4 ms·

I'm curious about the higher level issues at play as well, which I'm just guessing about since I only have a lay understanding of the field Seems like there's

by kmod 4y ago

I'm curious about the higher level issues at play as well, which I'm just guessing about since I only have a lay understanding of the field Seems like there's probably a "winners curse" variant: conditioned on a bid being accepted, it is likely to be underestimated even if unconditioned there's an equal probability of under and over resonating Also seems like there's an incentive alignment problem as well: there's a clear benefit to bidding lower, and if the costs of underestimating are not born by the bidder then you would expect to see systematic underbidding even if we have on-average-accurate ways of estimating. These feel like important issues that are orthogonal to our ability to estimate things accurately and I wonder how much of the effect is from which type of cause (I wish the author had talked about this)

jaclaz 4y ago
From experience, what you mention is part of the perverse effects of the bid mechanisms, but these have nothing to do with the initial underestimation and is comparatively minor in relevance. Something is designed (years time), then it is estimated (often wrongly, however it takes some time as well), then it is approved/funded (some more years). But, IF the design was valid AND the estimation was correct at the most you could adjust the total amount by inflation/increases of prices. So, pure theory, you calculate a cost of 100 units (million dollars, whatever), then, since a few years have passed you should add (say) 10% for inflation/price adjustment before proceeding to a public bid (but it is extremely rare that this is actually done). The construction firms/contractors will offer between 15% and 25% rebate, if the bid is (as often is) given at max rebate, the one with 25% rebate will get the job. Now, assuming that the estimation is perfectly accurate (it isn't at all) the winner is (on the first day) under for at least 20%, 10% (the price adjustment, if included) and another 10% (their profit), while the remaining 5% is debatable, in the sense that it is what is actually possible that one construction company can save when compared to another for a number of reasons (machinery, organization, whatever). So, from day one the winner has two possible ways out, the first is to actually optimize the work to recover this 20-25% (difficult) the second is to find errors in the project (relatively easy). But, once the project is analyzed, what is found is not just the "missing" 20 to 25% (which if spent would not make the project over budget) but more likely something like 100%. So, at the end of the day, you are going to spend 175 instead of 100, of which 15 is connected to the perversions of the bidding, but 60 is errors in the project or in its estimation.
- staunton 4y ago
  You don't explain why the estimation itself (error of "60") is always an underestimate. The article does discuss this but your comment sounded to me like it was meant to be self-contained.
  - jaclaz 4y ago
    I was only trying to expand on the article with an example, the article is quite good in evidencing the common reasons why. In theory every single price is estimated analitically adding to market costs for materials and labor a percentage (usually between 13 and 17 %) and on this total another percentage for the profit of the contractor (usually 10 %)[1]. The quantities are calculated geometrically. So in theory the estimation quantities by price should be exact or within a very small approximation to reality. But in practice this sometimes (often) doesn't happen, the main reasons (cited in the article) are time passed since the estimation, optimism about ground/soil conditions, use of "wrong" price lists (and more generally inexperience of the estimator or Law/bureaucratic restrictions), optimism about weather, optimism about interferences with other infrastructures, change of building code, added features, actual mistakes in the project or in the contract, the list is long, but all the items in it tend to be "negative" ones implying more time and money needed for the construction. Of course there are also many projects that are fine, properly estimated and that are built in time and within budget but essentially all the estimation, even if done in good faith and by experienced technicians, are predictions (and predictions about the future, the most difficult ones). A better, more accurate process during designing and estimating (the 3 to 5% talked in the article) might help, but likely it would be not enough or reduce the incertainty by the same 3-5% that is spent for it. To all the above you add what is a sort of the S.E.P. (someone else's problem) syndrome, the one(s) that make the project and the estimation, by the time the site is ready to start are already elsewhere, in another job or another project, it is not unlike what happens nowadays in many industries, a new CEO is chosen, he does some cuts on costs, changes something in the products, and after two years, by the time the effects of the changes become actually evident he leaves and someone else starts again. [1] it depends on countries, but usually what are used are pre-made price lists that are a (supposed) average for a given work and are either outdated or do not apply to the specific work
acover 4y ago
There are governments that don't open bids that don't meet basic requirements to succeed.

by kmod 4y ago

jaclaz 4y ago

From experience, what you mention is part of the perverse effects of the bid mechanisms, but these have nothing to do with the initial underestimation and is comparatively minor in relevance. Something is designed (years time), then it is estimated (often wrongly, however it takes some time as well), then it is approved/funded (some more years). But, IF the design was valid AND the estimation was correct at the most you could adjust the total amount by inflation/increases of prices. So, pure theory, you calculate a cost of 100 units (million dollars, whatever), then, since a few years have passed you should add (say) 10% for inflation/price adjustment before proceeding to a public bid (but it is extremely rare that this is actually done). The construction firms/contractors will offer between 15% and 25% rebate, if the bid is (as often is) given at max rebate, the one with 25% rebate will get the job. Now, assuming that the estimation is perfectly accurate (it isn't at all) the winner is (on the first day) under for at least 20%, 10% (the price adjustment, if included) and another 10% (their profit), while the remaining 5% is debatable, in the sense that it is what is actually possible that one construction company can save when compared to another for a number of reasons (machinery, organization, whatever). So, from day one the winner has two possible ways out, the first is to actually optimize the work to recover this 20-25% (difficult) the second is to find errors in the project (relatively easy). But, once the project is analyzed, what is found is not just the "missing" 20 to 25% (which if spent would not make the project over budget) but more likely something like 100%. So, at the end of the day, you are going to spend 175 instead of 100, of which 15 is connected to the perversions of the bidding, but 60 is errors in the project or in its estimation.

staunton 4y ago
You don't explain why the estimation itself (error of "60") is always an underestimate. The article does discuss this but your comment sounded to me like it was meant to be self-contained.
- jaclaz 4y ago
  I was only trying to expand on the article with an example, the article is quite good in evidencing the common reasons why. In theory every single price is estimated analitically adding to market costs for materials and labor a percentage (usually between 13 and 17 %) and on this total another percentage for the profit of the contractor (usually 10 %)[1]. The quantities are calculated geometrically. So in theory the estimation quantities by price should be exact or within a very small approximation to reality. But in practice this sometimes (often) doesn't happen, the main reasons (cited in the article) are time passed since the estimation, optimism about ground/soil conditions, use of "wrong" price lists (and more generally inexperience of the estimator or Law/bureaucratic restrictions), optimism about weather, optimism about interferences with other infrastructures, change of building code, added features, actual mistakes in the project or in the contract, the list is long, but all the items in it tend to be "negative" ones implying more time and money needed for the construction. Of course there are also many projects that are fine, properly estimated and that are built in time and within budget but essentially all the estimation, even if done in good faith and by experienced technicians, are predictions (and predictions about the future, the most difficult ones). A better, more accurate process during designing and estimating (the 3 to 5% talked in the article) might help, but likely it would be not enough or reduce the incertainty by the same 3-5% that is spent for it. To all the above you add what is a sort of the S.E.P. (someone else's problem) syndrome, the one(s) that make the project and the estimation, by the time the site is ready to start are already elsewhere, in another job or another project, it is not unlike what happens nowadays in many industries, a new CEO is chosen, he does some cuts on costs, changes something in the products, and after two years, by the time the effects of the changes become actually evident he leaves and someone else starts again. [1] it depends on countries, but usually what are used are pre-made price lists that are a (supposed) average for a given work and are either outdated or do not apply to the specific work

acover 4y ago

There are governments that don't open bids that don't meet basic requirements to succeed.