bkudrzycki / 2_survey Goto Github PK

• wage -> self-employment
• wage -> wage (vs wage -> self or wage -> neet)
• apprentice/school -> work
  • apprentice -> wage vs school -> wage
  • apprentice -> self vs school -> self
  • apprentice -> neet vs school -> neet
  • school -> wage vs school -> self vs school -> neet
  • apprentice -> wage vs apprentice -> self vs apprentice -> neet
• only a single occupation
• school -> work -> school

• Has nobody done high frequency surveys of youth? Double check

• Do separately for men and women
• Control for age (NOT at baseline)
• Control for what activity people are doing, not the change in occupations
• start with model 4 and then add wage - Isa (we don't have good wage data -ed)
• add education - employment transition

Transition propensity tables "not totally clear" - Isa

Use different base rates: e.g. what is the propensity to transition to formal work conditional on being NEET in the previous period?
Do we need the p-values?
How did Cunningham and Salvagno present the data?

Ask question before detailing method

a) what do we know
b) what question asked in this paper
c) what do we do in the paper
d) result

• - 87% self-employed transitions: stable/absorbing
• - RARELY do self-employed transition to wage employment
• - look at it by gender
• - take out absolute numbers

Table \ref{tab:eductable} in the Appendix details the educational attainment of youth in our survey by their activity at baseline. The similarity in upper education completion rates for employed and inactive youth are strongly suggestive of a queuing phenomenon, in which qualified youth endure long stints in unemployment while waiting for an offer from a limited pool of wage employers @serneels2007}. Moreover, educated youth are more likely to be inactive than to be self-employed: self-employed youth in our sample have completed primary education at half the rate of the NEET youth, and are 50% less likely to hold a baccalauréate or university diploma. In fact, a greater proportion of NEET youth holds a Bachelor's degree (License) than the employed, exemplary of the preponderance of the so-called "educated unemployed" observed throughout sub-Saharan Africa (see, for example, @matsumoto2010}).

(in contrast to: "full-time work following a spell of inactivity is the most common path taken")

Isa suggestions:

• male vs female
• retrospective data vs panel data
• high- vs. low-income countries

Also look for comparable statistics in low income settings!

Isn't there an updated paper from Quintini from around 2017 that includes more countries?

And say what is missing (i.e. open questions, what we don't know)

• Reflect on this briefly and consider only reporting raw means if it makes more sense
• This will be questioned by committee - if decide to keep, have to at least have an explanation of what it's doing (even if it's not justified)

May be smart to cut losses on just use a single of the stats: maybe combined? And explain the thought process in detail in the Appendix?

"Finally, we find that youth who have transitioned to the labor market experience more life satisfaction, though no effect is detected for particular types of transitions."

What do we know about HICs?

Structure (provisional)

2.1 sampling

2.2 method of data collection

2.3 data

OPTION 1

3.1 how long from education to labour and what determines length of transition

3.2 aspirations and what determines aspirations (maybe can be included in transition paths)

3.3 neet and what determines neet

3.4 transition paths

3.5 job quality

3.6 life satisfaction

OPTION 2

3.1 how long from education to labour and what determines length of transition

3.2 neet, transition paths and aspirations

3.3 job quality and life satisfaction

OPTION 3

3.1 duration and neet and aspirations

3.2 transition paths

3.3 job quality

3.4 life satisfaction

1. start with the question/gap/puzzle
2. what do we do
3. result
4. what does it mean

• Less stable than in yearly data, may indicate recall bias
• NEET stats: how many were ever NEET, and how many went "back and forth with unemployment"?
• Perhaps this could mean the percentage that was wage (or self) then NEET then wage (or self) again?
• Include NEET results in this section

More ideas:

• Average age of NEET for pooled sample
• Average % NEET by age for pooled sample

How many & for which countries?

"Only about a quarter of youth is observed to have transitioned to the labor market, despite a mean respondent age of over 24 years."

vs.

"we find that youth in Cotonou graduate at the age of 22 years, four months, and find their first employment opportunity about about 5 months later, on average, but find attachment to the labor market only after about two years on the job search."

I think so, double check

also:
HOW DONE, WHICH COUNTRIES, WHICH YEARS?