Does Climate Change Bolster the Case for Fishery Reform in Asia?

Global fisheries have diverged sharply over recent decades. High governance, wealthy economies have largely adopted output controls or various forms of catch shares, which has helped fisheries in these economies overcome inefficiencies arising from overfishing (Worm et al. 2009) and capital stuffing (Homans and Wilen 1997), and allowed them to turn the corner toward sustainability (Costello, Gaines, and Lynham 2008) and profitability (Costello et al. 2016). But the world's largest fishing region, Asia, has instead largely pursued open access and input controls, achieving less long-run fishery management success (World Bank 2017). Recent estimates show that many Asian fisheries continue to languish under outdated management regimes and could benefit from economically optimized fishery management systems such as catch shares. World Bank (2017) estimates that Asian fisheries lose $55 billion per year in inefficient management, which accounts for 65% of the estimated global loss of $85 billion. Figure 1 shows the potential gains from catch shares in the nine economies with the largest economic surplus, all of which are in Asia.

All of the aforementioned benefits of fishery reform were calculated assuming a stationary environment. Yet, climate change promises to dramatically alter the productivity and spatial distribution of most Asian fish stocks (Molinos et al. 2016). These climate-induced changes are expected to play out over the next 100 years or more, but are already starting to take hold. For example, range shifts have been noted in several of the world's oceans, coral bleaching appears to be accelerating, and the productivity of many stocks has sharply changed in recent years. These findings raise an important dilemma for Asian economies interested in the long-run sustainability, food security, and profitability of their fisheries: Should they aggressively pursue fishery management reforms in advance of the most serious predicted effects of climate change? Or does the prospect of climate change weaken the case for reforms such that aggressive reform is no longer necessary?

To shed light on this dilemma, I join newly available data on Asian fishery status with state-of-the-art climate forecasts and bioeconomic models. I largely draw on data and methods in Gaines et al. (2018), though that paper does not single out any results for Asian fisheries, nor does it ask whether the case for reform is strengthened or weakened under climate change. This allows me to conduct a species-by-species analysis for 193 species of the most widely harvested fish in Asia, representing about 29 million metric tons in fish catch.¹

I begin by estimating biological status and trends for each of these species; this is accomplished by combining retrospective regression approaches (Costello et al. 2012) with dynamic structural models (Martell and Froese 2013). I then use these data as inputs into a bioeconomic model that estimates the potential benefits—in terms of fish conservation, fishery profit, and fish catch—from adopting economically efficient fishery management practices in Asia in the absence of climate change. Essentially, this involves comparing projected fishery performance under business-as-usual (BAU) management with fishery performance under economically optimized management.² Results of that analysis largely corroborate previous findings. But because I am primarily interested in how climate change affects these calculations, I then couple to this analysis projections of climate effects on each of the species in my data set from Molinos et al. (2016). These climate models suggest that about 55% of Asian fisheries will experience reductions from climate change, and 29% will experience significant range shifts in the coming decades. By combining the fishery status, models, and climate effects, I can then estimate the potential benefits from adopting fishery management reforms in the face of climate change. Naturally, this involves solving for the economically optimal feedback control rule in each fishery. The final step is to ask whether the strong case for fishery reform is maintained, or undermined, in a future with significant climate change.

Overall, the strong case for fishery management reform is maintained in a world with significant climate change.³ For the median fishery, both the economic and food provision cases for reform are slightly strengthened by climate change (though by less than 1 percentage point). However, because the effects are not symmetric, the aggregate case is somewhat weakened (by about 3 percentage points for harvest and 4 percentage points for value). While these results suggest that Asian fisheries would still do well to hasten the transition to economically optimized fishery management, they also point to substantial heterogeneity across fisheries due to differences in (i) current status of fish populations, (ii) BAU management, (iii) the biological effects of climate change, and (iv) anticipated geographic movement under climate projections. Taken together, these results suggest that for many Asian fisheries, climate change will strengthen the case for management reforms. But in some cases, I find that the case gets substantially weaker; in these places, motivating governments to undertake costly reforms will have to rely on other arguments or sources of reform capital.

