Lecture # 22 Cst-Benefit Analysis I. Intrductin t Cst-Benefit Analysis (CBA) Public fficials ften use cst-benefit analysis t decide whether a prject is wrthwhile. Cst-benefit analysis can be used t guide several decisins: Is a prject wrth ding? D the benefits utweigh the csts? Given alternative prjects that culd achieve the same gal, which prject shuld be undertaken? What is the apprpriate scale f the prjects t be undertaken? Lgic f CBA analysis: Calculate the ecnmic benefits and ecnmic csts f each alternative. These csts and benefits are discunted ver time t reflect pprtunity values. Chse prjects with the highest net benefit. Nte: wrks fr a discrete set f chices, nt a cntinuum. Fr cntinuus chices, use marginal analysis. The ultimate gal f CBA is t ensure that sciety s resurces are put t their mst highly valued uses. That is, scarcity is a cncern even fr gvernment fficials. We want t make the best use f the resurces available t us. CBA is ften criticized as being the final wrd and fr putting t much emphasis n mnetary values. Hwever, it des have a rle t play in guiding plicy decisins. If thught f as a guide t the final decisin, rather than a means f prviding definitive answers, CBA can be a useful plicy analysis tl. CBA required by federal law fr evaluating regulatins. In 1981, President Reagan required CBA fr all prpsed rules befre published in the Federal Register. Regulatry actin shuld nt be undertaken unless the ptential benefits t sciety f the regulatin utweigh the ptential csts f the regulatin. In 1993, President Clintn mdified this requirement t refer t nly t significant regulatry actin In the upcming lectures, we will discuss: The steps t perfrming CBA The rle that CBA shuld play in plicy analysis
II. Steps t Cst-Benefit Analysis 1. Define the Situatin T begin, a cmmunity has a set f resurces, such as mney, prperty, labr, envirnmental amenities, and gvernment services. After a prgram is cmplete, the set f resurces will change. CBA lks at hw these resurces change. A gd analysis cnsiders alternatives At a minimum, ne alternative is the status qu -- ding nthing. Hwever, cnsidering prjects f multiple sizes helps us assess which size prvides the mst benefit. E.g. we culd build a 4, 6, r 8 lane highway. Cmparing benefits in each prvides infrmatin n the marginal benefits f adding mre lanes. Thus, we need t cnsider what cmmunity s resurces are at stake. Fr example, shuld neighbring twns be included in an analysis? (E.g. shuld visitrs frm nearby twns be cnsidered beneficiaries f a new municipal park?) There is n crrect answer t this questin the answer needs t be determined in the plitical arena. The jurisdictin ding the analysis matters. The csts f damages frm greenhuse gases differ if dne fr a single cuntry, rather than fr the entire wrld. 2. Estimate the Csts Next, we need t cnsider the cst f the prject, in terms f the resurces the cmmunity will use t carry ut the prject. Types f csts: 1. Resurces purchased fr the prject 2. The pprtunity cst f resurces already wned by the gvernment We need t knw the value f these assets in their next mst valuable use. 3. Resurces wned r purchased by lcal residents 4. External csts These csts can be harder t measure, but their values are imprtant. Later we will talk abut ways t value sme f these csts. It is best t include mnetary values fr such csts, s that cmparisns are pssible. Hwever, if mnetary values are nt pssible, the csts shuld still be listed, s that they can still be cnsidered in plitical debate.
