Entrepreneurship and optimism Flashcards
What is the Paper for this topic?
Curb your enthusiasm: Optimistic entrepreneurs earn less de Meza et al (2019)
What is the paper about?
Entrepreneurs tend to be optimists.
The paper asks whether optimsits starting businesses do better or worse than realists or pessimists.
On average optimists entrepreneurs earn less than realists.
What is the meaning of optimism in the paper by Meza et al (2019)
When you have an upward bias in expectations ( you believe something is better than what it should realistically be.
How do you measure whether someone is an intrinsic optimist?
ask them to forecast their performance in some repeated task. If on average their forecasts exceed their actual performance, they are classed as an optimist. ( the paper uses this method)
We are going to do some theory so interpret the P(z,x) line?
This line is earnings in paid employment, there might be some increase in earnings if you have entrepreneurial ability, even in paid employment but not going to be dramatic.
First of all does being SE mean entrepreneur ability matters a lot for earnings?
Interpret this ( first look at the S(z,x) line?
Yes if you become SE then entrepreneur ability is going to matter a lot. As we see with the 2 lines.
If you are rational, and your entrepreneur ability is not high then you should be in paid employment, if your entrepreneur ability reaches a certain level then switch to self employment.
What happens when people are optimistic? So the O + S line?
If your optimist then you will switch to SE when you think you have entrepreneur ability above Z^o not, but they think there earnings will be the diagonal vertical line but it isn’t, so they have switched too soon into SE, you would of been better off staying in paid employment.
What is one issue of measuring optimism?
So optimists are below the line, so if they predict higher then what they actually got then they are classified as an optimist.
However it might just be because of luck, so you might become an optimist because of bad luck.
Lets say you have stochastic income with expected value of 50k, a rational person will predict that there income next period is 50k, but they might experience a shock so income is 49k, thus classified as an optimistic, but if next period, distrubtion doesn’t change than income increases to 50k then on average optimistis are observed to have income increases? Whats the problem?
Its not informative and doesn’t fit with hypthoesis as optimism is positively correlated with earnings, which shouldn’t be the case, it should be negative.
So lets look the experiment so first of all they get Data from 18 annual waves of British Household Panel Survey from 1991 to 2008. ( includes a lot of data such as people being SE )
In the survey they ask 2 key questions used to calculate optimism:
1) Looking ahead, how do you think you yourself will be financially a year from now; better than you were now, worse than you are now, or about the same?’
2) Would you say that you that you yourself are better off, worse off or about the same financially than you were a year ago
Now what is our optimism error/ forecast error?
Who are futures and who are nevers?
Futures are people at some point who get SE.
Nevers are those who never go into SE.
How can we measure intrinsic optimism for futures?
We use the forecast error the periods before they become SE, so in paid employment.
How can we measure intrinsic optimism for nevers?
we use forecast error for the first half of the available years in paid employment.
What us a problem of using forecast error that we have already said?
It can be noisy estimate of intrinsic optimism e.g. if you have had an unlucky realisation you are classed as an optimist. ( hence why they use several periods, but they even do better than that)
How does Meza et al combat Forecast error noisiness.
The paper estimates fixed effect regression . This is when for each individual you use their forecast error and regress that on a bunch of variables which can capture environment (recession) and individual fixed effects. So the coefficients of these dummy variables will be defined as the measure of intrinsic optimism.
This table provides summary statistics for future and nevers in their first phase where everyone is an employee. In total there are 31,968 observations from 7985 individuals. Of these, there are 3138 futures observations from 618 individuals. For nevers there are 28,830 observations from 7367 individuals, therefore intrinsic optimism is constructed from an average of 5.1 observations per individual for future and 3.9 observations per individual for never. Interpret the results.
On average the financial forecasts of futures exceed those of nevers but average realisations are only marginally lower for futures. The forecast error is in the optimistic direction for both groups but futures are more optimistic than nevers
This is the full regression equation of optimism estimates, with a number of controls that could influence financial forecasting or realisations . What should we be looking at here?
The controls are jointly significant in some places and a few are individually significant.
Show the final SE equation?
What does figure 2 show?
displays distubtions of our measure of standardised optimism, Oi^ for futures and nevers, and this shows that futres are significantly more optimistic at all points on the lower 3/4 of the distubtion.
Dependent variable here is who becomes self employed , its a 0 1 variable, what can we interpret?
standard optimism has a significant positive coefficient, so the more optimistic people are the more likely they are to become self employed.
The ability measures ( past earnings and past realizations have a significant and positive effect of selection of self employment.
Also if both parents are self employed then that hugely increases chance of being self employed.
How do we interpret the coefficient for ‘Both parents self-employed’?
We look at it on the base that no parent is self employment so add 0.068 + 0.026 + 0.022. - this just shows the incremental effect.
If actually both parents are self employed then we we add 1 + 1 + 1
Now lets look at the SE earnings?
So running a regression on earnings and a bunch of other stuff, we find that, optimism has a significant negative effect on SE.
In paid employment, the more optimistic they are the more they earn. Significant and positive.
Someone in SE on average would of earned more if they said in paid employment
Extending the SE earnings regression why does ‘married/cohabiting-partner not employed’ have a significant positive coefficient’?
If your partner doesn’t work, then you have to work harder to offset this, hence earnings are higher potentially or because partner earns more you can afford the partner not working.