The rest of this paper is organized as follows. Section II discusses the status and trends of major Asian fisheries, and their management. Section III provides theoretical guidance about the conditions under which climate change might strengthen, or weaken, the case for fishery management reform. Section IV then focuses on the empirical estimates of the effects of climate change on Asian fisheries. The estimates of reform with and without climate change are presented in section V. Finally, section VI concludes.

Official data from the Food and Agriculture Organization (FAO) show a surprising and underrecognized trend in Asian versus non-Asian fish catch. While global catch has been relatively constant over the past few decades (at approximately 80 million metric tons per year), the fraction of global catch produced in Asia has steadily increased (Figure 2). Over the past 5 years, Asian catch has surpassed the rest of the world combined, which represents a dramatic feat for a region focused intently on increasing protein production from the sea (Cao et al. 2017). Yet, questions remain about the underlying reasons for this dramatic divergence in trends between Asia and non-Asian regions. The most common explanation is that Asia's catch is being propped up by increasingly aggressive fishing efforts. Under this explanation, fisheries are progressively being overfished and will eventually collapse. The second possibility is that many large Asian fisheries are thought to have fished-down their immense stocks of predatory fish and that this allows for a “predatory release” (Szuwalski et al. 2016). Under this explanation, catches of smaller-bodied fish can be sustained at a much higher level than was previously thought because their predator numbers have been reduced. But owing to the immense diversity in Asian fish species, fishery management institutions, and economic conditions, the truth is almost certainly somewhere in between.⁴ The model I use here will not allow us to distinguish between these underlying causes, but it will allow us to track the likely species-by-species consequences of climate change on Asia's fisheries.

Drawing concrete conclusions about Asian fisheries is significantly hampered by the paucity of evidence on the biological status and trends for species of fish harvested in Asia. While individual economies conduct some scientific surveys (Melnychuk et al. 2017), almost no Asian fisheries conduct or report stock assessments. While Asian fisheries supply over half of the global fish catch, among the more than 500 fish stocks represented in the global Ram Legacy Stock Assessment Database (Ricard et al. 2012), only about 1% are from Asia. To overcome these extreme data gaps, recent contributions have provided data-poor methods for estimating stock status and backing out the fishing mortality rate that is implied by reported fish catches. Costello et al. (2016) merge methods from Costello et al. (2012) and Martell and Froese (2013) to estimate current biological status (biomass of a fish stock relative to its biomass under maximum sustainable yield (MSY), denoted as $B/BMSY$⁠) and current fishing mortality rate (as a fraction of fishing mortality under MSY, denoted as $F/FMSY$⁠).

For an estimate of the current status of Asian stocks, we follow Gaines et al. (2018) and aggregate fisheries from Costello et al. (2016) at the species level, and extract species whose geographic range extends into Asian waters. Here I make some brief comments about the data underlying this analysis. Biomass and fishing mortality estimates are derived using a panel regression model (Costello et al. 2012) as priors for a structural model from fishery science called the Catch–MSY method (Martell and Froese 2013). This model also provides estimates of the biological parameters for each individual stock, which are then aggregated at the species level for species known to exist in Asian waters. Catch data are from FAO (2014) and the Ram Legacy Stock Assessment Database (Ricard et al. 2012). Price and cost parameters are species-level aggregations from Costello et al. (2016); the resulting database of global fish prices has been published in Melnychuk et al. (2016) and cost parameters are derived to rationalize the level of fishing observed as formalized in Costello et al. (2016). The relevant climate data, which describe the spatial footprint of fish species now and in the future under alternative climate scenarios, are from Molinos et al. (2016), who estimate the change in ocean temperatures over time and associate that with species’ temperature preferences to estimate the geographic range of a species in the future. After filtering for the species that reside in Asian waters, this leaves us with 193 species-level bioeconomic models with biological parameters, spatial distributions, and changes in each over time under different climate scenarios.⁵