T measure csts, we want t find the shadw price f the input used. The shadw price f an input r utput is its value t the ecnmy as a whle. Fr inputs, the shadw price is the marginal pprtunity cst the value f the best alternative use f the input. Shadw prices are related t the market price, but nt exactly the same because f market imperfectins. Cnsider, fr example, inputs prduced by a mnplist. In that case, P > MC. If the gvernment s purchase affects the amunt available fr private cnsumptin, it is the market price, which measures the value t cnsumers, that matters. Hwever, the pprtunity f using the gds is simply the MC. Thus, if the gvernment is prducing mre f the gd, it is the pprtunity cst (MC) that matters. Resurces frm the cmmunity are being taken frm ther uses. MC represents that pprtunity cst. When taxes are used t raise revenue fr a prject, we shuld cnsider the marginal excess tax burden. This accunts fr the csts f raising mney thrugh taxatin, including: inefficiency (deadweight lss) administrative csts It is imprtant t distinguish between csts and transfers. When resurces are nt used up nr created, but just shifted frm ne set f individuals t anther, we say the resurces are transferred. A transfer des nt change the ttal amunt f resurces in a cmmunity. It just changes the distributin f the resurces. Transfers d have equity implicatins, but we will cnsider them later
3. Estimate the Benefits Benefits are new resurces made available t the cmmunity. Ecnmists typically measure benefits by a cmmunity s willingness t pay. Willingness t pay the amunt that individuals are willing t pay t receive the benefits (r the csts they are willing t pay t avid negative effects). The demand curve tells us the marginal willingness t pay fr cnsumers. Ttal benefits are the area under the demand curve. This is equal t cnsumer surplus + ttal expenditures n the gd. Nte that it includes CS plus a bit mre because we are nly measuring benefits, nt csts. It is the ttal benefit, nt just the benefit abve and beynd the cst f prducing the gd, that we want here. Nte that, if there is market pwer, the price paid by cnsumers tells us the value f the gd t cnsumers. Therefre, the shadw price f benefits is the price paid by cnsumers. Types f benefits: Resurces measurable in mnetary terms Fr gds & services sld in the marketplace, their benefit is the market price f the gd. Resurces measurable in physical units, but nt in mnetary terms. We may be able t infer dllar values fr these benefits. Hwever, even if we can t place a dllar value n a benefit, infrmatin n the amunt f benefit is useful. We can ask whether the estimated csts are wrth this level f benefit. Nte that this requires persnal judgments. Resurces valued by the cmmunity, but nt measurable by any means. Yu shuld generally be suspicius f these claims.
Once again, it is imprtant t distinguish between benefits and transfers. Are jbs a benefit? Often, the emplyment pprtunities created by a prject are cnsidered benefits. Hwever, isn t emplyment really a cst f the prject? The key thing here is t distinguish between benefits and transfers. If the ecnmy is at full-emplyment, the pprtunity cst f using wrkers is that they are nt used smewhere else. Thus, the new jbs are simply a transfer f wrkers frm ne sectr t anther. Nte, fr example, hw bth sides ften use jbs gained r lst in their argument (e.g. jbs lst due t envirnmental regulatin vs. "green jbs.") This suggests that the labr market adjusts, with wrkers mving t new jbs as necessary. At the lcal level, this culd be mre difficult. If there is chrnic unemplyment in a regin, a prject in that regin culd prvide emplyment benefits t thse in the regin, even thugh there is n change in the natinal unemplyment rate. Thus, the scpe f the study is imprtant. A better way t deal with this issue is t adjust the csts f labr. If sme f the lcal wrkers are drawn frm the pl f unemplyed wrkers, the pprtunity cst f emplying them is less than the pprtunity cst f emplying ther wrkers.
Often, benefits r csts t be included in CBA are nt traded in the marketplace. Hw can these be valued? Use infrmatin n demand frm neighbring cmmunities. Surveys Ask peple hw much they are willing t pay fr services. Prblem: what peple say (stated preference) and what peple d (revealed preference) is nt always the same Use infrmatin n related markets Fr example, we can cmpare prperty values in neighbrhds with and withut the service. Cnsider tw huses with the same features. One is in a neighbrhd with clean air, and the ther is in a plluted neighbrhd. The huse in the cleaner neighbrhd will sell fr a higher price. The difference in prices is the value f the cleaner air. Nte use f ecnmic thery here. It allws us t infer values frm actual behavir. Example: the value f time Cnsider labr/leisure thery: At the ptimal pint, MULeisure/MUIncme = wage Therefre, wage equals value placed n leisure time. It is imprtant nt t duble cunt benefits. Fr example, the benefits f a new park are the willingness t pay fr the new recreatinal pprtunities. In additin, suppse that the new park increases prperty values in the neighbrhd by $95,000. The increased prperty values ccur because f the additinal recreatinal pprtunities. Listing them as an additinal benefit wuld duble-cunt the benefits f increased recreatin. Hwever, since there isn t a clear market fr recreatinal pprtunities, the $95,000 increase in prperty values culd be used t put a dllar value n thse pprtunities.