However, how many 1st tier universities can you find? Supply and demand determine the price. If Sisi wants to get the brand equity of 1st tier universities, Sisi has to pay the premium. Sisi is ready to pay for it, her husband is still hesitating.
And as a female, Sisi puts higher utility on security (on this part, Sisi is not alone ). Getting tenure in whatever university will give Sisi’s family low but guaranteed income and legal status. As immigrants without a green card, Sisi’s first priority is: Security! Security!! Security!!!
Still, the problem of assigning probability remains. I know that many fresh PhDs are facing Sisi’s decision tree. In fact, many friends of ours are on the same boat right now. Assigning these probabilities requires experience, across time, across people. I hope that senior readers who have experienced all these can help Sisi fill in the probability in the tree. If we can get a large sample, the law of large number will work for us.