The resulting 193 Asian fish species are displayed in Figure 3, where bubble size indicates the potential size of the species’ fish catch (MSY) and shading foretells the future climate effects estimated from the climate model that will be described later (lighter shade for positive effects on carrying capacity and darker shade for negative effects on carrying capacity).⁶ Using this approach, the median values for $B/BMSY$ and $F/FMSY$ are both near 1; this may initially suggest that Asian fisheries are in reasonable condition. But a closer inspection of Figure 3 reveals a stark contrast between two classes of fisheries. Those in the top left of Figure 3 are in poor condition. According to this model, these fisheries have been overfished, driving their biomass below levels that which would maximize food provision, and they continue to be fished at an excessive rate.⁷ Many of the medium-sized and large Asian fisheries (bubble size), and the fisheries that will be negatively impacted by climate change (darker shade), are in this region of the figure. The second major group consists of fisheries in the bottom right of Figure 3. These fisheries appear to be underfished, at least so far as food production is concerned. Many of these biologically abundant species are expected to be positively affected by climate change. When combined, these features suggest that there may be important possibilities for future growth in some of these fisheries.

This biological growth equation (known as the Pella–Tomlinson model) contains three parameters: (i) g, which is related to the maximum (or “intrinsic”) growth rate of the stock; (ii) ϕ, which governs the skewness of the familiar hump-shape of growth function; and (iii) $Kt$⁠, which is the carrying capacity of the stock.⁹ This functional form is quite general and nests two familiar examples. First, in equation (2), I have allowed the carrying capacity $Kt$ to vary over time; in this paper, $Kt$ reflects the climate state in year t. For example, if climate change is expected to reduce the overall suitable geographic range of a stock by 2% per year, I follow Gaines et al. (2018) and interpret this as a change in carrying capacity (so $Kt$ declines by 2% per year). This interpretation of carrying capacity allows climate impacts to have year-by-year effects on fish stock growth. Second, the special case where $φ=1$ delivers the familiar logistic growth equation (with carrying capacity $Kt$ and intrinsic growth rate 2g).

Naturally, the consequences of climate change on any given fishery will hinge not only on the environmental effects, but also on the way in which the fishery is managed. As a measure of the economic benefit of fishery reform without climate change, I calculate the net present value (NPV) under the BAU fishing mortality rate (again without climate change); denoted as $F¯-CC$⁠, these are the “Fishing Pressure” values in Figure 3 and are compared to the NPV under economically optimized fishery management, denoted as $F-CC*Bt$⁠. To calculate the optimized feedback control rule, $F-CC*Bt$⁠, I use a discrete-time dynamic programming approach, with numerical-value-function iteration and backward induction using $Kt=K0$⁠, thus assuming that climate change is not occurring. I work backward until the value and policy functions converge. I then forward simulate using the converged policy function from the starting conditions shown in Figure 3 to obtain $V¯-CC$ (NPV under BAU without climate change), $V-CC*$ (optimized NPV without climate change), $H¯-CC$ (cumulative harvest 2012–2100 under BAU without climate change), and $H-CC*$ (cumulative harvest 2012–2100 under optimized NPV without climate change).

In a similar manner, when climate change is present, I calculate the NPV and harvest under a BAU policy and an optimized policy. But which policies to use? For the optimized policy, since $Kt$ can change each year in the climate change scenario, it could be treated as a state variable, which would give rise to a policy function that conditioned on $Kt$ (as well as $Bt$⁠). Under that fully adaptive assumption, the fully optimal policy function would take the form $FCC*Bt,Kt$⁠, so effectively there would be a different optimized harvest control policy function every year that fully anticipated the future effects of climate change. While this may seem farfetched, it would provide a useful benchmark because it would represent the highest possible NPV that any fishery could attain under climate change. But, I do not conduct this additional optimization for three reasons. First, doing so would presume that the fishery manager had perfect foresight about climate effects in all fisheries over the next 80 years and was able to perfectly reoptimize her policy function every year in anticipation of those changes. This seems implausible because of information and policy constraints that often prevent such nimble policy responses. The second reason is that I have conducted this optimization for three fisheries (representing the 5th, 50th, and 95th percentiles of change in K due to climate change) and found that it makes almost no difference in the ultimate NPV of the fishery. The percentage increases in value from using the $FCC*Bt,Kt$ policy instead of the $FCC*Bt$ policy are 0.36%, 0.001%, and 0.02%, respectively, for the 5th, 50th, and 95th percentile fisheries; the commensurate differences in aggregate harvest in 2012–2020 are 3.9%, 0.01%, and 0.04%, respectively.¹⁰ The final reason is that conducting this optimization for all 193 fisheries is very time consuming.

For these reasons, I continue to use the same optimized feedback control rule derived above, so $FCC*Bt=F-CC*Bt$ from the dynamic programming value function iteration procedure described above. For the BAU policy under climate change, I allow for the possibility raised in Gaines et al. (2018), who argue that range shifts induced by climate change could lead to institutional failures that increase fishing pressure. At the same time, it seems irrational to assume that fishing would extend beyond what is economically viable.

To capture these features, I analyze two different models of BAU fishing pressure (Table 1). In both models, BAU fishing pressure is initially $F¯-CC$ (as in the case without climate change). In the first model, I assume fishing pressure for shifting stocks gradually shifts to the open access level of fishing pressure (for which economic profit is zero in steady state) over time as range shifts take hold. In the second model, I assume BAU fishing pressure is unaffected by climate change, so $F¯CC=F¯-CC$ forever.

Across these models, I evaluate two different measures of fishery performance. The first is the NPV of the fishery from 2012 to 2100, and the second is the cumulative harvest over the same time period. For any given fishery, these values will depend on the starting conditions (Figure 3), policy function (Table 1), and climate change impact on carrying capacity (Figure 6).

This paper seeks to determine the first differences:

• •

  percentage loss from failing to optimize management without climate change: $ΔΩ-CC≡Ω-CC*-Ω¯-CCΩ-CC*HCC*(4)$

• •
  percentage loss from failing to optimize management with climate change:

  $ΔΩCC≡ΩCC*-Ω¯CCΩ-CC*$

  (5)

For example, if ΔV = 5 percentage points for a particular fishery, this would indicate that the case for fishery reform is 5 percentage points stronger in a world with climate change than it is in a world without climate change. Of course, we expect this statistic to be positive for some fisheries and negative for others. These values are represented in Figures 9 and 10 below.

Does theory provide any guidance about how we might expect climate change to affect the value of fishery management optimization? First, whether or not climate change occurs, we expect that optimizing the management of a fishery will lead to an increase in economic value. In other words, we expect $ΔV-CC>0$ and $ΔVCC>0$⁠. And while we generally expect fishery profit and fishery catch to go hand-in-hand, fishing costs (c in equation 1) imply that it is possible for an intervention to increase profit but decrease catch.¹² But as a general rule, we expect $ΔH-CC>0$ and $ΔHCC>0$ for most fisheries; when these values are negative, we expect them to be small in absolute value.

But how will $ΔV-CC$ and $ΔVCC$ compare with each other? In other words, calculating ΔV ≶ 0 will determine whether the presence of climate change increases or decreases the economic case for fishery management reform. Similarly, calculating ΔH ≶ 0 will determine whether the presence of climate change increases or decreases the food production case for reform. While the answers will turn out to depend on current conditions, BAU management, and the dynamic effects of climate change for any particular fishery, some broad generalizations are possible. First, for fisheries that will experience a reduction in carrying capacity from climate change, it is reasonable to expect a reduction in both the maximum fish catch and the maximum value of the fishery. For example, Figure 4 shows the production function for a fishery with the median global parameters, where the solid line uses current parameters (no climate change) and the dashed line assumes a 60% reduction in carrying capacity resulting from climate change. Production with climate change is everywhere below production without climate change, reflecting the reduction in carrying capacity. This logic seems to suggest that fisheries that will suffer reductions in carrying capacity are likely to gain less from management reform than are fisheries that will experience increases in carrying capacity.

But this logic turns out to depend on the current level of fishing pressure. If BAU fishing pressure is very high (e.g., if the fishery is in open access equilibrium), then the logic holds firmly because the fishery is currently experiencing low (or zero) profit and low catch. Optimizing such a fishery eventually brings about positive increases whether or not climate change occurs, but climate change increases the case for reform only if it will increase carrying capacity. Alternatively, if BAU fishing pressure is very low (take the extreme case when it is zero), then the logic also holds because under BAU both profit and catch are low (or zero). But for intermediate levels of fishing pressure, it turns out that the logic can break down. Figure 5 shows the increase in steady-state profit (again for a fishery with globally median parameters) that arises from fishery management reform as a function of BAU fishing pressure. The solid line depicts the upside from reform without climate change and the dashed line depicts it with a 60% drop in carrying capacity resulting from climate change. To see how a deleterious climate shock could actually increase the benefits from reform, consider the following example. Suppose BAU fishing pressure is about 0.9, which is near the profit-maximizing fishing pressure in the absence of climate change (solid line in Figure 5). In that case, the economic upside from reform (in the absence of climate change) is near zero. But how does the economic upside from reform change after a deleterious climate shock (dashed line)? After climate change, the optimal level of fishing pressure declines (to about 0.75) and the upside from reform, given that BAU fishing pressure is 0.9, is nonnegligible. This example is simply meant to illustrate the possibility that a negative climate shock does not necessarily imply a lower benefit from fishery management reform.

The bioeconomic models I apply to Asian fisheries are substantially more complicated than the simple illustrative examples from Figures 4 to 5. The effects of climate change play out over time, starting conditions differ across fisheries, BAU and optimized policies have effects that evolve over time, and optimal policies are dynamically (not statically) optimized. While the intuition provided above can provide some guidance, it is ultimately an empirical question whether the presence of climate change will strengthen or weaken the case for management reform in any given Asian fishery.

Following Gaines et al. (2018), the myriad effects of climate change on global fisheries can be distilled into two categories. The first, and most widely studied, is that climate change may alter the stock growth of a fishery, which is often interpreted as a change in carrying capacity. This can occur through changes in prey abundance, ocean temperatures, acidification, or via other mechanisms. The second consequence of climate change is that it can alter the spatial range of a species in the ocean. Even in the absence of carrying capacity changes, range shifts can have significant consequences to fishery sustainability because as a fish stock crosses international boundaries, institutional failures can lead to overexploitation.

I thus assume that the major effects of climate change can be captured by changes in carrying capacity and shifts in geographic range over time. Changes in the total geographic area suitable for a species correspond to changes in carrying capacity over time. Figure 6 shows the individual trajectories of carrying capacity for each of the 193 species in this analysis under a representative concentration pathway 6.0 climate change scenario; among these species, 55% will decline in carrying capacity and 45% will increase. The line thickness in Figure 6 corresponds to MSY (so it varies over time for each stock, in accordance with changes in $Kt$⁠) and the shading corresponds to scaled $Kt$⁠; all values are relative to the 2012 value. Among the species studied, 61% will experience a reduction in carrying capacity and/or significant range shifts as a result of climate change; the other 39% will experience positive effects.

To capture these effects of climate change on a species’ carrying capacity and range, we must somehow translate them into the bioeconomic model presented above. First, we keep track of the carrying capacity in each year for each species (Figure 6), and this becomes an input into the model itself (see equation 2). To capture range shifts, we make no changes to the model when a stock is a stationary stock (that is, when it stays within a country's waters). But for the 29% of species that are shifting stocks, I run two scenarios. In the first scenario, the BAU fishery policy gets progressively worse as these transboundary shifts start to take hold. In the second scenario, fishing pressure for shifting stocks is unaffected by climate change. All of these assumptions are summarized as follows:

• •

  Changes in fish stock growth

  • –

    Changes in carrying capacity, K, over time: $Kt$ (Figure 6) is an input to the biological model (equation 2) and thus to the forward simulations.

  • –

    BAU policy under climate change: fish at the current fishing mortality rate (except for shifting stocks, see below)

  • –

    Optimized policy under climate change: use the dynamically optimized harvest control rule under current conditions.

• •

  Range shifts

  • –

    “Stationary stocks” have policy functions as indicated above.

  • –

    “Shifting stocks,” or those that are expected to cross significantly into multiple jurisdictions (Gaines et al. 2018), are treated as follows:

    • ·

      Under BAU, the initial fishing mortality rate is the current fishing mortality rate. It either gradually transitions to the fishing mortality rate under open access according to when the shifts are expected to occur, or it is maintained at the current fishing mortality rate; both scenarios are examined below.

    • ·

      Under optimized management, the harvest policy is optimized (under current conditions), so range shifts are internalized into the policy.

Detailed information on fishery management in Asia is extremely hard to come by. Most available evidence suggests that fishery management institutions are somewhat outdated and rely heavily on input controls such as season length; gear restrictions; and, in some cases, limited licenses. But there seem to be very few cases of feedback control rules, such as harvest control rules, that are now the backbone of fishery policy in Australia, Canada, the United States, and much of Europe and Latin America. I use the model described above to estimate the economic and food provision benefits of adopting fishery management reforms in Asian fisheries.

In the absence of climate change, the benefits of management reforms vary by fishery, but adopting economically rational fishery management generally increases both cumulative harvest (horizontal axis of Figure 7) and economic value (vertical axis of Figure 7) relative to BAU. The average effect of implementing optimized fishery management is expected to increase catch by about 24% and economic value by 34%, though these values range widely across fisheries. The comparable results in a world with climate change are shown in Figure 8, where the shading refers to whether climate change is expected to have a positive (lighter shade) or negative (darker shade) effect on fish stock growth. With climate change, the benefits of reform are still large (visually, there is little difference between Figures 7 and 8). But the average effects of reform are slightly muted here (reform increases catch by 21% and economic value by 30%). The next section explicitly focuses on the difference between these two sets of results.

The main question this paper seeks to ask is: does climate change undermine the case for fishery management reform in Asia? I conclude with an emphatic “no.” Perhaps the best evidence is from Figure 8, which shows that there remains a large benefit of fishery management reform in nearly all Asian fisheries despite the onset of climate change. A more nuanced question is: does climate change strengthen or weaken the case for fishery management reform? Essentially, this amounts to the difference between Figure 8 and Figure 7, which is depicted in Figure 9 as percentage point changes for each individual fishery.

For stationary stocks (triangles in Figure 9), climate change only affects carrying capacity (it does not affect BAU management). For these stocks, the intuition provided in section III was that carrying capacity increases and the case for reform typically go hand-in-hand. Indeed, this seems to be the case for Asian fisheries: those for which carrying capacity shocks will be positive (lighter triangles) tend to have a stronger case for reform (in both harvest and economic value), and those for which carrying capacity shocks will be negative (darker triangles) tend to have a weaker case for reform. For stationary stocks, the overall conclusion is that climate change will generally bolster the case for fishery management reform in Asia.

But the story can be considerably different for Asian stocks for which we anticipate future range shifts resulting from climate change (circles in Figure 9, which reflect the assumption that BAU fishing pressure gradually shifts to open access for shifting stocks). For those stocks, climate change induces potentially devastating institutional failure, which drives a possibly large wedge between the value of the fishery with and without reform. This complicates the calculus. While many shifting stocks are also negatively affected by climate change, the case for reform can either be strengthened (circles in upper right of Figure 9) or weakened (circles in lower left of Figure 9) by the onset of climate change. Taken together, these results suggest that despite climate change, the case for fishery reform remains strong in Asia, though the case can be weakened for some stocks.

To test the importance of the BAU assumption for shifting stocks, I repeat the same analysis for the alternative BAU scenario. In the results depicted in Figure 9, the BAU policy under climate change was for stationary stocks to continue at their current fishing mortality rate and for shifting stocks to transition to open access fishing pressure. The alternative is to treat shifting stocks in the same manner as stationary stocks (so they maintain the current fishing mortality rate). In that case, the basic story stands but the case for reform is even stronger. In both the no-climate change and the climate change scenarios, the benefits of reform are 21%–24% (increase in harvest from reform) and 30%–34% (increase in economic value from reform), suggesting that climate change does not dramatically alter the case for reform. The fishery-by-fishery results for this scenario are depicted in Figure 10, which conform to the theoretical expectation that the case for reform will generally be strengthened for stocks that experience a positive climate shock (lighter shade) and weakened for stocks that experience a negative climate shock (darker shade).

Returning to the original BAU assumption, we can aggregate the data underlying Figure 9 to the FAO fish category level to provide a glimpse into the types of fish for which climate change is likely to strengthen or weaken the case for fishery reforms (recognizing that the case for reform remains strong in nearly all cases). Table 2 reports ΔH and ΔV for the seven fish categories with MSY>1 million metric tons (reported as percentage point gains as a consequence of climate change). These data suggest that the case for reform is strengthened for the large class of cods, hakes, and haddocks, but weakened (sometimes substantially) for other groups such as salmon and smelts. Some groups show the interesting pattern that the case for harvest is weakened but the case for economic value is strengthened (e.g., herrings, sardines, and anchovies). The table also provides the number of species composing each category and measures of fishery size (MSY) and overall biomass (BMSY). Four of the five largest classes of fish are expected to have a stronger economic rationale for reform with climate change than without climate change.

The focus of this paper has been on whether climate change undermines the case for fishery management reform in Asia. While the Asia-wide answer is “no,” the answer for any given species turns out to hinge on the exact manner in which climate change will influence the species. For sedentary stocks, the main effect of climate change is on the carrying capacity, and thus the overall growth of the fish stock. If the carrying capacity of a stock is expected to decline under climate change, then the case for fishery reform is generally weakened; the opposite holds for cases when the carrying capacity will increase in the future. While the model results support this prediction, the weakening of the case for reform is quite small (less than 5 percentage point changes), even when climate change will have deleterious effects. The other significant implication of climate change, which has largely gone unnoticed by the previous literature, is that the ranges of some stocks will change. When fish stocks move into new jurisdictions, this can cause a race to fish and may result in worse outcomes than if the same stock had not crossed a jurisdictional boundary. Fisheries for which this second effect is present see a much wider range of outcomes, which largely hinge on how aggressively they are currently managed.

Overall, these results suggest that the vast majority of Asian fisheries, including its largest ones, would benefit economically and in terms of food security by engaging in fishery management reforms. Across Asia, I find that such reforms could lead to increases of 30% in the present value of fisheries and 21% in food provision, even under impending climate change. This implies that Asian fisheries should hasten the transition to sensible, economically rational fishery management under current climate conditions; this will simultaneously secure food and livelihoods across Asia's diverse fisheries, even in the face of climate change.

This table contains the scientific and common names for each of the 193 species in the Asia data set used in this paper. Among these are several names of economies, which represent aggregated species according to the Food and Agriculture Organization's not elsewhere included (nei) category (e.g., Singapore nei 110).